B R O O K I N G S PA P E R S O N

William C. Brainard and George L. Perry, Editors

2005

Ryan D. Nunn, Statistical Associate

Theodore Papageorgiou, Assistant to the Editors
Michael Treadway, Editorial Associate
Lindsey B. Wilson, Production Associate

BROOKINGS INSTITUTION

WA S H I N G T O N , D . C .
Copyright © 2005 by
THE BROOKINGS INSTITUTION
1775 Massachusetts Avenue, N.W., Washington, D.C. 20036

ISSN 0007-2303

ISBN-10: 0-8157-1350-9
ISBN-13: 978-0-8157-1350-0

Authorization to photocopy items for internal or personal use or the internal or personal use of specific
clients is granted by the Brookings Institution for libraries and other users registered with the Copyright
Clearance Center Transactional Reporting Service, provided that the basic fee is paid to the Copyright
Clearance Center, 222 Rosewood Drive, Danvers, MA 01923. For more information, please contact CCC at
(508) 750-8400. This authorization does not extend to other kinds of copying, such as copying for general
distribution, for creating new collective works, or for sale. Specific written permission for such copying
must be obtained from the Permissions Department, Brookings Institution Press, 1775 Massachusetts Avenue,
N.W., Washington, DC 20036; FAX (202) 797-6195; E-mail permissions@brookings.edu.
THE BROOKINGS INSTITUTION
The Brookings Institution is a private nonprofit organization devoted to research,
education, and publication on important issues of domestic and foreign policy. Its
principal purpose is to bring the highest quality independent research and analy-
sis to bear on current and emerging policy problems. The Institution was founded
on December 8, 1927, to merge the activities of the Institute for Government
Research, founded in 1916, the Institute of Economics, founded in 1922, and the
Robert Brookings Graduate School of Economics and Government, founded in
1924. Interpretations or conclusions in Brookings publications should be under-
stood to be solely those of the authors.

Board of Trustees
John L. Thornton Lawrence K. Fish Michael H. Jordan
Chair Cyrus F. Freidheim Jr. Harold Hongju Koh
Strobe Talbott Bart Friedman William A. Owens
President David Friend Frank H. Pearl
Zoë Baird Ann M. Fudge John Edward Porter
Alan R. Batkin Jeffrey W. Greenberg Steven Rattner
Richard C. Blum Brian L. Greenspun Haim Saban
Geoffrey T. Boisi William A. Haseltine Leonard D. Schaeffer
James W. Cicconi Teresa Heinz Lawrence H. Summers
Arthur B. Culvahouse Jr. Samuel Hellman David F. Swensen
Kenneth W. Dam Glenn Hutchins Larry D. Thompson
Vishakha N. Desai Joel Z. Hyatt Laura D’Andrea Tyson
Thomas E. Donilon Shirley Ann Jackson Antoine W. van Agtmael
Mario Draghi Kenneth Jacobs Beatrice W. Welters
Kenneth M. Duberstein Suzanne Nora Johnson Daniel Yergin

Honorary Trustees
Leonard Abramson Robert A. Helman J. Woodward Redmond
Elizabeth E. Bailey Roy M. Huffington Charles W. Robinson
Rex J. Bates James A. Johnson James D. Robinson III
Louis W. Cabot Ann Dibble Jordan Judith Rodin
A. W. Clausen Vernon E. Jordan Jr. Warren B. Rudman
William T. Coleman Jr. Breene M. Kerr B. Francis Saul II
Alan M. Dachs Marie L. Knowles Ralph S. Saul
D. Ronald Daniel James T. Lynn Henry B. Schacht
Robert A. Day Jessica Tuchman Mathews Michael P. Schulhof
Bruce B. Dayton David O. Maxwell Joan E. Spero
Charles W. Duncan Jr. Donald F. McHenry Vincent J. Trosino
Walter Y. Elisha Robert S. McNamara John C. Whitehead
Robert F. Erburu Mary Patterson McPherson Stephen M. Wolf
Henry Louis Gates Jr. Arjay Miller James D. Wolfensohn
Robert D. Haas Mario M. Morino Ezra K. Zilkha
Lee H. Hamilton Maconda Brown O’Connor
F. Warren Hellman Samuel Pisar
B R O O K I N G S PA P E R S O N

William C. Brainard and George L. Perry, Editors 2005

Editors’ Summary ix

Articles
OLIVIER BLANCHARD, FRANCESCO GIAVAZZI, and FILIPA SA
International Investors, the U.S. Current Account,
and the Dollar 1
Comments by Ben S. Bernanke and Hélène Rey 50
General Discussion 62
MAURICE OBSTFELD and KENNETH S. ROGOFF
Global Current Account Imbalances
and Exchange Rate Adjustments 67
Comments by Richard N. Cooper and T. N. Srinivasan 124
General Discussion 141
MICHAEL DOOLEY and PETER GARBER
Is It 1958 or 1968? Three Notes on the Longevity
of the Revived Bretton Woods System 147
Comments by Barry Eichengreen and Jeffrey A. Frankel 188
General Discussion 204
SEBASTIAN EDWARDS
Is the U.S. Current Account Deficit Sustainable?
If Not, How Costly Is Adjustment Likely to Be? 211
Comments by Kathryn M. E. Dominguez and Pierre-Olivier Gourinchas 272
General Discussion 282
DEAN BAKER, J. BRADFORD DELONG, and PAUL R. KRUGMAN
Asset Returns and Economic Growth 289
Comments by N. Gregory Mankiw and William D. Nordhaus 316
General Discussion 325
 Purpose Brookings Papers on Economic Activity contains the articles, reports,
and highlights of the discussions from conferences of the Brookings
Panel on Economic Activity. The panel was formed to promote pro-
fessional research and analysis of key developments in U.S. economic
activity. Prosperity and price stability are its basic subjects.
The expertise of the panel is concentrated on the “live” issues of
economic performance that confront the maker of public policy and the
executive in the private sector. Particular attention is devoted to recent
and current economic developments that are directly relevant to the
contemporary scene or especially challenging because they stretch
our understanding of economic theory or previous empirical findings.
Such issues are typically quantitative, and the research findings are
often statistical. Nevertheless, in all the articles and reports, the rea-
soning and the conclusions are developed in a form intelligible to the
interested, informed nonspecialist as well as useful to the expert in
macroeconomics. In short, the papers aim at several objectives: metic-
ulous and incisive professional analysis, timeliness and relevance to
current issues, and lucid presentation.
Articles appear in this publication after presentation and discussion
at a conference at Brookings. From the spirited discussions at the con-
ference, the authors obtain new insights and helpful comments; they
also receive searching criticism about various aspects of the papers.
Some of these comments are reflected in the published summaries of
discussion, some in the final versions of the papers themselves. But in
all cases the papers are finally the product of the authors’ thinking and
do not imply any agreement by those attending the conference. Nor
do the papers or any of the other materials in this issue necessarily rep-
resent the views of the staff members, officers, or trustees of the
Brookings Institution.

Correspondence Correspondence regarding papers in this issue should be addressed to

the authors. Manuscripts are not accepted for review because this jour-
nal is devoted exclusively to invited contributions.

Subscriptions For information on subscriptions, standing orders, and individual copies,

contact Brookings Institution Press, P.O. Box 465, Hanover, PA 17331-
0465. Or call 866-698-0010. E-mail brookings@tsp.sheridan.com. Visit
Brookings online at http://bookstore.brookings.edu.

Brookings periodicals are available online through both Online Com-

puter Library Center (contact OCLC subscriptions department at
800-848-5878, ext. 6251) and Project Muse (http://muse.jhu.edu).
 Panel Dean Baker Center for Economic and Policy Research
members and Olivier Blanchard Massachusetts Institute of Technology
staff: William C. Brainard Yale University
J. Bradford DeLong University of California, Berkeley
Michael Dooley University of California, Santa Cruz
Sebastian Edwards University of California, Los Angeles
Peter Garber Deutsche Bank
Francesco Giavazzi Universitá Commerciale Luigi Bocconi
Paul R. Krugman Princeton University
Maurice Obstfeld University of California, Berkeley
George L. Perry Brookings Institution
Kenneth S. Rogoff Harvard University
Filipa Sa Massachusetts Institute of Technology
________
Ryan D. Nunn Brookings Institution
Theodore Papageorgiou Yale University
Michael Treadway Brookings Institution
Lindsey B. Wilson Brookings Institution

Panel advisers Martin Neil Baily Institute for International Economics

participating Richard N. Cooper Harvard University
in the Benjamin A. Friedman Harvard University
seventy-ninth Robert J. Gordon Northwestern University
conference: N. Gregory Mankiw Harvard University
William D. Nordhaus Yale University
Edmund S. Phelps Columbia University

Guests whose Henry J. Aaron Brookings Institution

writings or Ben S. Bernanke Board of Governors of the Federal Reserve System
comments Kathryn M. E. Dominguez University of Michigan
appear in this Barry Eichengreen University of California, Berkeley
issue: Jeffrey A. Frankel Harvard University
Pierre-Olivier Gourinchas University of California, Berkeley
Gian Maria Milesi-Ferretti International Monetary Fund
Richard Portes London Business School
Hélène Rey Princeton University
T. N. Srinivasan Yale University
Editors’ Summary

The brookings panel on Economic Activity held its seventy-ninth con-

ference in Washington, D.C., on March 31 and April 1, 2005. This issue of
Brookings Papers on Economic Activity includes the papers and discussions
presented at the conference. The first four articles address the position of
the United States in the global economy, an increasingly controversial sub-
ject in the research, financial, and policy communities. Since the early
1990s, U.S. current account deficits have grown almost without interrup-
tion, reaching $666 billion, or about 6 percent of GDP, in 2004. The U.S.
international investment position is now one of net indebtedness approach-
ing 30 percent of GDP, and in recent years a substantial portion of the
buildup in net debt has come in the form of additions to dollar reserves by
foreign central banks. Some observers see the present situation as unsus-
tainable and warn of an abrupt depreciation of the dollar, which could
destabilize financial markets and disrupt the global economy. Others are
more sanguine, arguing that the present situation reflects the relative
strength of the U.S. economy, consumer and business preferences, and
rational financial decisions, all of which could evolve so as to make any
needed adjustments gradual.
Each of the four articles takes a different approach to analyzing the sit-
uation, focusing on issues that the authors see as key. The first article
models portfolio choices and how they moderate the pace of adjustment in
exchange rates and current accounts. The second stresses the relative price
changes that will be needed, both in the United States and abroad, to move
the U.S. current account toward balance. The third considers the motiva-
tions of policymakers in China and elsewhere for accumulating dollar
reserves. The fourth assesses the likelihood of an abrupt depreciation of the
dollar and the economic instability that might result in the United States
and abroad. The volume concludes with an article on the possible impact of
slowing labor force growth on stock market returns.
ix
 x Brookings Papers on Economic Activity, 1:2005

The u.s. international investment position is affected by develop-

ments in both foreign trade and international capital flows—the market for
imports and exports of goods and services and the market for foreign and
domestic assets. The sustainability of the U.S. current account deficit and
the consequences of reducing that deficit depend on features of both those
markets. Most economic models that have been used to analyze the cur-
rent account deficit assume imperfect substitutability between foreign and
domestic goods and services but perfect substitutability between foreign
and domestic assets. These assumptions carry strong implications for how
the economy adjusts to new developments. In the first article in this
volume, Olivier Blanchard, Francesco Giavazzi, and Filipa Sa provide a
distinctive analysis that allows for imperfect substitutability between
domestic and foreign assets and between domestic and foreign goods. With
this feature, movements in exchange rates and asset prices have poten-
tially important effects on the portfolios of international investors and
strong implications for the speed with which exchange rates adjust to
shocks. Compared with popular discussion and with earlier, simpler mod-
els, this rich specification provides a better understanding of past develop-
ments in the U.S. current account balance and the dollar exchange rate
and a more realistic framework for assessing future prospects.
In its simplest form the authors’ model has just two regions—the United
States and the rest of the world—each of which supplies interest-bearing
assets. The wealth of each region is given by the value of domestic assets
plus net claims on foreigners. Investors diversify their portfolios, holding
both foreign and U.S. assets, but exhibit home bias: given equal expected
returns, they place a larger fraction of their wealth in domestic than in for-
eign assets. As a result, a shift in wealth to foreigners reduces the demand
for U.S. assets, causing the dollar to depreciate. Similarly, an increase in
private or government demand for dollar assets causes the dollar to appre-
ciate. Because of imperfect substitutability, the relative returns on foreign
and U.S. assets can vary with changes in relative supplies or shifts in the
distribution of world wealth, and uncovered interest parity does not hold.
In the model the effects of a depreciation on the path of the current
account balance and changes in U.S. net foreign indebtedness are conven-
tional. The current account balance is the sum of the trade balance and net
interest earnings. Dollar depreciation improves both, immediately reducing
the dollar value of net interest payments and eventually reducing the U.S.
trade deficit. Changes in U.S. net foreign indebtedness reflect the sum of the
William C. Brainard and George L. Perry xi

current account balance and the revaluations of U.S. and foreign portfo-
lios that arise from exchange rate movements. In the real world, asset val-
ues and therefore net debt will also change with changes in domestic
interest rates, but the model ignores these so as to focus on exchange rate
movements, which are the key for understanding the model’s distinctive
implications.
Whereas the response of the current account in the model is quite famil-
iar, the effect of depreciation on asset demands is quite different than in
conventional models where assets are perfect substitutes. Depreciation of
the dollar reduces U.S. net indebtedness directly, increasing the dollar value
of foreign assets held in U.S. portfolios while decreasing the value of U.S.
assets in foreign portfolios. If assets were perfect substitutes, these changes
in portfolio shares would be of no importance, and the expected returns on
U.S. and foreign assets would always have to be equal. With fixed domes-
tic interest rates, the expected change in exchange rates would then be zero.
In such a world, real exchange rate changes are always unexpected. With
imperfect substitutability, in the absence of compensating changes in
expected relative rates of return, investors in both regions will want to
rebalance their portfolios following an unexpected exchange rate move-
ment. Thus an unexpected depreciation of the dollar in response to a trade
shock actually increases the relative demand for U.S. assets, reducing but
not reversing the depreciation. Unlike in the case of perfect substitutability,
the expected returns on U.S. and foreign assets do not have to be the same
after the initial adjustment. Rather than jump all the way to a new equilib-
rium from which no further change is expected, the dollar undergoes a
sharp initial, unexpected depreciation followed by a more gradual, expected
depreciation. The expected depreciation merely reduces the desired shares
of U.S. assets in investors’ portfolios rather than causing massive flight
from dollars. The rate at which the dollar depreciates after its initial
response to an adverse shock depends on the elasticity of asset demands
with respect to the relative rates of return: the lower the elasticity, the more
gradual the depreciation and the improvement in the current account.
Since observed outcomes are always the result of past and present
shocks, the dynamics of adjustment toward the steady state are of particu-
lar interest. The authors analyze two representative cases. In response to a
shock that increases the trade deficit, such as an increase in U.S. economic
activity or an enlarged preference for imports, there is, as explained above,
an initial, unexpected depreciation of the dollar, followed by a gradual
 xii Brookings Papers on Economic Activity, 1:2005

further, anticipated depreciation and an increase in U.S. net debt. How

much of the depreciation is immediate and how much takes place on the
subsequent path of adjustment depend on the response of trade to the depre-
ciation and on the responsiveness of portfolio demands to the anticipated
changes in relative rates of return. The less substitutability between for-
eign and U.S. assets, the smaller will be the initial depreciation, and the
more rapid the subsequent depreciation. However, the eventual deprecia-
tion in the new steady state is the same, and large enough to generate a
sufficient trade surplus to offset the higher interest payments on the larger
debt.
The second case involves a response to a shock that increases the
demand for U.S. assets, such as an increase in demand by foreign govern-
ments. In this case the reduced supply available to private portfolios leads
to an initial dollar appreciation. This enlarges the trade deficit, adding to the
future flow of dollar assets supplied. The subsequent path is one of a grad-
ual, anticipated depreciation and increase in net debt. Despite the initial
favorable portfolio shift, the new steady state requires a weaker dollar,
since, as in the previous example, the trade surplus must be larger to offset
the interest payments on the now-larger debt.
The authors suggest that the U.S. experience of recent years can be
understood as responses to shocks like those just described. In their view a
shift in private portfolio preferences toward U.S. assets led initially to an
appreciation of the dollar. Independently, a shift in the preferences of U.S.
consumers toward foreign goods worsened the trade balance by more than
can be explained by exchange rate and income effects. As described above,
both kinds of shifts predict an eventual sustained dollar depreciation to a
level below that prevailing before the shift. Although the accumulation of
reserves by foreign governments has supported the dollar against some cur-
rencies, the authors argue that the United States has entered the deprecia-
tion phase of the adjustment that their model predicts.
To assess future prospects, and in particular how large an eventual dollar
depreciation should be expected, the authors quantify their model using
estimates of present wealth, assets, portfolio shares, and net debt for the
United States and the rest of the world, together with estimates of model
parameter values based on existing empirical studies and some assumptions
about adjustment speeds and policy preferences. For 2003 these estimates
include the following: U.S. assets of $36.8 trillion, foreign assets of $33.3
trillion, and U.S. net foreign debt of $2.7 trillion; 77 percent of U.S. wealth
William C. Brainard and George L. Perry xiii

invested in U.S. assets, and 71 percent of foreign wealth invested in foreign

assets. In the model these shares imply that a transfer of one dollar of U.S.
wealth to foreigners leads to a decrease of 48 cents in demand for U.S.
assets. The estimated trade elasticities imply that a 1-percentage-point
reduction in the ratio of the trade deficit to GDP requires a depreciation of
15 percent.
Armed with these quantifications of their model, the authors use it to
predict where the U.S. international position is headed. First they calcu-
late the exchange rate adjustment that would be needed to maintain the pre-
sent net debt position as a steady state, under the implicit assumption that
the economy has already adjusted to past shocks, and introducing no impor-
tant asymmetries between foreign and U.S. interest rates or growth rates. In
this case the ratio of the current account deficit to GDP that can be sustained
indefinitely is given by the economy’s growth rate times the ratio of net
debt to GDP. With 3 percent annual growth in U.S. GDP, maintaining a
net debt ratio of about 25 percent requires reducing the current account
deficit from its present 6 percent to 0.75 percent of GDP. With annual inter-
est rates at 4 percent, this requires a depreciation of the dollar of 56 percent.
The authors note some important qualifications to this calculation. To the
extent that the present current account deficit reflects J-curve effects in
response to the dollar’s recent depreciation (in which a depreciation at
first worsens the current account balance before improving it), it over-
states the additional depreciation required. Noting that the current
account continued to worsen for nearly two years after the depreciation
of the mid-1980s began, they estimate that a similar path this time would
mean that only a 34 percent further depreciation is needed. They also
note that if the U.S. net debt ratio were allowed to stabilize at a higher
level than the present, the equilibrium current account deficit could be
larger.
As an alternative way to assess the dollar’s prospects, the authors under-
take dynamic simulations of the response to trade and portfolio shocks in
which the equilibrium debt-to-GDP ratio is endogenous. Simulating per-
manent shocks to the trade deficit, they calculate that a 1-percent-of-GDP
shift away from U.S. goods increases the equilibrium net debt ratio by
17 percentage points and causes the dollar to depreciate by 12.5 percent.
Simulating shifts in asset preferences, they calculate that, in response to a
shift that raises the share of U.S. assets in both U.S. and foreign portfolios
by 5 percentage points, the dollar initially appreciates and then eventually
 xiv Brookings Papers on Economic Activity, 1:2005

reaches an equilibrium depreciation of 15 percent with a 35-percentage-

point increase in the net debt ratio. A striking feature of both simulations
is how long it takes to reach equilibrium. After fifty years the adjustment
is still far from complete, with the dollar still above its pre-shock level after
the shift toward dollar assets, and the depreciation only about two-thirds
complete after the shift in trade away from U.S. goods. Although they ques-
tion the realism of these extraordinary adjustment periods, the authors
believe they do correctly show that the adjustment process can be very
long.
Such gradualism contrasts with the predictions of some observers that
the dollar is likely to fall abruptly in the near future. To evaluate this pos-
sibility, the authors examine under what conditions their model would pre-
dict a faster depreciation than in their baseline simulations. As discussed
above, the anticipated rate of depreciation is faster, the less substitutability
there is between U.S. and foreign assets, with the extreme case of constant
shares providing an upper bound. For this case the authors show that the
anticipated rate of depreciation depends on the change in the ratio of U.S.
net debt to U.S. assets: a faster rise in the debt ratio requires a more rapid
depreciation to maintain portfolio balance. In a situation where the net
debt ratio is rising by 5 percent a year, and with a ratio of gross assets to
GDP of 3—both rough approximations of recent experience—they calcu-
late an anticipated rate of depreciation of 2.7 percent a year. This estimate
is based on anticipated portfolio shares remaining constant. In the model,
however, the rate of depreciation will also be affected by any anticipated
change in the relative demand for U.S. assets—a shock imposed on top of
the constant-shares assumption in the previous calculation. If the demand
for shares of U.S. assets in foreign or domestic portfolios is expected to
decline, the expected depreciation can be much faster. For example, if the
share of U.S. assets demanded in foreign portfolios is expected to decline
by 2 percentage points over the coming year, the expected depreciation
rises to 8.7 percent.
The authors note that there is considerable disagreement about the share
of U.S. assets that foreigners will want to hold in the future. Some
observers argue that foreign central banks will continue their recent policy
of adding to dollar holdings. Others see a latent demand for U.S. assets
by private Chinese investors who are currently restrained by capital con-
trols. Although the authors consider these outcomes possible, they find it
more likely that the relative demand for U.S. assets will decline in the near
William C. Brainard and George L. Perry xv

future, as foreign central banks stop pegging the dollar or diversify their
portfolios away from U.S. assets, or both. The calculations just provided
for a shift in shares are then relevant. The authors also observe that the
longer the peg continues, the larger both the initial and the eventual depre-
ciation will be.
The depreciation of the dollar since its 2002 peak has been very uneven
against different currencies: the dollar has fallen 45 percent against the
euro, 25 percent against the yen, and not at all against the Chinese ren-
minbi. To investigate how future adjustments would impact each of these
important currencies, the authors extend the essentials of their model to
include four regions rather than just two. The analysis focuses on the inter-
relations among the United States, Japan, the euro region, and China, ignor-
ing the rest of the world. The authors assume that half the U.S. current
account deficit is with China and a quarter with each of the others, values
that approximate recent actual deficits. These deficits transfer wealth, and
how that wealth is invested drives exchange rate movements. The model
allows for two special features of the Chinese economy: capital controls
on private financial capital inflows and outflows, and the pegging of the
renminbi to the dollar. Asset preferences in each of the other three regions
are allowed to differ, but all are assumed to have the same marginal
response to changes in expected returns, and interest rates measured in the
domestic currency are assumed to be the same in each. The authors illus-
trate the main forces at work using a simplified version of the model in
which asset demands do not depend on expected returns. For a given U.S.
current account deficit, the more dollar assets China holds, the smaller is the
appreciation of the euro and yen. Surprisingly, if China holds only dollar
assets, a U.S. current account deficit actually causes the dollar to appreciate
against both the euro and the yen, since most of the U.S. deficit is with the
region with extreme dollar preferences. If only Japan accumulates dollars,
both the yen and euro appreciate, with the yen appreciating more. In this
case a transfer of wealth to Japan leaves the real effective exchange rate of
the euro unchanged, as the euro rises against the dollar and falls against
the yen.
The authors also use this framework to analyze the effects of prospective
changes in China’s policies. If China stops pegging but maintains capital
controls, it will have a zero current account surplus, which would require an
appreciation of the renminbi against the dollar. Reserve accumulation
would then cease, and the U.S. current account deficit would have to be
 xvi Brookings Papers on Economic Activity, 1:2005

financed entirely by investors in Japan and Europe. This shift in wealth

accumulation away from the region with extreme dollar preferences would
strengthen the euro and the yen against the dollar. A diversification of
China’s portfolios away from all dollars would have a similar effect. The
same qualitative results are also found in simulations that allow for the
endogenous response of portfolio choices to expected relative returns.
Thus, in the authors’ analysis, China’s pegging to the dollar has limited
the appreciation of the euro and yen against the dollar, in contrast to the
opinion of some commentators that it has increased the pressure on the euro
to appreciate.
The authors briefly address the connections between domestic fiscal and
monetary policy and the U.S. international position. As the U.S. current
account and budget deficits have risen together in the past five years, they
have frequently been paired in discussions of needed policy changes, with
some commentators identifying the latter as the cause of the former and
calling for reduced fiscal deficits as a possible substitute for depreciation.
The authors point out, however, that these are complementary changes
rather than substitutes, with interest rates a key link between the two. With
the dollar depreciating under the pressure of excessive current account
deficits, demand for U.S. output expands, requiring a combination of higher
interest rates and fiscal deficit reduction to maintain domestic balance.
Because higher interest rates would limit the immediate depreciation while
requiring more in the future, smaller budget deficits are the appropriate bal-
ancing change. But, if fiscal policy is tightened without dollar depreciation,
the economy is likely to weaken.
The authors conclude by summarizing the implications of their findings
for understanding the recent past and projecting the future. In their view the
path of the dollar since the late 1990s has been supported by increases in the
demand for U.S. assets, first by private investors for equities and more
recently by central bank demands for bonds. A shift in preferences away
from U.S. goods has also contributed to growing trade deficits in this
period. Imperfect substitution in portfolios helps account for the gradualism
of exchange rate adjustments and for the persistent U.S. current account
deficits that have been observed. The model predicts that a gradual depre-
ciation of the dollar will be the prevailing trend for an extended period.
However, if the expected demand for U.S. assets falls, as it would if cen-
tral bank policies changed, the decline in the dollar would be more abrupt.
Similarly, the gradual depreciation could be interrupted by a temporary
William C. Brainard and George L. Perry xvii

appreciation if investors’ preferences shifted toward dollar assets, although

the resulting larger trade deficits would lead to an even larger depreciation
eventually. For the same reason, a rise in U.S. interest rates would
strengthen the dollar only temporarily and require a larger depreciation in
the longer run. The authors thus reason that a better policy mix would
combine a reduction of budget deficits with a reduction of interest rates to
maintain growth.
Turning to China, the authors argue that eventually the government will
find it difficult to continue to sterilize interventions and will abandon its
dollar peg. But the longer the peg will have supported the dollar, the larger
the eventual dollar depreciation will have to be in order for the United
States to service the larger accumulated foreign debt. The authors also
observe that a large dollar depreciation would not necessarily be a major
problem for the United States. By improving the trade balance, it would
permit a reduction of budget deficits without causing a recession. However,
dollar depreciation might pose a bigger problem for Japan and Europe,
which are already growing slowly and which have limited scope for expan-
sionary stabilization policies.

Some lay commentators have suggested that eliminating the federal

budget deficit would automatically reduce today’s massive deficit in the
U.S. current account. In a 1987 paper, James Tobin identified this as one
of eight “myths” about exchange rates and the current account, because it
ignores the fact that improvements in the current account balance have to
be earned in competition with foreign producers and will, if employment
is to be maintained, require changes in exchange rates and terms of trade. In
the second article in this issue, Maurice Obstfeld and Kenneth Rogoff pur-
sue this theme. They first provide a wide-ranging discussion of recent eco-
nomic developments, concluding that the U.S. current account deficit will
before long have to be substantially reduced, if not eliminated. They then
model the price adjustments that would be required to change import and
export patterns in the United States and abroad so as to eliminate or sub-
stantially reduce the U.S current account deficit without reducing aggregate
economic activity.
Although most analysts recognize that improving the trade balance will
require a real depreciation of the dollar, less attention has been paid to the
likely need for changes in the relative price of traded and nontraded goods
both in the United States and among its trading partners. In earlier work
 xviii Brookings Papers on Economic Activity, 1:2005

Obstfeld and Rogoff have argued that these adjustments are, if anything,
likely to be larger than the changes in the relative prices of domestic and
foreign tradable goods—the terms of trade. It is easy to show why this
might be so. Without changes in production anywhere, eliminating the U.S.
current account deficit, which today stands at roughly 6 percent of GDP,
implies something like a 20 percent reduction in U.S. consumption of
traded goods. Assume for simplicity that the traded goods of different coun-
tries are perfect substitutes, so that exchange rate changes do not change the
relative price of different traded goods, but only the prices of nontraded
goods relative to traded goods within countries. Then, with a unitary elas-
ticity of substitution between traded and nontraded goods and hence con-
stant shares, this 20 percent reduction in consumption of traded goods
requires a fall in the price of nontraded goods relative to traded goods of the
same percentage. In foreign countries, where, under these assumptions,
consumption of traded relative to nontraded goods has to rise, the relative
price of the latter must also rise. If the traded goods of different countries
are not perfect substitutes, the calculations are more complicated, and the
required terms of trade and real exchange rates need to be determined
simultaneously. But the qualitative nature of the needed adjustment is the
same.
To capture the salient features of the current international environment,
the authors develop their model by assuming three world regions, repre-
senting the United States, Asia, and Europe, all linked by trade and by a
matrix of international asset and liability positions. This enables the authors
to model asymmetries in the trading relationships between regions and to
analyze the implications of dividing the improvement in the U.S. trade
account between Europe and Asia in different ways. The model is short run
and static. Each region produces two goods: a nontraded good consumed
only by its residents, and a traded good that is both consumed domestically
and exported. Hence there are a total of six goods in the world economy. The
regions are endowment economies with flexible prices, implicitly assum-
ing factor immobility between sectors and full employment.
The preferences of consumers, and in particular the elasticities of sub-
stitution among the different goods, play the central role in determining
price adjustments associated with changes in the current account. Four
commodities are available to consumers in each region—their own region’s
traded and nontraded goods and the traded goods of the other two regions.
The authors model goods preferences in each region by means of two
William C. Brainard and George L. Perry xix

constant-elasticity-of-substitution (CES) consumption indexes: the first is

an index of overall traded good consumption derived from a bundle of the
three traded goods, and the second aggregates this index with nontraded
good consumption to provide a utility measure for total consumption.
Although the functional form of these CES functions is the same across
regions, the weights on the commodities differ. In particular, the traded
goods index displays home bias: consumers in each region have a relative
preference for the traded good that it produces and exports. Even though the
law of one price (individual traded goods have the same price everywhere)
holds, the price indexes for each region’s bundle of traded goods will dif-
fer across regions, because each depends on the region’s own consump-
tion weighting of individual traded goods. This implies that an increase in
a region’s income and expenditure improves its terms of trade, raising the
price of its exports relative to that of its imports. The United States and
Europe exhibit mirror symmetry in their preferences for each other’s traded
good but place the same weight on the Asian traded good. Asia meanwhile
weights the U.S. and the European traded goods the same, and the model
allows the weight it places on those goods to be changed, reflecting changes
in openness to trade. Whereas the weights on different goods thus differ
across regions, elasticities of substitution among goods are assumed to be
the same for all regions. The authors review a range of empirical studies
to arrive at informed judgments about the size of these elasticities. In their
baseline calculations the elasticity of substitution among tradables is
assumed to be 2, implying that, ceteris paribus, a 10 percent change in the
consumption of, say, an Asian import to the United States would be asso-
ciated with a 5 percent change in its price relative to that of the U.S. traded
good. The elasticity of substitution between nontraded goods and the index
of consumption that aggregates the three traded goods is assumed to be 1,
as in the simple example above.
Given these preferences, the authors can solve for the prices that equate
demand to supply for any global allocation of the six commodities. The
bilateral terms of trade are simply the relative prices of any two regions’
traded goods. Given the assumption of CES utility, the authors compute
exact price indexes for each region’s consumption bundle of traded goods
and for its overall consumption. Ratios of the latter give the corresponding
bilateral real exchange rates.
As noted earlier, even though the law of one price holds, the price
indexes for the bundle of traded goods differ across regions because of
 xx Brookings Papers on Economic Activity, 1:2005

differences in consumption weighting of the three traded goods. The

authors assume that each region’s bundle of traded goods has a 0.25 weight
in total consumption. This means that a change in a region’s bilateral real
exchange rate is 0.25 of the change in the region’s relative price index for
traded goods. That is, changes in the terms of trade, through their differing
effects on regions’ price levels for traded goods, can be traced directly to
real exchange rates. For example, if the price of the U.S. traded good falls
relative to the price of Europe’s traded good—an improvement in Europe’s
bilateral terms of trade—the relative price of the United States’ traded
goods index will also fall. Hence there will be a real depreciation of the dol-
lar relative to the euro.
Most of the burden of reducing the U.S. current account deficit has to
be borne by U.S. consumers reducing their consumption of traded goods,
but part of the adjustment is accomplished through valuation effects. The
United States is a net debtor, with its liabilities predominantly denominated
in dollars and more than half of the foreign assets held by U.S. residents
denominated in foreign currencies. Depreciation of the dollar therefore
actually decreases U.S. net indebtedness. Although this decrease in U.S. net
worth might be expected to affect demand gradually over time, it has an
immediate effect on the current account. Since foreign-denominated U.S.
assets exceed foreign-denominated U.S. liabilities, and interest payments
are largely denominated in the same currency as the underlying asset, the
dollar value of U.S. net interest receipts rises with a depreciation.
To estimate the effect on the terms of trade and real exchange rates of
reducing the U.S. current account deficit by 5 percent of GDP, the authors
have to make assumptions about how the offsetting reduction in current
account surpluses is distributed between Europe and Asia. They consider
three scenarios: a global rebalancing scenario, where the current accounts
of all three regions go to zero; a “Bretton Woods II” scenario, where Asia’s
currencies remain pegged to the dollar (a hypothesis analyzed at length in
the paper by Michael Dooley and Peter Garber in this volume); and a muted
version of the latter, where Asia maintains its current account surplus so
that a reduction in Europe’s surplus just balances the U.S. deficit reduction.
Given the assumed baseline elasticities, the changes in consumption
implied by global rebalancing imply very large real exchange rate changes.
The euro appreciates in real terms by over 28 percent, and the Asian cur-
rencies by over 35 percent. The greater real appreciation for Asia reflects
the fact that, initially, Asia has a much larger surplus than Europe, so that
William C. Brainard and George L. Perry xxi

moving to balance requires a much larger increase in its consumption of

traded goods. Although these may seem like large numbers, they are not
so different from what would be expected if traded goods were perfect
substitutes as in the earlier example. The fact that they are not, and that
there is home bias, does result in a deterioration of the U.S. terms of trade
with both Europe and Asia of about 14 percent; the result is a slightly larger
real depreciation than would otherwise be required.
The authors’ model is developed entirely in real terms, but they are able
to translate the real exchange rate changes into nominal changes by making
assumptions about how domestic price levels change. If, for example, each
central bank targets stability in its region’s overall consumer price index,
then real exchange rate changes are the only source of nominal exchange
rate change. Stabilizing the GDP deflator, which has different weights than
the consumer price index, gives much the same result.
The substantial depreciation of the dollar predicted by the model has a
large effect on the Asian net foreign asset position: because 80 percent of
Asia’s foreign asset holdings, but only 34 percent of its foreign liabilities,
are denominated in dollars, Asia’s net foreign asset position is reduced in
value by 60 percent. Although this is a substantial wealth effect, it produces
only a small decline in the current account, in turn only slightly reducing
the required exchange rate adjustment. Europe is in a more balanced posi-
tion and suffers a much smaller loss of wealth, with a negligible effect on
the required dollar depreciation.
The changes in consumption implied by the other two scenarios require
quite different real exchange rate adjustments. Under the authors’ Bretton
Woods II scenario, Asia raises its surplus in the process of pegging to the
dollar, increasing the adjustment that Europe must make. Specifically, Asia
increases its surplus as a percentage of U.S. traded-goods output from its
current 15 percent to 31 percent, and Europe has to move from its current
5 percent surplus, measured the same way, to a 31 percent deficit. This
adjustment requires an appreciation of the euro by roughly 60 percent
against both the dollar and the Asian currencies. Europe’s terms of trade also
rise dramatically, on the order of 25 percent. The authors conclude that sus-
taining Asia’s peg in the context of a substantial reduction in the U.S. current
account deficit is likely to be politically unacceptable for Europeans. Fur-
thermore, even though Europe’s net foreign asset position is much less sen-
sitive than Asia’s to exchange rate changes, the required appreciation of the
euro would be large enough to result in a significant loss in wealth.
 xxii Brookings Papers on Economic Activity, 1:2005

In the third scenario, Asia allows its currencies to rise against the dol-
lar by roughly 20 percent, just enough to keep its current account sur-
plus constant as the United States moves to current account balance, thus
placing less of a burden on Europe than in the Bretton Woods II sce-
nario. Although the euro still has to appreciate by nearly 45 percent
against the dollar, Europe’s effective exchange rate is affected much less,
by 32 percent rather than 60 percent, because of Europe’s substantial
trade with Asia.
Although the authors believe they have made fairly optimistic assump-
tions about elasticities, which, if anything, understate the required price
adjustments, they also report results under some alternative assumptions.
Raising the elasticity of substitution between nontraded and traded goods
from 1 to 2 reduces the required depreciation of the dollar significantly. In
the global rebalancing scenario the real dollar-euro rate rises by 19.3 per-
cent rather than 28.6 percent, and the Asian currencies appreciate by 22.5
percent rather than 35.2 percent—still quite significant adjustments. Elim-
inating the changes in the terms of trade by assuming that traded goods
are near-perfect substitutes for each other has similar quantitative effects,
revealing that the terms-of-trade effects were responsible for about a third
of the real dollar depreciation in the baseline model.
As previously noted, because the United States’ foreign debts are mostly
denominated in dollars and its foreign asset holdings mostly in foreign
currencies, valuation effects dampen the depreciation of the dollar required
to eliminate its current account deficit. The authors show that this effect is
modest: in their baseline estimates, the depreciation in terms of the U.S.
effective exchange rate is about 13 percent less than it would be in the
absence of valuation effects, implying that improvements in the trade bal-
ance still have to do the heavy lifting. They also show the effect of the
United States losing its historical ability to borrow at a low interest rate: the
effect of putting the United States on a par with other debtors is of roughly
the same magnitude as the valuation effects estimated above.
The authors recognize that their model ignores some effects that might
significantly change their estimates of needed dollar depreciation. In par-
ticular, they note that the realism of two of their key assumptions depends
on the time horizon over which adjustments take place. In the short run,
perhaps one or two years, the assumption of factor immobility appears
reasonable, but the assumption of completely flexible prices seems less so.
For the longer run, price flexibility seems more reasonable, but factor
William C. Brainard and George L. Perry xxiii

immobility is implausible. The authors recognize that factor reallocation

between sectors will dampen the expected real exchange rate adjustments
compared with the adjustments estimated for the core model. Thus, if cur-
rent account adjustments take place slowly and over many years, a smaller
reduction in the real exchange rate will be required to achieve current
account balance. Although they recognize that their model is incomplete
in this and other ways, the authors believe that the framework they have
developed for understanding needed relative price changes will be essential
in any analysis of major adjustments in current accounts.

The determinants of current account balances have been analyzed

much more extensively than those of international capital flows. But the
interaction of the capital and current accounts in determining exchange
rates highlights that understanding capital flows is just as important as
understanding trade flows. For emerging economies the role of such flows
has received considerable attention and has been identified as an impor-
tant factor in currency crises. However, for the U.S. dollar, which has been
the world’s dominant reserve currency for over half a century, there is less
empirical evidence and considerable uncertainty about how its exchange
rate responds to U.S. current account deficits and the accumulation of dol-
lar assets abroad. In the third article of this issue, Michael Dooley and Peter
Garber expand on what they have elsewhere called the “Revived Bretton
Woods” hypothesis, which stresses the willingness of foreign official sec-
tors to accumulate dollar liabilities, and they marshal support for their pre-
diction that U.S. current account deficits need not trigger a major dollar
devaluation or currency crisis in the foreseeable future.
Dooley and Garber first describe the key features of the global econ-
omy that underpin their Revived Bretton Woods view. The first is China’s
ongoing transformation from a centrally planned to a nascent market econ-
omy, which has moved hundreds of millions of underemployed workers
into the global market for labor. China’s continued economic development
depends on employing such workers, and, in the authors’ view, pursuit of
this goal will override the conventional pressures of trade imbalances in
China’s exchange rate strategy. The second is that most successful emerg-
ing economies have been net exporters of capital, contradicting the usual
assumption that successful development involves poor countries borrowing
from rich ones. The authors relate this seeming anomaly to the emerging
economies’ need for international capital by arguing that the export of sav-
 xxiv Brookings Papers on Economic Activity, 1:2005

ing supports two-way trade in financial assets that improves the productiv-
ity of their domestic capital formation. In China’s case access to needed
international capital is currently inhibited by the country’s geopolitical past
and its primitive financial system. The third key feature is that the large and
growing current account deficit of the United States has been funded at low
interest rates by foreign private and official lenders, suggesting that the
large foreign holdings of U.S. assets have not diminished the demand for
further accumulation.
The motivation for emerging economies, and particularly China, in
building foreign reserves is central to the authors’ argument. Export pro-
motion has long been an accepted strategy for a developing economy, and
the value of building reserves became apparent when foreign capital flight
from East Asia and elsewhere led to the crises of the 1990s. However, these
motives by themselves neither explain recent developments nor predict
how far the reserve buildups will go. Export-led growth alone does not
imply the need for a trade surplus and net export of capital, nor does the
precautionary motive require an indefinite buildup of reserves. However,
the authors hypothesize an additional motive for building reserves, which
today applies most clearly to China. Growth requires efficient capital for-
mation, yet the domestic financial system will, for a long time, not be up
to the task of channeling China’s high rate of saving into a high rate of
productive domestic investment. International financial intermediation
can substitute for the inadequate domestic financial system, but potential
foreign investors are put off by political risk. Dooley and Garber argue
that China’s foreign reserves act as collateral that reduces this risk. They
provide an extensive discussion of the role of private collateral arrange-
ments and the uncertainties that inhibit financial investments in their
absence, citing earlier work by Ricardo Caballero and Arvind Krishna-
murthy on the role of international collateral for private financing in devel-
oping economies. The authors observe that the U.S. authorities are legally
empowered to freeze or seize foreign-owned assets under a range of
unusual circumstances, and they identify many occasions when this has
been done. Although the conditions for taking such action are not well
defined or even generally understood, market participants and other gov-
ernments believe that the United States will take similar action in the future
against foreign governments that expropriate private foreign assets. By
holding dollar reserves that are vulnerable to seizure, a country thus pro-
vides effective collateral to potential investors.
William C. Brainard and George L. Perry xxv

Dooley and Garber concede that the Chinese authorities may have stum-
bled on this role of foreign reserves inadvertently. But they argue that,
having done so, the authorities now accept a continuing buildup of foreign
reserves as support for continued growth in gross inward foreign invest-
ment. Hence the authors’ collateral hypothesis provides a connection
between China’s current account surpluses (or net capital flows) and gross
capital flows.
In the present geopolitical climate, the collateral hypothesis would seem
more relevant to China than to the more developed Asian economies, espe-
cially Japan. But Dooley and Garber note that Japan has managed its
exchange rate for many years as a way of dealing with its own employ-
ment problem, and they see some of the other Asian economies as moti-
vated to keep their currencies aligned with China’s. They also see little
pressure from market forces that would cause Japan and the other
economies to abandon their reserve buildups. History provides many exam-
ples of market forces overwhelming official attempts at intervention to sup-
port a weak currency, but the analysis of those cases does not necessarily
apply to interventions to repress a strong currency. Nor have the constraints
that often arise when undervaluation or intervention leads to excessive
monetary expansion and overheating been a problem for China or for other
developing Asian economies, and those constraints are clearly irrelevant for
today’s cyclically depressed Japan.
Having thus explained why, in their view, these buildups of dollar
reserves abroad may continue, the authors turn to historical evidence of past
episodes of buildups and how they have ended. From a sample of 115
developing and industrial countries for the period 1970–2004, they identify
episodes in which a country ran current account surpluses for several con-
secutive years and the government increased its net foreign asset position
by at least 25 percent of the change in national net foreign assets. They
find several regularities in these episodes, one of which is that the typical
episode of reserve buildup has a relatively benign ending. With few excep-
tions, current account surpluses grew during the period of reserve accu-
mulation. When the accumulation stopped, current account surpluses
declined on average by 2 percent of GDP in the first year, suggesting that
the accumulation typically ended as a result of some shock to the previous
situation. On average, a real appreciation occurred in the last three years
of reserve accumulation, which in itself suggests a fundamental disequi-
librium in the exchange rate and the current account. But rather than sub-
 xxvi Brookings Papers on Economic Activity, 1:2005

sequently appreciating, as might be expected from such a disequilibrium,

currencies on average depreciated and economic growth moderated.
From this analysis of the typical experience, the authors turn to a more
detailed look at Japan, China, and Korea, which together accounted for
45 percent of global reserve holdings at the end of 2004. Since 1970 Japan
has had three episodes of extended reserve accumulation: the first starting
in 1986 and lasting three years, the second starting in 1992 and lasting
five years, and the current episode, which started in 1999. Based on the pat-
tern of the first two episodes, the authors project a moderate real apprecia-
tion of the yen and moderate economic growth in Japan in the immediate
future, with reserve accumulation ending when the current account deteri-
orates, at which time the real value of the yen will fall. Korea experienced
net reserve accumulation from 1986 to 1989 and again from 1998 to the
present. The end of the first episode coincided with a decline in the current
account surplus and a slowdown in GDP growth. The authors find nothing
unusual about the present episode, with reserve accumulation roughly
matching the current account surpluses and the won strengthening moder-
ately in real terms.
China’s episode of reserve accumulation is the longest in the entire sam-
ple, extending from 1990 to the present. Small current account surpluses
were roughly matched by reserve accumulation from 1990 to 2001. Since
then, however, China’s experience has been without modern precedent:
the current account surplus has grown rapidly, and private capital inflows
have been roughly as large; hence reserve accumulation has been about
twice as large as the surplus. Because the authorities have been able to
control inflation, the authors see no pressure to end the accumulation. How-
ever, they suggest that the buildup might end if an interruption of direct
investment inflows or liberalization of capital outflows were to lead to a real
depreciation of the renminbi.
From these examinations of past episodes, the authors draw several gen-
eralizations. They find almost no support for the idea that reserve accumu-
lations end with speculative attacks that force the currency to appreciate.
Rather, they typically end when the current account surplus declines sub-
stantially or swings into deficit, and they are followed by a real depreciation
and a modest downturn in the economy. One implication is that episodes
of reserve buildups do not end with capital losses on the government’s
reserves. Nor do they end with recessions generated by a sharp real appre-
ciation. From this evidence the authors judge that there are no constraints or
William C. Brainard and George L. Perry xxvii

obvious risks based on historical experience that would keep today’s cur-
rent account surplus economies from continuing to finance U.S. current
account deficits, as the Revived Bretton Woods hypothesis predicts. Dooley
and Garber expect that China’s present financial repression of capital flows
and distortion of the real exchange rate will end when the country’s indus-
trial sector has grown sufficiently and the domestic financial system has
become capable of efficient intermediation. But they also expect that this
will take a long time.

Many analysts and commentators in the business and financial press

see the U.S. international net debtor position as a major risk on the eco-
nomic horizon, not only for the United States but possibly also for the
global economy. In the fourth article in this volume, Sebastian Edwards
examines the recent history of the U.S. current and capital accounts, mod-
els some likely paths for them in the years ahead, and examines historical
episodes of sustained deficits and growing foreign indebtedness in other
countries for clues to how serious a risk the present U.S. situation entails.
Edwards starts with a brief history of the U.S. international position in
the three decades since exchange rates began floating in the early 1970s.
Focusing on the real trade-weighted exchange rate of the dollar and the
ratio of the U.S. current account balance to GDP, he identifies two extended
episodes of major imbalance. The first began in the early 1980s, when the
current account went deeply into deficit following a sharp real appreciation
of the dollar. This episode resolved when a steep depreciation that began
in 1985 returned the current account briefly to balance by 1991. The second
is the present enlargement of the deficit to record levels, which started
with a period of appreciation of the dollar from the mid-1990s to 2002 and
has continued despite the real depreciation that followed.
Each year’s current account deficit worsens the U.S. net international
investment position (NIIP) by a corresponding amount. But the NIIP, which
is measured in dollars, is also affected by changes in the valuation of assets
held across borders. These valuation effects occur as exchange rate move-
ments change the dollar value of foreign assets held by U.S. nationals. In
the 1980s valuation effects that were predominantly positive partly offset
the adverse effects of large current account deficits on the U.S. NIIP. Nev-
ertheless, by 1986 the United States had become a net debtor, and the mas-
sive deficits of the current episode have increased the net debt position to
about 30 percent of GDP. However, because the returns on U.S. assets held
 xxviii Brookings Papers on Economic Activity, 1:2005

by foreigners have been systematically lower than the returns on foreign

assets held by U.S. nationals, the income component of the current account
has to date remained positive. The entire current account deficit thus con-
sists of an enormous deficit in goods and services trade and a modest deficit
in transfers to foreigners.
Edwards turns next to an analysis of where the U.S. current account, real
exchange rate, and NIIP are likely to go from here. The three are, of course,
interrelated, with the main linkages coming from the exchange rate affect-
ing the trade balance, portfolio investments affecting the demand for dol-
lars, and the balance between portfolio and trade flows affecting the
exchange rate. Edwards captures these interrelationships in a model whose
main features include elements of the models of Blanchard, Giavazzi, and
Sa and of Obstfeld and Rogoff in their papers in this volume. Asset
demands are driven by wealth, with a bias for home assets and exogenously
determined portfolio shares. (The inclusion of demand by foreign central
banks can be treated as a shift in this home bias.) Trade flows are driven
by the real exchange rate, which affects the relative prices of traded and
nontraded goods and services, and by fluctuations and growth in incomes;
the magnitude of these effects is determined by price and income elastici-
ties in the United States and abroad. With this model Edwards is able to
analyze long-run equilibriums, that is, the eventual adjustments to real
exchange rates and current accounts that can be expected in response to
various shocks. With some simplifying assumptions, the sustainable ratio of
the U.S. current account to GDP is proportional to the growth rate of U.S.
nominal GDP, with the proportionality depending on the relative returns
and riskiness of its assets and the degree of integration of capital markets—
factors captured in the portfolio balance parameters.
To go further and characterize the dynamic path of such adjustments to
equilibrium, Edwards includes partial adjustments for asset holdings, which
allow for imperfections in countries’ capital markets, and for the current
account, which allow for consumption smoothing. As an example of the
resulting rich dynamics, he shows that a decline in home bias in the rest of
the world, which would increase the sustainable U.S. current account
deficit, would lead initially to the deficit overshooting its new equilibrium
level.
Edwards applies his model to the current situation by calibrating its para-
meters using values from earlier studies and values for the dynamic adjust-
ment terms that best explain the behavior of the U.S. current account since
William C. Brainard and George L. Perry xxix

1996. He then uses the calibrated model to simulate the effect of shocks to
portfolio choices, focusing on a specification in which the desired propor-
tions of foreign and domestic assets remain fixed after the shock. The sim-
ulations also assume that annual economic growth rates in the United States
and abroad average 3 percent and that the terms of trade do not change. In
his base case, foreigners are assumed to gradually increase the desired
proportion of dollar asset holdings in their portfolios from the present
30 percent to 40 percent in 2010, while U.S. nationals reduce their desired
holdings of U.S. assets from 73 percent to 71 percent over the same period.
These portfolio shifts have the effect of doubling foreigners’ net demand for
U.S. assets to an amount equal to 60 percent of U.S. GDP by the end of
the period. With this increase in demand for U.S. assets, the dollar appre-
ciates in real terms for the first four years and then depreciates rapidly,
eventually approaching a new equilibrium 19 percent below its initial 2005
level. The current account deficit initially continues to grow, peaking at
7.3 percent of GDP after four years. It declines sharply thereafter,
approaching an equilibrium ratio of 3.2 percent of GDP after a few more
years. The reversal of the trade deficit is even sharper and larger because the
growing net debt position raises net income payments to foreigners. The
main qualitative findings from this base case are robust under a range of
alternative assumptions about the model’s parameters.
Edwards simulates alternative assumptions about portfolio choice to test
what difference they make to the outcome. If, after the initial five years, for-
eign investors gradually reduce their holdings of U.S. assets to 50 percent
rather than 60 percent of U.S. GDP, the real depreciation and current
account reversals are steeper and eventually greater. After three years the
depreciation is 24 percent and the current account deficit has shrunk by
5.3 percent of GDP. Both changes continue for two more years, overshoot-
ing their eventual equilibrium values: an exchange rate about 23 percent
below, and a current account deficit about 3.5 percent of GDP smaller than,
2004 values. Although the size of these changes is within U.S. historical
experience, once the changes get under way, their abruptness, which comes
from dollar accumulations abroad reaching an assumed limit, could be
destabilizing. Edwards notes that different parameters for the adjustment
process could produce less abrupt changes, but he regards the qualitative
characteristics of the simulations as representative of the model’s dynam-
ics. And he notes that all the reported simulations assume foreign demand
for U.S. assets far exceeding today’s 30 percent of GDP.
 xxx Brookings Papers on Economic Activity, 1:2005

Large, abrupt swings in current account balances have often been

accompanied by disruptions to employment and growth in the affected
economies. Edwards looks to international experience with such reversals
to see whether it offers any insight into what is in store if the United States
undergoes the kind of changes predicted by his simulations. He defines two
types of current account reversal, one in which the current account deficit
declines by at least 6 percent of GDP within a three-year period, and one
in which it declines by at least 4 percent in a single year. For the period
1971-2001 he finds that the first type of reversal occurs in 9.2 percent of
all country-years, and the second type in 11.8 percent of all country-years.
He reports a number of other interesting findings, including a close associ-
ation of reversals and currency crises, a particular exchange rate pattern that
typifies reversals, and a correlation of reversals with economic growth.
However, the great majority of the reversals he finds are for small or less
developed countries. The corresponding incidences of the two types of cur-
rent account reversal for industrial countries are only 2.7 percent and 2.0
percent, and most of those reversals occurred in small countries. Among the
larger industrial countries, only Italy (in 1975) and Canada (in 1982) expe-
rienced reversals in this thirty-year period. Thus, although Edwards’s rich
analysis of historical reversals illuminates the structural and economic con-
ditions—and the problems—typically associated with them, the relevance
to the current U.S. situation is unclear.
Edwards believes nonetheless that it is very likely that the United States
will undergo a major adjustment in the not-too-distant future, which will
modify the present global imbalances between the U.S. and rest-of-world
current accounts. He identifies three main unresolved issues that will shape
how that adjustment unfolds. One is how central banks conduct their
reserves policy in a global economy with mostly flexible exchange rates.
He notes that, in contrast to the argument made by Dooley and Garber in
this volume, many observers believe that foreign central banks that have
been accumulating dollar reserves will reduce their demand for dollar
assets in the future, unleashing an abrupt collapse in the value of the dol-
lar. Another issue is how world interest rates, which influence global invest-
ment, will be affected by a major adjustment to the U.S. current account.
And the third is how private sector saving and government budget balances
evolve, and whether, in tandem with interest rate adjustments, they will suc-
ceed in maintaining global economic growth as the correction of today’s
current account imbalances works itself out.
William C. Brainard and George L. Perry xxxi

Most informed observers forecast a substantial decline in the growth of

the labor force in future decades. The baby-boom generation is approach-
ing retirement, no new explosion of fertility is in sight, and the growth in
female labor force participation that began in the 1960s is seen as largely
complete. Some forecasters also predict a slowing of productivity growth
in the longer run, although this is more controversial: the Social Security
Administration’s 2005 trustees’ report projects that hours worked will grow
by only 0.3 percent a year from 2015 to 2045, a slowdown of 1.2 percent-
age points from the average for this measure from 1958 to 2004. The trustees
also predict that long-run productivity growth will moderate from its pace of
the last fifteen years and that together these two factors will lead to a slow-
down of GDP growth of between 1.6 and 2.2 percentage points a year. Along
with increases in longevity, these projections are the major reason that the
Social Security system in its present form will be unable to maintain cur-
rent benefits into the indefinite future. Given this outlook, if equity invest-
ments earn as high a return as they have over the postwar period on average,
then investing a portion of the Social Security trust fund in equities, or cre-
ating private accounts to allow individual workers to do so, seems an attractive
and almost costless way to improve the system’s prospects. However, in
the fifth article in this issue, Dean Baker, Bradford DeLong, and Paul Krug-
man question this reasoning, arguing that rates of return on equities are
unlikely to match their historical levels if the pessimistic projections of labor
force and productivity growth in the trustees’ report are correct.
Standard economic growth models provide a relevant framework for
analyzing the long-run effects of labor force and productivity growth on
national income growth and the rate of return to capital. The authors begin
by reviewing the predictions of the mainstay of growth analysis, the Solow
model. In steady state the growth rate of national income is the sum of the
growth rates of hours worked and labor-augmenting technical progress, and
the capital stock grows at the same rate as income. At any income growth
rate the net saving rate determines the capital-to-output ratio, which in
turn determines wage rates and the rate of return to capital. It is easy to
show that a change in the rate of growth of labor input or labor productiv-
ity results in a proportional change in the rate of return to capital, with the
proportionality being the ratio of capital’s share of income (which is
assumed constant) to the gross saving rate.
The Solow model takes the saving rate as given, so that the output-to-
capital ratio and the rate of return to capital fall with a reduction in growth.
 xxxii Brookings Papers on Economic Activity, 1:2005

But a sufficient decline in the saving rate will keep the return to capital con-
stant. This leads the authors to analyze two canonical models that address
the effect of changing demographics on saving and hence on the rate of
return. The first is the Ramsey model, a highly stylized model that assumes
that the representative household lives forever, maximizing utility over a
consumption path into the indefinite future. Population growth in this
model is captured by assuming that the size of the representative household
grows over time. Household saving decisions maximize the welfare of this
dynastic household, given the projected growth in household size. Usually
it is assumed that the household’s utility in a given period is simply the sum
of the utilities of the members present in that period, and that the household
decisionmaker, contemplating the future, gives the same weight to the util-
ity of the new members as to his or her own. (The authors call this “per-
fect familial altruism.”) With these assumptions and the assumption that
utility is proportional to the logarithm of consumption (log utility), the
steady-state rate of return rises one for one with labor productivity growth,
as does the growth rate of consumption per worker. However, the rate of
return is unaffected by population growth. The reason for this can most eas-
ily be seen by abstracting from productivity growth, so that consumption
per capita is constant through time. Without population growth, the infi-
nitely lived individual will want to accumulate capital to the point where
the rate of return equals the rate of time preference. With population
growth, the same condition will hold. The fact that a forgone unit of con-
sumption by each household member today has to be divided among 1 + n
members tomorrow is just balanced by the fact that there will be 1 + n
fully weighted members tomorrow. So, as with a single individual, the
rate of return will be driven to the rate of time preference.
Why are these results different from those of the Solow model? In both
models, in the steady state, each new member of the labor force has to be
equipped with capital. In the Solow model the saving rate is constant, so
that the capital-to-labor ratio is higher when there are fewer workers to
equip with capital. In the Ramsey model the saving rate falls to keep the
capital-to-labor ratio and the rate of return to capital unchanged. Baker,
DeLong, and Krugman find the assumption of perfect familial altruism in
the Ramsey model implausible, particularly when many of the future mem-
bers of society are expected to be immigrants unrelated to today’s members.
They show that if there is less than perfect altruism, so that current gener-
ations give less weight to future generations than to themselves, then, when
William C. Brainard and George L. Perry xxxiii

population growth slows, saving does not fall enough to maintain the pre-
vious rate of return on capital.
There are several reasons why the Ramsey model is ill suited for ana-
lyzing the effects of demographic change on saving and the rate of return.
There is no meaningful way to analyze saving for retirement in a model
where individuals live and work forever. Nor can the model analyze the
effects on saving of changes in birth and death rates, the age of retirement,
or uncertainty about the length of life. However, the second canonical
model the authors examine, the Diamond overlapping-generations model,
can readily incorporate such features. In this model, versions of which have
been used by other authors to estimate empirically the effects of demo-
graphic change in the United States and elsewhere, individuals are assumed
to go through a life cycle of earning, saving, and consuming. Each agent
lives two periods, working and saving when young and consuming the
returns on capital acquired through that saving when old. Generations all
have the same preferences but differ in consumption opportunities as pro-
ductivity grows over time. Individuals are assumed to maximize the pre-
sent value of the utility of consumption over their two-period lifetimes,
using log utility and with no bequest motive. Output is given by a Cobb-
Douglas function combining the labor input of the young and the capital
owned by the old.
Even though it abstracts from some realistic features of the typical life
cycle, the Diamond model shows the fundamental differences that arise
from assuming finite rather than infinite horizons. First, the rate of return
bears no necessary relationship to the pure rate of time preference. House-
holds optimally allocate their income over two periods, and so determine
individual household saving, but aggregate saving depends crucially on
the demographic structure.
With log utility, the fraction of income saved is independent of the rate
of return—income and substitution effects just balance—making the analy-
sis quite simple. In the absence of population or productivity growth, aggre-
gate (net) saving would be zero: in every period the dissaving of retirees
just balances the saving of the equally numerous young. With population
growth there are more savers relative to dissavers, so that the capital stock
grows along with the labor force in steady state. However, since the sav-
ing of one generation of workers is used by the more numerous next gen-
eration, the capital-to-labor ratio is lower and the rate of return higher than
with a constant population. If saving per worker were fixed, the capital-to-
 xxxiv Brookings Papers on Economic Activity, 1:2005

labor ratio would fall at a rate proportionate to the growth in the number
of workers. The resulting increase in the steady-state real rate of return
would have no effect on saving with log utility, but the lower wage would
reduce the saving of workers, resulting in an even greater increase in the
rate of return than would otherwise be the case. And, if the utility function
were less elastic than under log utility, as most analysts believe, the increase
in the return to capital would in turn reduce saving, raising the rate of return
even further. Productivity growth—an increase in the labor equivalence of
a worker—divides one generation’s saving among a greater number of
equivalent workers and has the same effect as labor force growth on the rate
of return and wages.
The authors conclude that there are good reasons to believe that the rate
of return on capital will fall if population growth and productivity growth
slow. Since capital is the underlying asset generating returns for the owners
of firms, it is hard to construct a scenario where a permanent decline in the
rate of return on capital does not imply a similar decline in equity returns.
The authors illustrate this by examining the implications of the standard
Gordon equation for equity prices. The rate of return on an equity claim is
its current yield plus capital gains. In the absence of news that affects a
company’s prospects, the price of its stock grows with its dividends. The
Gordon equation simply shows that the price of a stock is equal to its cur-
rent dividend divided by the expected rate of return minus the growth rate
of dividends. Applying this equation to projections for the aggregate econ-
omy, the authors calculate that, if real GDP grows at the 1.5 percent annual
rate consistent with the 2005 trustees’ report, and assuming a constant cap-
ital share, real earnings on capital will likewise grow at 1.5 percent a year,
as will dividends in the absence of changes in firm’s debt-equity ratio or the
dividend payout rate. With this growth rate and the current dividend yield,
the Gordon equation implies an expected annual real rate of return on
stocks of 4.4 percent, considerably less than the 6.5 percent annual real
return averaged over the past half-century. Thus the authors conclude that
this measure of market expectations is consistent with the fall in the rate
of return on capital that they infer from their analysis of growth models.
How might future stock market returns be higher than this calculation
suggests? The authors discuss several possibilities but in the end are skep-
tical of their importance. The capital in the growth models corresponds to
all productive assets in the economy, including, for example, those of unin-
corporated enterprises. Thus the rate of return to capital in traded firms
William C. Brainard and George L. Perry xxxv

could be higher than the return to capital as a whole in the economy. But the
authors see no reason to expect that this is so. A firm’s earnings and divi-
dends typically go through a life cycle, so that the growth in dividends for
the economy as a whole, reflecting the emergence of new firms, may dif-
fer from that of existing firms. But the authors suggest that this difference
may mean lower rather than higher dividend growth for a broad stock
index. Nor do the authors see much room to increase dividend payout rates.
With a decline in the rate of return to capital, increased payouts to stock-
holders have to come either at the expense of bondholders or from a reduc-
tion in retained earnings. Although reducing leverage could temporarily
raise the fraction of earnings paid to stockholders, such reductions could
not continue indefinitely.
As the authors show using the Solow model, a reduction in the saving
rate could maintain a higher rate of return to capital by avoiding the
increase in the capital-to-labor ratio. In that case, although growth in output
would still fall as a result of the fall in growth rates of population and pro-
ductivity, with growth in earnings and dividends following suit, the divi-
dend yield would be higher, with less retained earnings and household
saving required to grow capital at the lower rate. Because there would be
less capital along the economy’s growth path, the rate of return would be
maintained despite the lower growth rate.
The most interesting possibility for maintaining a higher rate of return
is a shift in the distribution of world investment away from the United
States to regions where the labor force is growing faster and potential
returns are higher. The authors reason that, if American companies were to
increase their investment abroad, the growth of earnings of companies in
the index could exceed the rate of growth of the domestic economy. How-
ever, they calculate that achieving the historical 6.5 percent return by this
approach would require that companies increase their foreign investment
by historically unprecedented proportions, unless that investment substi-
tuted for U.S. domestic investment.
The authors also point out that, unless the U.S. trade balance changes,
any such increase in U.S. firms’ investment abroad will have to be balanced
by increased capital inflows of the same magnitude, reducing returns on
domestic capital. Hence, if there is no change in U.S. saving, there will be
no net effect on the growth of the domestic capital stock, and thus no effect
on the rate of return in the United States. They do not address the possibil-
ity that any improvement in the trade balance, coming perhaps from dollar
 xxxvi Brookings Papers on Economic Activity, 1:2005

depreciation, would reduce domestic investment for a given level of saving.

Restoring balance in U.S. trade would reduce net capital inflows, domes-
tic investment, and growth in the domestic capital stock by the same
amount. Since the current account deficit today is near 6 percent of GDP, a
major fraction of the capital deepening associated with slowed population
and productivity growth could be avoided in this way.
Historically, equities have paid a significantly higher return than bonds,
resulting in a large (and, many argue, excessive) risk premium on equities.
If this equity premium can be counted on to persist, it would seem to pro-
vide a good reason for private investors, or the Social Security trust fund, to
invest more heavily in equities and less heavily in bonds. But the authors
observe that the reason for the high premium remains a puzzle, a fact that
argues for caution in adopting any strategy to capitalize on the premium in
the future. Insofar as the premium reflects a failure of markets to efficiently
allocate risk among individuals, it could make sense for the government,
the agent with the greatest ability to manage systematic risk, to take a direct
position in equities. But if, as some believe, the growing sophistication of
markets is already in the process of eliminating the equity premium, the
gains from switching to equities from bonds will disappear. In that case any
attempt to exploit the premium would fail in the long run.
The authors acknowledge that much uncertainty remains about what
the future holds for economic growth. But, they argue, the main inference
of their analysis is that it is precisely in those cases when growth slows
that returns to equities are likely to be lower than historical experience.
Thus, if slower growth does contribute to the Social Security problem,
investment in equities is likely to disappoint as a solution.
OLIVIER BLANCHARD
Massachusetts Institute of Technology
FRANCESCO GIAVAZZI
Universitá Commerciale Luigi Bocconi
FILIPA SA
Massachusetts Institute of Technology

International Investors, the

U.S. Current Account, and the Dollar

TWO MAIN FORCES underlie the large U.S. current account deficits of the
past decade. The first is an increase in U.S. demand for foreign goods,
partly due to relatively faster U.S. growth and partly to shifts in demand
away from U.S. goods toward foreign goods. The second is an increase in
foreign demand for U.S. assets, starting with high foreign private demand
for U.S. equities in the second half of the 1990s, and later shifting to
foreign private and then central bank demand for U.S. bonds in the
2000s. Both forces have contributed to steadily increasing current account
deficits since the mid-1990s, accompanied by a real dollar appreciation
until late 2001 and a real depreciation since. The depreciation acceler-
ated in late 2004, raising the issues of whether and how much more is to
come and, if so, against which currencies: the euro, the yen, or the Chinese
renminbi.
We address these issues by developing a simple model of exchange
rate and current account determination, which we then use to interpret
the recent behavior of the U.S. current account and the dollar and explore
what might happen in alternative future scenarios. The model’s central
assumption is that there is imperfect substitutability not only between

An earlier version of this paper was circulated as MIT working paper WP 05-02, January
2005. We thank Ben Bernanke, Ricardo Caballero, Menzie Chinn, William Cline, Guy
Debelle, Kenneth Froot, Pierre-Olivier Gourinchas, Søren Harck, Maurice Obstfeld, Hélène
Rey, Roberto Rigobon, Kenneth Rogoff, Nouriel Roubini, and the participants at the Brook-
ings Panel conference for comments. We also thank Suman Basu, Nigel Gault, Brian Sack,
Catherine Mann, Kenneth Matheny, Gian Maria Milesi-Ferretti, and Philip Lane for help with
data.

1
 2 Brookings Papers on Economic Activity, 1:2005

U.S. and foreign goods, but also between U.S. and foreign assets. This
allows us to discuss the effects not only of shifts in the relative demand
for goods, but also of shifts in the relative demand for assets. We show
that increases in U.S. demand for foreign goods lead to an initial real dol-
lar depreciation, followed by further, more gradual depreciation over time.
Increases in foreign demand for U.S. assets lead instead to an initial ap-
preciation, followed by depreciation over time, to a level lower than be-
fore the shift.
The model provides a natural interpretation of the recent behavior of
the U.S. current account and the dollar exchange rate. The initial net effect
of the shifts in U.S. demand for foreign goods and in foreign demand for
U.S. assets was a dollar appreciation. Both shifts, however, imply an even-
tual depreciation. The United States appears to have entered this depreci-
ation phase.
How much depreciation is to come, and at what rate, depends on how
far the process has come and on future shifts in the demand for goods and
the demand for assets. This raises two main issues. First, can one expect
the deficit to largely reverse itself without changes in the exchange rate?
If it does, the needed depreciation will obviously be smaller. Second, can
one expect foreign demand for U.S. assets to continue to increase? If it
does, the depreciation will be delayed, although it will still have to come
eventually. Although there is substantial uncertainty about the answers,
we conclude that neither scenario is likely. This leads us to anticipate, in
the absence of surprises, more dollar depreciation to come at a slow but
steady rate.
Surprises will, however, take place; only their sign is unknown. We again
use the model as a guide to discuss a number of alternative scenarios, from
the abandonment of the renminbi’s peg against the dollar, to changes in the
composition of reserves held by Asian central banks, to changes in U.S.
interest rates.
This leads us to the last part of the paper, where we ask how much of
the dollar’s future depreciation is likely to take place against the euro, and
how much against Asian currencies. We extend our model to allow for
four “countries”: the United States, the euro area, Japan, and China. We
conclude that, again absent surprises, the path of adjustment is likely to be
associated primarily with an appreciation of the Asian currencies, but also
with a further appreciation of the euro against the dollar.
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 3

A Model of the Exchange Rate and the Current Account

Much of economists’ intuition about joint movements in the exchange rate

and the current account is based on the assumption of perfect substitutability
between domestic and foreign assets. As we shall show, introducing im-
perfect substitutability changes the picture substantially. Obviously, it allows
one to think about the dynamic effects of shifts in asset preferences. But
it also modifies the dynamic effects of shifts in preferences with respect
to goods.
We are not the first to insist on the potential importance of imperfect
substitutability. Indeed, the model we present builds on an older (largely
and unjustly forgotten) set of papers by Paul Masson, Dale Henderson and
Kenneth Rogoff, and, especially, Pentti Kouri.1 These papers relax the
interest parity condition and instead assume imperfect substitutability of
domestic and foreign assets. Masson and Henderson and Rogoff focus mainly
on issues of stability; Kouri focuses on the effects of changes in portfolio
preferences and the implications of imperfect substitutability between assets
for shocks to the current account.
The value added of this paper is in allowing for a richer description of
gross asset positions. By doing this, we are able to incorporate into the analy-
sis the “valuation effects” that have been at the center of recent empirical re-
search on gross financial flows,2 and that play an important role in the context
of U.S. current account deficits. Many of the themes we develop, including
the roles of imperfect substitutability and valuation effects, have also been
recently emphasized by Maurice Obstfeld.3

1. Masson (1981); Henderson and Rogoff (1982); Kouri (1983). The working paper
version of the paper by Kouri dates from 1976. One could argue that there were two funda-
mental papers written that year, the first by Dornbusch (1976), who explored the implica-
tions of perfect substitutability, and the other by Kouri, who explored the implications of
imperfect substitutability. The Dornbusch approach, with its powerful implications, has
dominated research since then. But imperfect substitutability seems central to the issues we
face today. Branson and Henderson (1985) provide a survey of this early literature.
2. See, in particular, Gourinchas and Rey (2005) and Lane and Milesi-Ferretti (2002,
2004).
3. Obstfeld (2004). We limit our analysis of valuation effects to those originating from
exchange rate movements. Valuation effects can and do also arise from changes in asset
prices, particularly stock prices. The empirical analysis of a much richer menu of possible
valuation effects has recently become possible, thanks to the data on gross financial flows
and gross asset positions assembled by Lane and Milesi-Ferretti.
 4 Brookings Papers on Economic Activity, 1:2005

The Case of Perfect Substitutability

To see how imperfect substitutability of assets matters, it is best to start
from the well-understood case of perfect substitutability. Consider a world
with two “countries”: the United States and a single foreign country com-
prising the rest of the world. We can think of the U.S. current account and
exchange rate as being determined by two relations. The first is the un-
covered interest parity condition:

(1 + r ) = (1 + r*) EE e
,
+1

where r and r* are U.S. and foreign real interest rates, respectively (asterisks
denote foreign variables), E is the real exchange rate defined as the price
of U.S. goods in terms of foreign goods (so that an increase in the exchange
e
rate denotes an appreciation of the dollar), and E+1 is the expected real ex-
change rate in the next period. The condition states that expected returns
on U.S. and foreign assets must be equal.
The second relation is the equation giving net debt accumulation:
F+1 = (1 + r ) F + D ( E+1 , z+1 ) ,

where D(E, z) is the trade deficit. The trade deficit is an increasing func-
tion of the real exchange rate (so that DE > 0). All other factors—changes
in total U.S. or foreign spending, as well as changes in the composition
of U.S. or foreign spending between foreign and domestic goods at a
given exchange rate—are captured by the shift variable z. We define
z such that an increase worsens the trade balance (DZ > 0). F is the net
debt of the United States, denominated in terms of U.S. goods. The con-
dition states that net debt in the next period is equal to net debt in the
current period times 1 plus the interest rate, plus the trade deficit in the
next period.
Assume that the trade deficit is linear in E and z, so that D(E, z) =
θE + z. Assume also, for convenience, that U.S. and foreign interest
rates are equal (r* = r) and constant. From the interest parity condition, it
follows that the expected exchange rate is constant and equal to the cur-
rent exchange rate. The value of the exchange rate is obtained in turn by
solving out the net debt accumulation forward and imposing the condi-
tion that net debt does not grow at a rate above the interest rate. Doing
this gives
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 5

r 1 ∞ ( 
1 + r ) z+e i  .
−i
E = −  F−1 +
θ
∑
1+ r 0 

That is, the exchange rate depends negatively on the initial net debt position
and on the sequence of current and expected shifts in the trade balance.
Replacing the exchange rate in the net debt accumulation equation in
turn gives

 r ∞ ( 
1 + r ) z+e i  .
−i
F − F−1 =  z −

∑
1+ r 0 

That is, the change in the net debt position depends on the difference
between the current shift and the present value of future shifts in the trade
balance.
For our purposes these two equations have one main implication. Con-
sider an unexpected, permanent increase in z at time t—say, an increase in
the U.S. demand for Chinese goods (at a given exchange rate)—by ∆z.
Then, from the two equations above,
∆z
E − E−1 = − ; F − F−1 = 0.
θ
In words: permanent shifts lead to a depreciation large enough to maintain
current account balance. By a similar argument, shifts that are expected
to be long lasting lead to a large depreciation and only a small current
account deficit. As we argue later, this is not what has happened in the
United States over the last ten years. The shift in z appears to be, if not per-
manent, at least long lasting. Yet it has not been offset by a large depreci-
ation but has been reflected instead in a large current account deficit. This,
we shall argue, is the result of two factors, both closely linked to imper-
fect substitutability. The first is that, under imperfect substitutability, the
initial depreciation in response to an increase in z is more limited, and, by
implication, the current account deficit is larger and longer lasting. The
second is that, under imperfect substitutability, asset preferences matter.
An increase in foreign demand for U.S. assets, for example—an event that
obviously cannot be analyzed in the model with perfect substitutability
we have just presented—leads to an initial appreciation and a current
account deficit. And such a shift has indeed played an important role since
the mid-1990s.
 6 Brookings Papers on Economic Activity, 1:2005

Imperfect Substitutability and Portfolio Balance

We now introduce imperfect substitutability between assets. Let W denote
the wealth of U.S. investors, measured in units of U.S. goods. W is equal to
the stock of U.S. assets, X, minus the net debt position of the United States, F:
W = X − F.
Similarly, let W* denote foreign wealth and X* denote foreign assets, both
in terms of foreign goods. Then the wealth of foreign investors, expressed
in terms of U.S. goods, is given by
W* X*
= + F.
E E
Let Re be the relative expected gross real rate of return on holding U.S.
assets versus foreign assets:
1 + r E+e1
(1) Re ≡ .
1 + r* E
Under perfect substitutability, the case studied above, Re was always equal
to 1; this need not be the case under imperfect substitutability.4
U.S. investors allocate their wealth W between U.S. and foreign assets.
They allocate a share α to U.S. assets and, by implication, a share (1 − α) to
foreign assets. Symmetrically, foreign investors invest a share α* of their
wealth W* in foreign assets and a share (1 − α*) in U.S. assets. Assume
that these shares are functions of the relative rate of return, so that
α = α ( R e, s ) , α Re > 0, α s > 0 α* = α* ( R e, s ) , α*Re < 0, α*s < 0.
A higher relative rate of return on U.S. assets leads U.S. investors to in-
crease the share they invest in U.S. assets, and foreign investors to decrease
the share they invest in foreign assets. The variable s is a shift factor, stand-
ing for all the factors that shift portfolio shares for a given relative return. By
convention, an increase in s leads both U.S. and foreign investors to increase
the share of their portfolio in U.S. assets for a given relative rate of return.

4. One may wonder whether, even if many investors have strong asset preferences, the
effects of these preferences on expected returns are not driven away by arbitrageurs, so that
expected returns are equalized. The empirical work of Gourinchas and Rey (2005), which
we discuss later, strongly suggests that this does not happen, and that financial assets
denominated in different currencies are indeed imperfect substitutes.
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 7

An important parameter in the model is the degree of home bias in U.S.

and foreign portfolios. We assume that there is indeed home bias, and we
capture it by assuming that the sum of portfolio shares falling on own-
country assets exceeds 1:
α ( R e, s ) + α* ( R e, s ) > 1.
Equilibrium in the market for U.S. assets (and, by implication, in the
market for foreign assets) implies
X = α ( R e, s )W + [1 − α* ( R e, s )](W */ E ).
The supply of U.S. assets must be equal to U.S. demand plus foreign de-
mand for those assets. Given the definition of F introduced earlier, this
condition can be rewritten as
(2) X = α ( R e, s ) ( X − F ) + (1 − α* ( R e, s ))[( X */ E ) + F ] ,

where Re is given in turn by equation 1 and depends in particular on E and

e
E+1 . This gives us the first relation, which we refer to as the portfolio bal-
ance relation, between net debt, F, and the exchange rate, E.
To see its implications most clearly, consider the limiting case where
the degree of substitutability is zero, so that the shares α and α* do not
depend on the relative rate of return. In this case
—The portfolio balance condition fully determines the exchange rate
as a function of the world distribution of wealth, (X − F) and [(X*/E )
+F)]. In sharp contrast to the case of perfect substitutability, news about
current or future current account balances, such as a permanent shift in z,
has no effect on the current exchange rate.
—Over time, current account deficits lead to changes in F, and thus to
changes in the exchange rate. The slope of the relation between the ex-
change rate and net debt is given by
dE E α + α* − 1
=− < 0.
dF (1 − α*) X */ E

So, in the presence of home bias, an increase in net debt is associated with
a lower exchange rate. The reason is that, as wealth is transferred from the
United States to the rest of the world, home bias leads to a decrease in the
demand for U.S. assets, which in turn requires a decrease in the exchange
rate.
 8 Brookings Papers on Economic Activity, 1:2005

Outside this limiting case, the portfolio balance determines a relation

between net debt and the exchange rate for a given expected rate of depre-
ciation. The exchange rate is no longer determined myopically. But the
two insights from the limiting case remain: On the one hand, the exchange
rate will respond less to news about the current account than it does under
perfect substitutability. On the other, it will respond to changes either in
the world distribution of wealth or in portfolio preferences.

Imperfect Substitutability and Current Account Balance

Assume, as before, that U.S. and foreign goods are imperfect substi-
tutes and that the U.S. trade deficit, in terms of U.S. goods, is given by
D = D ( E , z ) , DE > 0, Dz > 0.
Turn now to the equation expressing the dynamics of the U.S. net debt
position. Given our assumptions, U.S. net debt is given by
W* ( E
F+1 = (1 − α*( R e, s )) 1 + r ) − (1 − α ( R e, s )) W (1 + r *) + D ( E+1 , z+1 ) .
E E+1

In words, net debt in the next period is equal to the value of U.S. assets
held by foreign investors next period, minus the value of foreign assets
held by U.S. investors next period, plus the trade deficit next period:
—The value of U.S. assets held by foreign investors next period is
equal to their wealth in terms of U.S. goods this period times the share
they invest in U.S. assets this period times the gross rate of return on U.S.
assets in terms of U.S. goods.
—The value of foreign assets held by U.S. investors next period is
equal to U.S. wealth this period times the share they invest in foreign
assets this period times the realized gross rate of return on foreign assets
in terms of U.S. goods.
The previous equation can be rewritten as
 1+ r * E 
(3) F+1 = (1+ r ) F + (1− α ( R e, s )) (1+ r ) 1− ( X − F ) + D ( E+1 , z+1 ).
 1+ r E+1 

We shall call this the current account balance relation.5

5. This appears to give a special role to α rather than α*, but in fact this is not the case.
A symmetrical expression can be derived with α* appearing instead of α. Put another way,
F, α*, and α are not independent. F+1 can be expressed in terms of any two of the three.
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 9

The first and last terms on the right-hand side of equation 3 are standard:
next-period net debt is equal to this-period net debt times the gross rate of
return, plus the trade deficit next period. The term in the middle reflects valu-
ation effects, recently stressed by Pierre-Olivier Gourinchas and Hélène Rey
and by Philip Lane and Gian Maria Milesi-Ferretti.6 Consider, for example,
an unexpected decrease in the price of U.S. goods—that is, an unexpected
decrease in E+1 relative to E. This dollar depreciation increases the dollar
value of U.S. holdings of foreign assets, decreasing the U.S. net debt position.
Putting things together, a depreciation improves the U.S. net debt posi-
tion in two ways: the first, conventional way through the improvement in
the trade balance, and a second way through asset revaluation. Note that
—The strength of the valuation effects depends on gross rather than net
positions and so on the share of the U.S. portfolio in foreign assets (1 − α)
and on U.S. wealth (X − F ). It is present even if F = 0.
—The strength of the valuation effects depends on our assumption
that U.S. gross liabilities are denominated in dollars, so that their value in
dollars is unaffected by a dollar depreciation. Valuation effects would
obviously be very different when, as is typically the case for emerging
market economies, gross positions are smaller and liabilities are denomi-
nated in foreign currency.

Steady State and Dynamics

Assume the stocks of assets X and X* and the shift variables z and s to
be constant. Assume also r and r* to be constant and equal to each other.
In this case the steady-state values of net debt F and E are characterized
by two relations.
The first is the portfolio balance relation (equation 2). Given the equal-
ity of interest rates and the constant exchange rate, Re = 1, the relation
takes the form
X = α (1, s ) ( X − F ) + (1 − α* (1, s ))[( X */ E ) + F ].

6. Gourinchas and Rey (2005); Lane and Milesi-Ferretti (2004). As a matter of logic,
one can have both perfect substitutability and valuation effects. (Following standard prac-
tice, we ignored valuation effects in the perfect substitutability model presented earlier by
implicitly assuming that, if net debt was positive, U.S. investors did not hold foreign assets
and net debt was therefore equal to the foreign holdings of dollar assets.) Under perfect
substitutability, however, there is no guide as to what determines the shares, and therefore
what determines the gross positions of U.S. and foreign investors.
 10 Brookings Papers on Economic Activity, 1:2005

This first steady-state condition implies a negative relation between net

debt and the exchange rate. As we showed earlier, in the presence of home
bias, a larger U.S. net debt, which transfers wealth to foreign investors,
shifts demand away from U.S. assets and thus lowers the exchange rate.
The second relation is the current account balance relation (equation 3).
Given the equality of interest rates, and given the constant exchange rate
and net debt, the relation takes the form
0 = rF + D ( E , z ) .
This second relation also implies a negative relation between net debt and
the exchange rate. The larger the net debt, the larger the trade surplus re-
quired in steady state to finance interest payments on the debt, and thus
the lower the exchange rate.7 This raises the question of the stability of the
system. The system is (locally saddle point) stable if, as drawn in figure 1,
the portfolio balance locus is steeper than the current account balance
locus. (Appendix A characterizes the dynamics.) To understand this con-
dition, consider an increase in U.S. net debt. This increase has two effects
on the current account deficit, and thus on the change in net debt: it in-
creases interest payments, but it also leads, through the portfolio balance
relation, to a lower exchange rate and thus a decrease in the trade deficit.
For stability to prevail, the net effect must be that the increase in net debt
reduces the current account deficit. This condition appears to be satisfied
for plausible parameter values (the next section explores this issue further),
and we assume that it is satisfied here. In this case the path of adjustment—
the saddle path—is downward sloping, as drawn in figure 1.

The Effects of a Shift toward Foreign Goods

We can now characterize the effects of shifts in preferences for goods or
assets. Figure 2 shows the effect of an unexpected and permanent increase
in z. One can think of this increase as coming either from an increase in
U.S. activity relative to foreign activity, or from a shift in exports or im-
ports at a given level of activity and a given exchange rate; we defer until

7. If we had allowed r and r* to differ, the relation would have an additional term and
take the form 0 = rF + (1 − α)(r − r*)(X − F) + D(E, z). This additional term implies that if,
for example, a country pays a lower rate of return on its liabilities than it receives on its
assets, it may be able to combine positive net debt with positive net income payments from
abroad—the situation in which the United States remains today.
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 11
Figure 1. Determination of Exchange Rate and Net Debt in Steady State

Exchange rate (E)

Portfolio balance

Current account balance

Net debt (F)

Source: Authors’ model described in the text.

later a discussion of the sources of the actual shift in z over the past decade
in the United States.
For any given level of net debt, current account balance requires a lower
exchange rate: the current account balance locus shifts down. The new
steady state is at point C, associated with a lower exchange rate and a
larger net debt.
Valuation effects imply that any unexpected depreciation leads to an
unexpected decrease in the net debt position. If we denote by ∆E the un-
expected change in the exchange rate at the time of the shift, it follows
from equation 3 that the change in net debt at the time of the shift is given
by
∆E
(4) ∆F = (1 − α ) (1 + r*) ( X − F ) .
E
 12 Brookings Papers on Economic Activity, 1:2005
Figure 2. Adjustment of Exchange Rate and Net Debt to an Increase in z

Exchange rate (E)

Portfolio balance

Current account balance

C

Net debt (F)

Source: Authors’ model described in the text.

The economy jumps initially from point A to point B and then converges
over time along the saddle path, from point B to point C. The shift in the
trade deficit leads to an initial, unexpected depreciation, followed by fur-
ther depreciation and net debt accumulation over time until the new steady
state is reached.
Note that the degree of substitutability between assets does not affect the
steady state; more formally, the steady state depends on α(1, s) and α*(1, s),
and so changes in αR and α*R that leave α(1, s) and α*(1, s) unchanged do
not affect the steady state. In other words, the eventual depreciation is the
same no matter how close substitutes U.S. and foreign assets are. But the
degree of substitutability plays a central role in the dynamics of adjustment
and in the relative roles of the initial unexpected depreciation and the antic-
ipated depreciation thereafter. This is shown in figure 3, which shows the
effects of three different values of αR and α*R on the path of adjustment.
(The three simulations are based on values for the parameters introduced in
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 13
Figure 3. Responses of the Exchange Rate and Net Debt to a Shift in z

Exchange rate
Percent changea

–2
Low αR
–4
Medium αR
–6

–8 High αR

Net debt
Percentage-point change

6 Low αR
Medium αR
4

2 High αR

5 10 15 20 25 30 35 40 45
Years
Source: Authors’ calculations.
a. All simulations are for a shift in z of 1 percent of U.S. GDP.

the next section. The purpose here is simply to show the qualitative proper-
ties of the paths. We return to the quantitative implications later.)
The less substitutable U.S. and foreign assets are—that is, the smaller
are αR and α*R—the smaller the initial depreciation and the higher the an-
ticipated rate of depreciation thereafter. To understand why, consider the
extreme case where the shares do not depend on rates of return: U.S. and
foreign investors want to maintain constant shares, no matter what the
relative rate of return is. In this case the portfolio balance relation (equa-
tion 2) implies that there will be no response of the exchange rate to the
 14 Brookings Papers on Economic Activity, 1:2005

unexpected change in z at the time it happens: any movement in the ex-

change rate would be inconsistent with equilibrium in the market for U.S.
assets. Only over time, as the deficit leads to an increase in net debt, will the
exchange rate decline.
Conversely, the more substitutable U.S. and foreign assets are, the larger
will be the initial depreciation, the lower the anticipated rate of deprecia-
tion thereafter, and the longer the time taken to reach the new steady state.
The limit of perfect substitutability—corresponding to the model discussed
at the start—is actually degenerate: the initial depreciation is such as to
maintain current account balance, and the economy does not move from
there on, never reaching the new steady state (and so the anticipated rate
of depreciation is equal to zero).
To summarize: In contrast to the case of perfect substitutability between
assets we saw earlier, an increase in U.S. demand for foreign goods leads
to a limited depreciation initially, a potentially large and long-lasting cur-
rent account deficit, and a steady depreciation over time.

The Effects of a Shift toward U.S. Assets

Figure 4 shows the effect of an unexpected and permanent increase in
s, that is, an increase in the demand for U.S. assets. Again we defer to
later a discussion of the potential factors behind such an increase.
By assumption, the increase in s leads to an increase in α(1, s) and a
decrease in α*(1, s). At a given level of net debt, portfolio balance requires
an increase in the exchange rate. The portfolio balance locus shifts up. The
new steady state is at point C, associated with a lower exchange rate and
larger net debt.
The dynamics are given by the path ABC. The initial adjustment of E and
F must again satisfy the condition in equation 4. So the economy jumps
from point A to point B and then converges over time from point B to
point C. The dollar initially appreciates, triggering an increase in the trade
deficit and a deterioration in the net debt position. Over time, net debt
continues to increase and the dollar depreciates. In the new equilibrium
the exchange rate is necessarily lower than before the shift: this reflects the
need for a larger trade surplus to offset the interest payments on the now-
larger U.S. net debt. In the long run the favorable portfolio shift leads to a
depreciation.
Again the degree of substitutability between assets plays an important
role in the adjustment. This is shown in figure 5, which plots the path of
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 15
Figure 4. Adjustment of Exchange Rate and Net Debt to an Increase in s

Exchange rate (E)

Portfolio balance

Current account balance A

C

Net debt (F)

Source: Authors’ model described in the text.

adjustment for three different values of αR and α*R. The less substitutable
are U.S. and foreign assets, the greater the initial appreciation and the higher
the anticipated rate of depreciation thereafter. Although the depreciation
is eventually the same in all cases (the steady state is invariant to the val-
ues of αR and α*R), the effect of portfolio shifts is more muted but longer
lasting when the degree of substitutability is high.

An Interpretation of the Past

Looking at the effects of shifts in preferences for goods and for assets
under imperfect asset substitutability suggests three main conclusions:
—Shifts in preferences toward foreign goods lead to an initial depreci-
ation, followed by a further anticipated depreciation. Shifts in preferences
toward U.S. assets lead to an initial appreciation, followed by an antici-
pated depreciation.
 16 Brookings Papers on Economic Activity, 1:2005
Figure 5. Responses of the Exchange Rates to a Shift in s

Exchange rate
Percent changea

20 Low αR

15
Medium αR

10
High αR
5

Net debt
Percentage-point change

20 Low αR
Medium αR
15

10 High αR

5 10 15 20 25 30 35 40 45
Years

Source: Authors’ calculations.

a. All specifications are for a 5-percentage-point shift in s.

—The empirical evidence suggests that both types of shifts have been
at work in the United States in the recent past. The first shift, by itself,
would have implied a steady depreciation in line with increased trade defi-
cits, whereas instead an initial appreciation was observed. The second shift
can explain why the initial appreciation has been followed by a deprecia-
tion. But it attributes the increase in the trade deficit fully to the initial appre-
ciation, whereas the evidence is of a large adverse shift in the trade balance
even after controlling for the effects of the exchange rate. (This does not do
justice to an alternative, and more conventional, monetary policy explana-
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 17

tion, in which high U.S. interest rates relative to foreign interest rates at
the end of the 1990s led to an appreciation, followed since by a deprecia-
tion. The observed relative interest rate differentials seem too small, how-
ever, to explain the movement in exchange rates.)
—Both shifts lead eventually to a steady depreciation, to a lower ex-
change rate than before the shift. This follows from the simple condition
that a larger net debt, no matter what its origin, requires larger interest
payments in steady state and thus a larger trade surplus. The lower the
degree of substitutability between U.S. and foreign assets, the higher the
expected rate of depreciation along the path of adjustment. The United
States appears to have indeed entered this depreciation phase.

How Large a Depreciation? A Look at the Numbers

The model is simple enough that one can insert some values for the param-
eters and draw the implications for the future. More generally, the model
provides a way of looking at the data, and this is what we do in this section.

Parameter Values
Consider first what we know about portfolio shares: In 2003 U.S. finan-
cial wealth, W, was $34.1 trillion, or about three times U.S. GDP of $11 tril-
lion.8 Non-U.S. world financial wealth is harder to assess. For the euro area
financial wealth was about t15.5 trillion in 2003, compared with GDP of
t7.5 trillion; Japanese financial wealth was about ¥1 quadrillion in 2004,
compared with GDP of ¥500 trillion.9 If one extrapolates from a ratio of
financial assets to GDP of about 2 for both Japan and Europe, and GDP for
the non-U.S. world of approximately $18 trillion in 2003, a reasonable esti-
mate for W*/E is $36 trillion—roughly the same as for the United States.
The net U.S. debt position, F, measured at market value, was $2.7 trillion
in 2003, up from approximate balance in the early 1990s.10 By implication,

8. Financial wealth data are from the Flow of Funds Accounts of the United States
1995–2003, table L100, Board of Governors of the Federal Reserve System, December 2004.
9. The figure for Europe is from ECB Bulletin, February 2005, table 3.1, and that for
Japan from Bank of Japan, Flow of Funds (www.boj.or.jp/en/stat/stat_f.htm).
10. The source for the numbers in this and the next paragraph is Bureau of Economic
Analysis, International Transactions, table 2, International Investment Position of the United
States at Year End, 1976–2003, June 2004.
 18 Brookings Papers on Economic Activity, 1:2005

U.S. assets, X, were W + F = $36.8 trillion ($34.1 trillion + $2.7 trillion),

and foreign assets, X*/E, were W*/E − F = $33.3 trillion ($36.0 trillion −
$2.7 trillion). Put another way, the ratio of U.S. net debt to U.S. assets,
F/X, was 7.3 percent ($2.7 trillion ÷ $36.8 trillion); the ratio of U.S. net
debt to U.S. GDP was 24.5 percent ($2.7 trillion ÷ $11.0 trillion).
In 2003 gross U.S. holdings of foreign assets, at market value, were
$7.9 trillion. Together with the value for W, this implies that the share of
U.S. wealth in U.S. assets, α, was 1 − (7.9/34.1), or 0.77. Gross foreign
holdings of U.S. assets, at market value, were $10.6 trillion. Together
with the value of W*/E, this implies that the share of foreign wealth in for-
eign assets, α*, was equal to 1 − (10.6/36.0), or 0.71.
To get a sense of the implications of these values for α and α*, note
from equation 2 that a transfer of one dollar from U.S. wealth to foreign
wealth implies a decrease in the demand for U.S. assets of (α + α* − 1)
dollars, or 48 cents.11
To summarize:
W = $34.1 trillion
W*/E = $36.0 trillion
X = $36.8 trillion
X*/E = $33.3 trillion
F = $2.7 trillion
α = 0.77
α* = 0.71.
We would like to know not only the values of the shares, but also their
dependence on the relative rate of return—the values of the derivatives αR
and α*R. Little is known about these values. Gourinchas and Rey provide
indirect evidence of the relevance of imperfect substitutability by show-
ing that a combination of the trade deficit and the net debt position helps
predict a depreciation (we return to their results later);12 this would not be
the case under perfect substitutability. However, it is difficult to back out
estimates of αR and α*R from their results. Thus, when needed below, we
derive results under alternative assumptions about these derivatives.

11. Note that this conclusion depends on the assumption we make in our model that
marginal and average shares are equal. This may not be the case.
12. Gourinchas and Rey (2005).
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 19

The next important parameter in our model is θ, the effect of the ex-
change rate on the trade balance. The natural starting point here is the
Marshall-Lerner relation:

dD dE
= [ ηimp − ηexp − 1] ,
Exports E

where ηimp and ηexp are, respectively, the elasticities of imports and exports
with respect to the real exchange rate.
Estimates of the ηs based on estimated U.S. import and export equations
range quite widely.13 In some cases the estimates imply that the Marshall-
Lerner condition (the condition that the term in brackets be positive, so that
a depreciation improves the trade balance) is barely satisfied. Estimates
used in macroeconometric models imply a value for the term in brackets
between 0.5 and 0.9. Put another way, together with the assumption that the
ratio of U.S. exports to U.S. GDP is 10 percent, they imply that a reduc-
tion of the ratio of the trade deficit to GDP by 1 percentage point requires
a depreciation of somewhere between 11 and 20 percent.
One may believe, however, that measurement error, complex lag struc-
tures, and misspecification all bias these estimates downward. An alterna-
tive approach is to derive the elasticities from plausible specifications of
utility and the pass-through behavior of firms. Using such an approach in
a model with nontradable goods, domestic tradable goods, and foreign
tradable goods, Obstfeld and Rogoff find that a 1-percentage-point decrease
in the ratio of the trade deficit to GDP requires a decrease in the real ex-
change rate of somewhere between 7 and 10 percent—a smaller deprecia-
tion than implied by the macroeconometric models.14
Which value to use is obviously crucial in assessing the scope of the
required exchange rate adjustment. We choose an estimate for the term in
brackets in the Marshall-Lerner equation of 0.7—toward the high range
of empirical estimates but lower than the Obstfeld-Rogoff elasticities.
This estimate, together with an exports-to-GDP ratio of 10 percent, implies
that a reduction in the ratio of the trade deficit to GDP of 1 percentage point
requires a depreciation of 15 percent.

13. See the survey by Chinn (2004).

14. Obstfeld and Rogoff (2004).
 20 Brookings Papers on Economic Activity, 1:2005

A Simple Exercise
We have argued that a depreciation of the dollar has two effects: a con-
ventional one through the trade balance, and another through valuation
effects. To get a sense of their relative magnitudes, consider the effects of
an unexpected depreciation in our model. More specifically, consider the
effects of an unexpected 15 percent decrease in E+1 relative to E on net
debt, F+1, in equation 3.
The first effect of the depreciation is to improve the trade balance.
Given our earlier discussion and assumptions, such a depreciation reduces
the trade deficit by 1 percent of GDP (which is why we chose to look at a
depreciation of 15 percent).
The second effect is to increase the dollar value of U.S. holdings of for-
eign assets (and to reduce the foreign currency value of foreign holdings
of U.S. assets) and thus reduce the U.S. net debt position. From equation
3 (with both sides divided by U.S. output, Y, to make the interpretation of
the magnitudes easier), this effect is given by
dF X − F dE
= − (1 − α ) (1 + r *) .
Y Y E
From the earlier discussion, (1 − α) is equal to 0.23, and (X − F )/Y to 3.
Assume that r* is equal to 4 percent. The effect of a 15 percent deprecia-
tion is then to reduce the ratio of net debt to GDP by 10 percentage points
(0.23 × 1.04 × 3 × 0.15). This implies that, after the unexpected deprecia-
tion, interest payments are lower by 4 percent times 10 percent, or 0.4 per-
cent of GDP.15 Putting things together, a 15 percent depreciation improves
the current account balance by 1.4 percent of GDP, with roughly one-third
of the improvement due to valuation effects.16
It is tempting here to ask how large an unexpected depreciation would
have to occur to lead to a sustainable U.S. current account deficit today?17

15. This computation assumes that all foreign assets held by U.S. investors are denom-
inated in foreign currency. In reality, some foreign bonds held by U.S. investors are de-
nominated in dollars. This reduces the valuation effects.
16. Lane and Milesi-Ferretti (2004) give a similar computation for a number of coun-
tries, although not for the United States.
17. This is also the question taken up by Obstfeld and Rogoff in this volume. Their
focus, relative to ours, is on the required adjustments in both the terms of trade and the real
exchange rate, starting from a micro-founded model with nontraded goods, exportables, and
importables.
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 21

Take the actual current account deficit of about 6 percent. What the “sus-
tainable” current account deficit is depends on the ratio of net debt to
GDP that the United States is willing to sustain, and on the growth rate of
GDP: if g is the growth rate of U.S. GDP, the United States can sustain a
current account deficit of g(F/Y). Assuming, for example, a nominal GDP
growth rate of 3 percent and a ratio of net debt to GDP of 25 percent (the
ratio prevailing today, but one that has no particular claim to being the
right one for this computation) implies that the United States can run a cur-
rent account deficit of 0.75 percent while maintaining a constant ratio of
net debt to GDP. In this case the depreciation required to shift from the
actual to the sustainable current account deficit would be roughly 56 per-
cent (6 percent − 0.75 percent) × (15 percent ÷ 1.4 percent).
This is a large number, and despite the uncertainty attached to the under-
lying values of many of the parameters, it is a useful number to keep in
mind. But one should be clear about the limitations of the computation:
—The United States surely does not need to shift to sustainable current
account balance right away. The rest of the world is still willing to lend to
it, if perhaps not at the current rate. The longer the United States waits,
however, the higher the ratio of net debt to GDP becomes, and thus the
larger the eventual required depreciation. In this sense our computation
gives a lower bound on the eventual depreciation.
—The computation is based on the assumption that, at the current ex-
change rate, the trade deficit will remain as large as it is today. If, for
example, we believed that part of the current trade deficit reflects the com-
bined effect of recent depreciations and J-curve effects, the computation
above would clearly overestimate the required depreciation.
The rest of this section deals with these issues. First, by returning to
dynamics, we try to get a sense of the eventual depreciation and of the rate
at which it may be achieved. Second, we look at the evidence on the ori-
gins of the shifts in z and s.

Returning to Dynamics
How large is the effect of a given shift in z (or in s) on the accumula-
tion of net debt and on the eventual exchange rate? And how long does it
take to get there? The natural way to answer these questions is to simulate
our model using the values of the parameters we derived earlier. This is
indeed what the simulations presented in figures 3 and 5 did; we now look
more closely at their quantitative implications.
 22 Brookings Papers on Economic Activity, 1:2005

Both sets of simulations are based on the values of the parameters

given above. Recognizing the presence of output growth (which we did
not allow for in the model), and rewriting the equation for net debt as an
equation for the ratio of net debt to output, we take the term in front of F
in the current account balance relation (equation 3) to stand for the inter-
est rate minus the growth rate. We choose an interest rate of 4 percent and
a nominal growth rate of 3 percent, so that their difference is 1 percent.
We write the portfolio shares as
α ( R e, s ) = a + bR e + s, α*( R e, s ) = a* − bR e − s.
The simulations show the results for three values (10, 1.0, and 0.1) of the
parameter b. A value of 1 implies that an increase in the expected relative
return on U.S. assets of 100 basis points increases the desired share in U.S.
assets by 1 percentage point.
Figure 3 showed the effect of an increase in z of 1 percent of U.S. GDP.
Figure 5 showed the effect of an increase in s of 5 percentage points, lead-
ing to an increase in α and a decrease in α* of 5 percentage points at a
given relative rate of return. Time is measured in years.
Figure 3 leads to two main conclusions. First, the effect of a permanent
increase in z by 1 percent is to eventually increase the ratio of net debt to
GDP by 17 percentage points and require an eventual depreciation of
12.5 percent. (Recall that the long-run effects are independent of the degree
of substitutability between assets—that is, independent of the value of b.)
Second, it takes a long time to get there: the figure is truncated at fifty
years, by which time the adjustment is still not complete.
Figure 5 leads to similar conclusions. The initial effect of the increase in
s is an appreciation of the dollar: by 23 percent if b = 0.1, and by 12 percent
if b = 10. The long-run effect of the increase in s is an increase in the
ratio of U.S. net debt to GDP of 35 percentage points and a depreciation of
15 percent. But even after fifty years the adjustment is far from complete,
and the exchange rate is still above its initial level.
What should one conclude from these exercises? We conclude that,
under the following assumptions—that there are no anticipated changes in
z or in α or α*, that investors have been and will be rational (the simula-
tions are carried out under rational expectations), and that there are no
surprises—the dollar will depreciate by a large amount, but at a steady
and slow rate. There are good reasons to question each of these assump-
tions, and this we do next.
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 23

A Closer Look at the Trade Deficit

To think about the likely path of z, and thus of the path of the trade
deficit at a given exchange rate, it is useful to write the trade deficit as
the difference between the value of imports in terms of domestic goods,
and exports:
D ( E , z ) ≡ E imp ( E , Z , z
) − exp ( E , Z *, z
*)
We have decomposed z into two components: total U.S. spending, Z, and
z̃, which represents shifts in the relative U.S. demand for U.S. versus for-
eign goods, at a given level of spending and a given exchange rate. Simi-
larly, z* is decomposed into Z* and z̃*, the latter measuring shifts in the
relative foreign demand for U.S. versus foreign goods.
Most of the large current account fluctuations in developed countries of
the last few decades have come from relative fluctuations in activity, that
is, in Z relative to Z*.18 It has indeed been argued that the deterioration of
the U.S. trade balance has come mostly from faster growth in the United
States than in its trade partners, leading imports by the United States to
increase faster than U.S. exports to the rest of the world. This appears,
however, to have played a limited role. Europe and Japan indeed have had
slower growth than the United States (U.S. output grew a cumulative
45 percent from 1990 to 2004, compared with 29 percent for the euro area
and 25 percent for Japan), but these countries account for only 35 percent
of U.S. exports, and meanwhile other U.S. trade partners have grown as
fast as or faster than the United States. Indeed, a study by the International
Monetary Fund finds nearly identical output growth rates for the United
States and its export-weighted partners since the early 1990s.19
Some have argued that the deterioration in the trade balance reflects
instead a combination of rapid growth both in the United States and abroad
and a U.S. import elasticity with respect to domestic spending that is higher
(1.5 or above) than the elasticity of U.S. exports with respect to foreign
spending. In this view rapid U.S. growth has led to a more than propor-
tional increase in imports and an increasing trade deficit. The debate about

18. For a review of current account deficits and adjustments for twenty-one countries
over the last thirty years, and references to the literature, see Debelle and Galati (2005).
19. International Monetary Fund, Article IV United States Consultation—Staff Report,
2004. As the case of the United States indeed reminds us, output is not the same as domes-
tic spending, but the differences in growth rates between the two over a decade are small.
 24 Brookings Papers on Economic Activity, 1:2005

the correct value of the U.S. import elasticity is an old one, dating back
to the estimates by Hendrik Houthakker and Stephen Magee; we tend to
side with the recent conclusion by Jaime Marquez that the elasticity is close
to 1.20 For our purposes, however, this discussion is not relevant. Whether the
growth in the U.S. trade deficit is the result of a high import elasticity or of
shifts in the z̃s, there are no obvious reasons to expect either the shift to
reverse or growth in the United States to drastically decrease in the future.
One way of assessing the relative roles of shifts in spending, the ex-
change rate, and other factors is to look at the performance of import and
export equations in detailed macroeconometric models. The numbers ob-
tained using the macroeconometric model of Global Insight (formerly the
Data Resources, Inc., or DRI, model) are as follows:21 The U.S. trade deficit
in goods increased from $221 billion in the first quarter of 1998 (annual-
ized) to $674 billion in the third quarter of 2004. Of this $453 billion
increase, $126 billion was due to the increase in the value of oil imports,
leaving $327 billion to be explained. When the export and import equa-
tions of the model are used, activity variables and exchange rates explain
$202 billion, or about 60 percent of the increase. Unexplained time trends
and residuals account for the remaining 40 percent, a substantial amount.22
Looking to the future, whether growth rate differentials, Houthakker-
Magee effects, or unexplained shifts are behind the increase in the trade
deficit is probably not essential. The slower growth in Europe and Japan
reflects in large part structural factors, and neither Europe nor Japan is
likely to make up much of the cumulative growth difference since 1995
over the next few years. One can still ask how much a given increase in
growth in Europe and Japan would reduce the U.S. trade deficit. A simple
computation is as follows. Suppose that Europe and Japan made up the
roughly 20-percentage-point growth gap they have accumulated since 1990
vis-à-vis the United States—an unlikely scenario in the near future—so that
U.S. exports to Europe and Japan increased by 20 percent. Given that U.S.

20. Houthakker and Magee (1969); Marquez (2000).

21. We thank Nigel Gault of Global Insight for communicating these results to us.
22. The model has a set of export and import equations disaggregated by product type.
Most of the elasticities of the different components with respect to domestic or foreign
spending are close to 1, indicating that Houthakker-Magee effects play a limited role (ex-
cept for imports and exports of consumption goods, where the elasticity of imports with re-
spect to consumption is 1.5 for the United States, but the elasticity of U.S. exports with
respect to foreign GDP is an even higher 2.0).
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 25

exports to these countries are currently about $350 billion, the improve-
ment would be 0.7 percent of U.S. GDP—not negligible, but not a major
increase either.
One other factor, however, may hold more hope for a reduction in the
trade deficit, namely, the working out of the J-curve. Nominal deprecia-
tions increase import prices, but these decrease the volume of imports only
with a lag. Thus, for a while, a depreciation can increase the value of im-
ports and worsen the trade balance, before improving it later.
One reason to think this may be important is the “dance of the dollar”
and the movements of the dollar and the current account during the 1980s.
From the first quarter of 1979 to the first quarter of 1985, the real exchange
rate of the United States (measured by the trade-weighted major currencies
index constructed by the Federal Reserve Board) increased by 41 percent (log
percentage change). This appreciation was then followed by a sharp depreci-
ation, with the dollar falling by 44 percent from the first quarter of 1985 to
the first quarter of 1988. The appreciation was accompanied by a steady de-
terioration in the current account deficit, from rough balance in the early
1980s to a deficit of about 2.5 percent of GDP when the dollar reached its
peak in early 1985. The current account continued to worsen, however, for
more than two years, reaching a peak of 3.4 percent of GDP in 1987. The
divergent paths of the exchange rate and the current account from 1985 to
1987 led a number of economists to explore the idea of hysteresis in trade:23
the notion that, once appreciation has led to a loss of market share, an equal
depreciation may not be sufficient to reestablish trade balance. Just as the
idea was taking hold, however, the current account position rapidly im-
proved, and trade was roughly in balance by the end of the decade.24
The parallels with more recent developments are clear from figure 6,
which plots the dollar exchange rate and the U.S. current account during
both episodes, aligned in the figure so that the dollar peak of 1985:1 co-
incides with the dollar peak of 2001:2. The figure suggests two conclusions:
—If the earlier episode is a reliable guide, and the lags today are simi-
lar to those that prevailed in the 1980s, the current account deficit may start

23. In particular, Baldwin and Krugman (1987).

24. These issues were discussed at length in the Brookings Papers at the time. Besides
Baldwin and Krugman (1987), see, for example, Cooper (1986), Dornbusch (1987), and
Sachs (1988), with post mortems by Lawrence (1990) and Krugman (1991). Another
much-discussed issue, to which we return later, was the relative roles of fiscal deficit reduc-
tion and exchange rate adjustment in closing the deficit.
 26 Brookings Papers on Economic Activity, 1:2005
Figure 6. Current Account Deficit and Effective Real Exchange Rate,
1978–93 and 1995–2004

Current account deficit

Percent of GDP

5 1995–2004

2
1978–93
1

–1

Exchange ratea
Index (March 1973 = 100)

1978–93
120

110

100
1995–2004
90

80

1980 1982 1984 1986 1988 1990 1992

1997 1999 2001 2003 2005
Year

Source: Bureau of Economic Analysis, Table 1, U.S. International Transactions; Federal Reserve data.
a. Price-adjusted Major Currencies index.
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 27

to turn around soon. Today’s deficit, however, is much larger than the ear-
lier deficit was at its peak in 1987 (approaching 6 percent of GDP versus
3.5 percent), and the depreciation so far has been more limited (23 percent
from 2001:2 to 2004:4, compared with 33 percent over the equivalent period
from 1985:1 to 1988:3).25
—Hence one can surely not conclude that the depreciation so far is
enough to restore the current account deficit to sustainable levels. But it
may be that, in our computation, the appropriate place to start is from a
J-curve-adjusted ratio of the current account deficit to GDP of 4 or 5 per-
cent instead of 6 percent.26 If we choose 4 percent—a very optimistic
assumption—the remaining required depreciation is 34 percent (4 percent
− 0.75 percent) × (15 percent ÷ 1.4 percent).27

A Closer Look at Portfolio Shares

One striking aspect of the simulations presented above is how slow the
depreciation is along the adjustment path. This is in contrast with some pre-
dictions of much more abrupt falls in the dollar in the near future.28 This
raises two issues: Can the anticipated depreciation be greater than in these
simulations? And are there possible surprises under which the depreciation
might be much faster (or slower), and, if so, what are they?
To answer the first question, we go back to the model. We noted earlier
that the lower the degree of substitutability between assets, the higher the
anticipated rate of depreciation. So, by assuming zero substitutability—that
is, constant asset shares except for changes coming from shifts in s—we can

25. On the other hand, the gross positions, and thus the scope for valuation effects from
dollar depreciation, are much larger now than they were then. In 1985 gross U.S. holdings
of foreign assets were $1.5 trillion, compared with $8 trillion today.
26. Forecasts by Macroeconomic Advisers, LLC, are for an improvement in the trade
balance of $75 billion, or less than 1 percent of GDP, over the next two years. (The forecast
is based on a depreciation of the dollar of 4 percent over that period.) The residuals of the
import price equations of the model, however, suggest an unusually low pass-through of
the dollar decline to import prices over the recent past, and the forecast assumes that the
low pass-through continues. If the pass-through were to return to its historical average, the
improvement in the trade balance would be larger.
27. This number is surprisingly close to the 33 percent obtained by Obstfeld and
Rogoff in this volume.
28. For example, by Roubini and Setser (2005).
 28 Brookings Papers on Economic Activity, 1:2005

derive an upper bound on the anticipated rate of depreciation. Differentiat-

ing equation 2 gives
dE (α + α* −1) X dF ( X − F ) dα − ( X * E + F ) dα*
= − + .
E (1 − α*) X * E X (1 − α*) X * E
In the absence of anticipated shifts in shares (so that the second term
equals zero), the anticipated rate of depreciation depends on the change in
the ratio of U.S. net debt to U.S. assets: the faster the increase in net debt,
the faster the decrease in the relative demand for U.S. assets, and there-
fore the higher the rate of depreciation needed to maintain portfolio bal-
ance. Using the parameters we constructed earlier, this equation implies
dE F
= −1.8d + ( 3.5 dα − 3.7 dα*) .
E X
Suppose shares remain constant. If we take the annual increase in the
ratio of net debt to U.S. GDP to be 5 percent and the ratio of U.S. GDP to
U.S. assets to be one-third, this gives an anticipated annual rate of depre-
ciation of 3 percent a year (1.8 × 0.05 ÷ 3).29
If, however, shares of U.S. assets in the portfolios of either domestic or
foreign investors are expected to decline, the anticipated depreciation can
clearly be much larger. If, for example, we anticipate that the share of U.S.
assets in foreign portfolios will decline by 2 percent over the coming year,
the anticipated depreciation is 8.7 percent (2.7 percent as calculated above,
plus 3.0 times 2 percent). This is obviously an upper bound on the size of
the anticipated depreciation, derived by assuming that private investors are
willing to keep a constant share of their wealth in U.S. assets despite a high
negative expected rate of return between now and then. (If, instead, antici-
pating this high negative rate of return, private investors decide to decrease
their share of dollar assets, then some of the depreciation will take place
now, rather than when the shift in portfolio composition occurs, and so the
anticipated depreciation will be smaller.) Still, it implies that, under imper-
fect substitutability, and under the assumption that desired shares in U.S.

29. Although comparison is difficult, this rate appears lower than that implied by the
estimates of Gourinchas and Rey (2005). Their results imply that a combination of net debt
and trade deficits 2 standard deviations from the mean—a situation that would appear to
characterize well the United States today—implies an anticipated annual rate of deprecia-
tion of about 5 percent over the following two years.
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 29

assets will decrease, it is logically acceptable to predict a substantial depre-

ciation of the dollar in the near future.
Are there good reasons to expect these desired shares to decrease in the
near future? This is the subject of a contentious debate. Some argue that the
United States can continue to finance its current account deficits at today’s
level for a long time to come at the same exchange rate. They argue that
the poor development of financial markets in Asia and elsewhere, together
with the need for Asian countries to accumulate international collateral,
implies a steadily increasing relative demand for U.S. assets. They point
to the latent demand for U.S. assets on the part of Chinese private investors,
currently limited by capital controls. In short, they argue that foreign in-
vestors will be willing to further increase their holdings of U.S. assets for
many years to come.30
Following this argument, we can ask what increase in shares—say,
what increase in (1 − α*), the share of U.S. assets in foreign portfolios—
would be needed to absorb the current increase in net debt at a given
exchange rate. From the relation derived above, setting dE/E and dα equal
to zero gives
(α* + α − 1) X
dα * = − (Y X ) d F .
X* E + F Y
For the parameters we have constructed, a change of 5 percentage points
in F/Y requires an increase in the share of U.S. assets in foreign portfolios
of about 0.8 percentage point a year (0.47 × 5 percent ÷ 3).31
We find more plausible the argument that the relative demand for U.S.
assets may actually decrease rather than increase in the future. This argu-
ment is based, in particular, on the fact that much of the recent accumula-
tion of U.S. assets has taken the form of accumulation of reserves by the
Japanese and Chinese central banks. Many worry that this will not last,

30. See, for example, Dooley, Folkerts-Landau, and Garber (2004) and Caballero,
Farhi, and Hammour (2004).
31. A related argument is that, to the extent that the rest of the world is growing faster
than the United States, an increase in the ratio of net debt to GDP in the United States is
consistent with a constant share of U.S. assets in foreign portfolios. This argument falls
quantitatively short: although some Asian countries are growing rapidly, their weight and
their financial wealth are still far too small to absorb the U.S. current account deficit while
maintaining constant shares of U.S. assets in their portfolios.
 30 Brookings Papers on Economic Activity, 1:2005

that the pegging of the renminbi will come to an end, or that both central
banks will want to change the composition of their reserves away from
U.S. assets, leading to further depreciation of the dollar. Our model pro-
vides a simple way of discussing the issue and thinking about the numbers.
Consider pegging first: the foreign central bank buys or sells dollar assets
so as to keep E = Ē.32 Let B denote the reserves (U.S. assets) held by the
foreign central bank, so that
X = B + α (1) ( X − F ) + (1 − α* (1)) ( X * E + F ) .
Figure 7 illustrates the resulting dynamics. Suppose that, in the absence of
pegging, the steady state is given by point A and that the foreign central
bank pegs the exchange rate at Ē. At that level the U.S. current account is in
deficit, and so F increases over time. Wealth gets steadily transferred to the
foreign country, and so the private demand for U.S. assets steadily decreases.
To keep E unchanged, B must increase further over time. Pegging by the
foreign central bank is thus equivalent to a continuous outward shift in the
portfolio balance schedule: in effect, the foreign central bank is keeping
world demand for U.S. assets unchanged by offsetting the fall in private
demand. Pegging leads to a steady increase in U.S. net debt and a steady in-
crease in the foreign central bank’s reserves, offsetting the steady decrease
in private demand for U.S. assets (represented by the path DC in figure 7).
What happens when the foreign central bank unexpectedly stops pegging?
From point C just before the peg is abandoned, the economy jumps to point
G (recall that valuation effects lead to a decrease in net debt, and therefore a
capital loss for the foreign central bank, when there is an unexpected depre-
ciation) and then adjusts along the saddle-point path GA′. The longer the
peg lasts, the larger the initial and the eventual depreciation.
In other words, an early end to the Chinese peg would obviously lead
to a depreciation of the dollar (an appreciation of the renminbi). But the
sooner it takes place, the smaller the required depreciation, both initially
and in the long run. Put another way, the longer the Chinese wait to aban-
don the peg, the larger the eventual appreciation of the renminbi.
The conclusions are very similar with respect to changes in the compo-
sition of reserves. We can think of such changes as changes in portfolio

32. Our two-country model has only one foreign central bank, and so we cannot discuss
what happens if one foreign bank pegs its currency and the others do not. The issue is, how-
ever, relevant in thinking about the paths of the dollar-euro and the dollar-yen exchange
rates. We discuss this further in the next section.
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 31
Figure 7. Adjustment of Exchange Rate and Net Debt to Abandonment of Foreign Peg

Exchange rate (E)

Portfolio balance

D
C
E
Current account balance
A G

A

Net debt (F)

Source: Authors’ model described in the text.

preferences, this time not by private investors but by central banks, and so
we can apply our earlier analysis directly. A shift away from U.S. assets
will lead to an initial depreciation, leading in turn to a lower current account
deficit, a smaller increase in net debt, and thus to a smaller depreciation in
the long run.
How large might these shifts be? Chinese reserves currently equal
$610 billion, and Japanese reserves are $840 billion. Assuming that these
reserves are now held mostly in dollars, if the People’s Bank of China and
the Bank of Japan reduced their dollar holdings to half of their portfolio,
this would represent a decrease in the share of U.S. assets in total foreign
(private and central bank) portfolios, (1 − α*), from 30 percent to 28 per-
cent. The computations we presented earlier suggest that this would be a
substantial shift, leading to a decrease in the dollar exchange rate possibly
as large as 8.7 percent.
 32 Brookings Papers on Economic Activity, 1:2005

To summarize: Avoiding a depreciation of the dollar would require a

steady and substantial increase in shares of U.S. assets in U.S. or foreign
portfolios at a given exchange rate. This seems unlikely to hold for very long.
A more likely scenario is the opposite, a decrease in shares, due in particular
to diversification of reserves by central banks. If and when this happens, the
dollar will depreciate. Note, however, that the larger the adverse shift, the
larger the initial depreciation but the smaller the accumulation of debt there-
after, and therefore the smaller the eventual depreciation. “Bad news” on the
dollar now may well be good news in the long run (and vice versa).

The Path of Interest Rates

Our model takes interest rates as given, and the discussion thus far has
taken them as constant.33 Yield curves in the United States, Europe, and
Japan indeed indicate little expected change in interest rates over the near
and the medium term. However, it is easy to think of scenarios where
changes in interest rates play an important role, and this leads us to dis-
cuss the role of budget deficit reduction in the adjustment process.
First, however, we briefly show the effects of an increase in the U.S.
interest rate in our model. Figure 8 shows the effects of an unexpected
permanent increase in r over r*. (In contrast to the case of perfect sub-
stitutability, it is possible for the two interest rates to differ even in the
steady state.) The portfolio balance locus shifts upward: At a given level
of net debt, U.S. assets are more attractive, and so the exchange rate in-
creases. The current account balance locus shifts down: the higher inter-
est rate implies larger payments on foreign holdings of U.S. assets and
thus requires a larger trade surplus, and in turn a lower exchange rate. The
adjustment path is given by ABC. In response to the increase in r, the
economy jumps from point A to point B and then moves over time from
point B to point C. As drawn, there is an appreciation initially, but, in gen-
eral, the initial effect on the exchange rate is ambiguous. If gross liabili-

33. Remember that, when financial assets are imperfect substitutes, the interest rate dif-
ferential no longer directly reflects expected exchange rate changes. It is thus perfectly
rational for the level of long-term interest rates in the United States and in other countries
to be very similar, even as the market anticipates a depreciation of the dollar. Therefore, if
we consider that financial assets denominated in different currencies can be imperfect sub-
stitutes, there is no “interest rate puzzle,” contrary to what is sometimes claimed in the
financial press.
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 33
Figure 8. Adjustment of Exchange Rate and Net Debt to an Increase in the Domestic
Interest Rate

Exchange rate (E)

Portfolio balance

Current account balance

C

Net debt (F)

Source: Authors’ model described in the text.

ties are large, for example, the effect of higher interest payments on the
current account balance may dominate the more conventional “overshoot-
ing” effects of increased attractiveness and lead to an initial depreciation
rather than an appreciation. In either case the steady-state effect is greater net
debt accumulation, and thus a larger depreciation than if r had not increased.
Thus, under the assumption that an increase in interest rates leads ini-
tially to an appreciation, an increase in U.S. interest rates beyond what is
already implicit in the yield curve would delay the depreciation of the
dollar, at the cost of greater net debt accumulation and a larger eventual
depreciation.
Interest rate changes, however, do not take place in a vacuum. It is more
interesting to think about what may happen to interest rates as the dollar
depreciates, either slowly along the saddle path or more sharply, in response,
 34 Brookings Papers on Economic Activity, 1:2005

for example, to adverse portfolio shifts. As the dollar depreciates, relative

demand shifts toward U.S. goods, reducing the trade deficit but also increas-
ing total demand for U.S. goods. Suppose also that output is initially at its
natural level (the level associated with the natural rate of unemployment),
which appears to be a good description of the United States today. Three
outcomes are possible:
—Interest rates and fiscal policy remain unchanged. The increase in
demand leads to an increase in output but also an increase in imports,
which partly offsets the effect of the depreciation on the trade balance. (In
terms of our model, it leads to an increase in domestic spending, Z, and
thus to a shift in z.)
—Interest rates remain unchanged, but fiscal policy is adjusted to off-
set the increase in demand and leave output at its natural level; in other
words, the budget deficit is reduced so as to maintain internal balance.
—Fiscal policy remains unchanged, but the Federal Reserve increases
interest rates so as to maintain output at its natural level. In this case, higher
U.S. interest rates limit the extent of the depreciation and mitigate the cur-
rent account deficit reduction. In doing so, however, they lead to larger net
debt accumulation and to a larger eventual depreciation.
In short, an orderly reduction of the current account deficit—that is, one
that occurs while maintaining internal balance—requires both a decrease
in the exchange rate and a reduction in the budget deficit.34 The two are not
substitutes: the depreciation is needed to achieve current account balance,
and budget deficit reduction is needed to maintain internal balance at the nat-
ural level of output.35 (The frequently heard statement that deficit reduction
would reduce the need for dollar depreciation leaves us puzzled.) If the
decrease in the budget deficit is not accompanied by a depreciation, the
result is likely to be lower demand and a recession. Although the recession

34. Many of the discussions at Brookings in the late 1980s were about the relative roles
of budget deficit reduction and exchange rate adjustment. For example, Sachs (1988)
argued that “the budget deficit is the most important source of the trade deficit. Reducing
the budget deficit would help reduce the trade deficit . . . [while] an attempt to reduce the
trade deficit by a depreciating exchange rate induced by easier monetary policy would pro-
duce inflation with little benefit on the current account,” a view consistent with the third
scenario above. Cooper (1986), in a discussion of the policy package best suited to elimi-
nate the U.S. imbalances, stated, “The drop in the dollar is an essential part of the policy
package. The dollar’s decline will help offset the fiscal contraction through expansion of
net exports and help maintain overall U.S. economic activity at a satisfactory level,” a view
consistent with the second scenario.
35. Obstfeld and Rogoff (2004) emphasize a similar point.
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 35

would reduce the current account deficit, this is hardly a desirable out-
come. If the depreciation is not accompanied by a reduction in the budget
deficit, one of two things can happen: demand will increase, and with it the
risk that the economy will overheat, or, more likely, interest rates will
increase so as to maintain internal balance. This increase would either limit
or delay the depreciation of the dollar, but, as we have made clear, this
would be a mixed blessing. Such a delay implies less depreciation in the short
run but more net debt accumulation and more depreciation in the long run.

The Euro, the Yen, and the Renminbi

The depreciation of the dollar since the peak of 2002 has been very
unevenly distributed: as of April 2005 the dollar had fallen 45 percent
against the euro, 25 percent against the yen, and not at all against the ren-
minbi. In this section we return to the questions asked in the introduction:
if substantially more depreciation is indeed to come, against which cur-
rencies will the dollar fall? If China abandons its peg, or if Asian central
banks diversify their reserves, how will the euro and the yen be affected?
The basic answer is simple. Along the adjustment path, what matters—
because of home bias in asset preferences—is the reallocation of wealth
across countries, and thus the bilateral current account balances of the
United States with its partners. Wealth transfers modify countries’ relative
demands for assets, thus requiring corresponding exchange rate movements.
Other things equal, countries with larger trade surpluses with the United
States will see a larger appreciation of their currency.
Other things may not be equal, however. Depending on portfolio prefer-
ences, a transfer of wealth from the United States to Japan, for example,
may change the relative demand for euro assets and thus the euro exchange
rate. In that context one can think of central banks as investors with dif-
ferent asset preferences. For example, a central bank that holds most of its
reserves in dollars can be thought of as an investor with strong dollar pref-
erences. Any increase in its reserves is likely to lead to an increase in the
relative demand for dollar assets and thus an appreciation of the dollar. Any
diversification of its reserves is likely to lead to a depreciation of the dollar.
It is beyond the scope of this paper to construct and simulate a realistic
multicountry portfolio model. But we can make some progress in thinking
about mechanisms and magnitudes. The first step is to extend our model
to allow for more countries.
 36 Brookings Papers on Economic Activity, 1:2005

Extending the Portfolio Model to Four Regions

In 2004 the U.S. trade deficit in goods (the only component of the cur-
rent account for which a decomposition of the deficit by country is avail-
able) was $665 billion. Of this, $162 billion was with China, $77 billion
with Japan, $85 billion with the euro area, and the remainder, $341 bil-
lion, with the rest of the world. We ignore the rest of the world here and
think of the world as composed of four countries or regions: the United
States, Europe, Japan, and China (indexed 1 through 4, respectively). We
shall therefore think of China as accounting for roughly half the U.S. cur-
rent account deficit, and Europe and Japan as accounting each for roughly
one-fourth.
We extend our portfolio model as follows. We assume that the share of
asset j in the portfolio of country i is given by
α ij(⋅) = aij + ∑ k β ijk R e ,
k

where Rek is the expected gross real rate of return, in dollars, from holding
assets of country k (so that Rek denotes a rate of return, not a relative rate of
return as in our two-country model).
We assume further that βijk = βjk, so that the effect of the return on asset
k on demand for asset j is the same for all investors, independent of the
country of origin. This implies that differences in portfolio preferences
across countries show up only as different constant terms, and derivatives
with respect to rates of return are the same across countries.
The following restrictions apply: From the budget constraint (the con-
dition that the shares sum to 1, for any set of expected rates of return), it
follows that Σj aij = 1 for all i, and Σj βjk = 0 for all k. The home bias
assumption takes the form Σi aii > 1. The demand functions are assumed
to be homogeneous of degree zero in expected gross rates of return, so
that Σk βjk = 0 for all j.
Domestic interest rates, in domestic currency, are assumed to be con-
stant and all equal to r. Exchange rates, Ek, are defined as the price of
U.S. goods in terms of foreign goods (so that E1 = 1, and an increase in E2,
for example, indicates an appreciation of the dollar against the euro—or,
equivalently, a depreciation of the euro against the dollar). It follows that
the expected gross real rate of return, in dollars, from holding assets of
country k is given by Rek = (1 + r)Ek/Ek+1. In steady state Rek = (1 + r), so that
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 37

Σk βjkRek = 0, and we can concentrate on the aij elements. The portfolio bal-
ance conditions, absent central bank intervention, are given by
Xj X 
= ∑ i aij  i − Fi  ,
Ej  Ei 
where Fi denotes the net foreign debt position of country i, so that Σi Fi = 0.
So far we have treated all four countries symmetrically. China, how-
ever, is special in two respects: it enforces strict capital controls, and it
pegs the renminbi to the dollar. We capture these two features as follows:
—We formalize capital controls as the assumption that a4i = ai4 = 0 for
all i ≠ 4; that is, capital controls prevent Chinese residents from investing
in foreign assets but also prevent investors outside China from acquiring
Chinese assets.36
—We assume that, to peg the renminbi-dollar exchange rate (E4 = 1),
the People’s Bank of China passively acquires all dollars flowing into
China: the wealth transfer from the United States to the euro area and
Japan is thus the U.S. current account minus the fraction that is financed
by the Chinese central bank: dF2 + dF3 = −dF1 − dF4.

Some Simple Computations

Consider now an increase in U.S. net debt equal to dF1. Assume that a
share γ of the U.S. net debt is held by China. Assume that a fraction x of
the remaining portion is held by the euro area and a fraction (1 − x) by
Japan, so that the changes in net debt are given by
dF2 = − x (1 − γ ) dF1 , dF3 = − (1 − x ) (1 − γ ) dF1 , dF4 = − γ dF1 .
Assume further that China imposes capital controls and pegs the renminbi,
that the other three economies are all the same size, and that the matrix of
aij elements is symmetric in the following way: aii = a and aij = c = (1 − a)/
2 < a for i ≠ j.37 In other words, investors want to put more than one-third

36. This ignores inflows of foreign direct investment into China, but since we are con-
sidering the financing of the U.S. current account deficit, this assumption is inconsequential
for our analysis.
37. The assumption of countries of equal size allows us to specify the matrix in a sim-
ple and transparent way. Allowing countries to differ in size, as they obviously do, would
lead to a more complex, size-adjusted matrix; but the results would be unaffected.
 38 Brookings Papers on Economic Activity, 1:2005

of their portfolio into domestic assets (the conditions above imply a > 1⁄3)
and allocate the rest of their portfolio equally among foreign assets.
Under these assumptions, dE4 = 0 (because of pegging), and dE2 and
dE3 are given by

dE2 ( a − c ) (1 − γ )[ x (1 − a ) + c (1 − x )] cγ
=− +
(1 − a ) − c − a−c
2
dF1 2 1

dE 3 ( a − c ) (1 − γ )[ xc + (1 − a ) (1 − x )] cγ
=− + .
(1 − a ) − c 2 1− a − c
2
dF1

Consider first the effects of γ, the share of U.S. net debt held by China:
—For γ = 0, dE2/dF1 and dE3/dF1 are both negative. Not surprisingly,
an increase in U.S. net debt leads to a depreciation of the dollar against
both the euro and the yen.
—As γ increases, the depreciation of the dollar against the euro and the
yen becomes smaller. This, too, is not surprising. What may be more sur-
prising, however, is that, for high values of γ, the depreciation turns into
an appreciation. For γ = 1, for example, the dollar appreciates against both
the euro and the yen. The explanation is straightforward and is found in
portfolio preferences: The transfer of wealth from the United States to
China is a transfer of wealth from U.S. investors, who are willing to hold
dollar, euro, and yen assets, to the People’s Bank of China, which holds
only dollars. This transfer to an investor with extreme dollar preferences
leads to a relative increase in the demand for dollars and hence an appre-
ciation of the dollar against both the euro and the yen.
Consider now the effects of x, the share of the U.S. net debt held by
Europe, excluding the net debt held by China (for simplicity, we set γ
equal to zero):
—Consider first the case where x = 0, so that the accumulation of net
debt is entirely vis-à-vis Japan. In this case, it follows that dE3/dF1 =
2 dE2/dF1. Both the yen and the euro appreciate against the dollar, with the
yen appreciating twice as much as the euro. This result might again be sur-
prising: why should a transfer of wealth from the United States to Japan
lead to a change in the relative demand for euros? The answer is that it
does not. The euro appreciates against the dollar but depreciates against
the yen. The real effective exchange rate of the euro remains unchanged.
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 39

—If x = 1⁄2 (which seems to correspond roughly to the ratio of trade deficits
and thus to the relative accumulation of U.S. net debt today), then obviously
the euro and the yen appreciate in the same proportion against the dollar.
This simple framework also allows us to think about what would hap-
pen if China stopped pegging, or diversified its reserves away from dol-
lars, or relaxed capital controls on Chinese and foreign investors, or any
combination of these. Suppose China stopped pegging but maintained cap-
ital controls. Because the end of the peg, together with the assumption of
maintained capital controls, implies a zero Chinese surplus, the renminbi
would have to appreciate against the dollar. From then on, reserves of the
Chinese central bank would remain constant. So, as the United States con-
tinued to accumulate net debt vis-à-vis Japan and Europe, relative net debt
vis-à-vis China would decrease. In terms of our model, γ, the proportion
of U.S. net debt held by China, would decrease.38 Building on our results,
this would lead to a decrease in the role of an investor with extreme dollar
preferences, the People’s Bank of China, and would lead to an appreciation
of the euro and the yen.
Suppose instead that China diversified its reserves away from dollars.
Then, again, the demand for euros and for yen would increase, leading to
an appreciation of both currencies against the dollar.
To summarize: The trade deficits of the United States with Japan and
the euro area imply an appreciation of both the yen and the euro against
the dollar. For the time being, this effect is partly offset by the Chinese
policies of pegging and keeping most of its reserves in dollars. If China
were to give up its peg or to diversify its reserves, the euro and the yen
would appreciate further against the dollar. This last argument is at odds
with the often-heard statement that the Chinese peg has “increased the pres-
sure on the euro-dollar exchange rate,” and that therefore the abandonment of
the peg would remove some of the pressure, leading to a depreciation of the
euro against the dollar. We do not understand the logic behind that statement.

Two Simulations and a Look at Portfolios

We have looked so far at equilibrium for a given distribution of Fs.
This distribution is endogenous, however, in our model, determined by

38. Marginal γ, the proportion of the increase in U.S. net debt absorbed by China,
would equal zero.
 40 Brookings Papers on Economic Activity, 1:2005

trade deficits and portfolio preferences. We now report the results of two
simulations of our extended model.
In the first simulation we keep the symmetric portfolio assumptions
introduced above. We take the three economies to be of the same size, and
we use the values for the portfolio parameters introduced above of 0.70 for
a and 0.15 for c. We consider a shift in the U.S. trade deficit, with half of the
change in the deficit falling on China, one-fourth on Japan, and one-fourth
on the euro area. We assume that each country trades only with the United
States, so that we can focus on the bilateral balances with the United States.
We perform this simulation under two alternative assumptions about
Chinese policy. In both we assume capital controls, but in the first case we
assume that China continues to peg the renminbi, and in the second we
assume that the renminbi floats; together with the assumption of capital
controls, this implies, as indicated above, a zero Chinese trade surplus.
The top panel of figure 9 presents the results. Because of symmetry,
the responses of the euro and the yen are identical and thus represented
by the same line. The lower line shows the depreciation of the dollar against
the euro and the yen when the renminbi floats. The higher locus shows
the more limited depreciation of the dollar (and more limited appreciation
of the euro and the yen) when the renminbi is pegged and the Chinese cen-
tral bank accumulates dollars.
One may wonder whether the preferences of private investors are really
symmetric, however. Constructing portfolio shares for Japanese, European,
and U.S. investors requires rather heroic assumptions. We have nevertheless
given it a try, and the results are reported in table 1. Appendix B presents
details of the construction.
Note in table 1 the much larger share of dollar assets in European than
in Japanese portfolios. Note also the small share of Japanese assets held
by euro-area investors relative to the share of euro-area assets held by
Japanese investors (the difference is much larger than the difference in
relative size of the two economies). Portfolio preferences appear indeed
to be asymmetric.
To show what difference this asymmetry makes, the bottom panel of fig-
ure 9 presents results of a second simulation. This simulation is identical
to that in the top panel but now takes into account the relative size of the
three economies (the Xs) and uses the shares reported in table 1.
The main conclusion we draw from the bottom panel is that it looks
very similar to the top, except that the dollar depreciates initially a bit
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 41
Figure 9. Effects of a Shift in the U.S. Trade Deficit on Euro-Dollar and Yen-Dollar
Exchange Rates, with and without Chinese Pega

Symmetrical portfolio weights

Percent change

–1

–2
Under peg
–3

–4
Under float
–5

Actual portfolio weights

Percent change

–1
Dollar-euro, under peg

–2

Dollar-yen, under peg

–3

–4 Dollar-euro, under float

Dollar-yen, under float
–5

–6

5 10 15 20 25 30 35 40 45
Years

Source: Authors’ calculations.

a. All simulations assume that China maintains capital controls.
 42 Brookings Papers on Economic Activity, 1:2005

Table 1. Calculated Portfolio Shares by Investment Destinationa

Investing country
Destination United States Euro area Japan
United States 0.77 0.42 0.22
Euro area 0.15 0.53 0.15
Japan 0.08 0.05 0.63
Source: Authors’ calculations using data in appendix table B-1.
a. Investment includes both portfolio investment and foreign direct investment.

more against the yen than against the euro. This difference is due to the
larger share of dollar assets in European than in Japanese portfolios: a
dollar transferred from the United States to Europe leads to a smaller
decrease in the demand for U.S. assets than does a dollar transferred from
the United States to Japan.

Summary and Conclusions

We have argued that there have been two main forces behind the large
U.S. current account deficits of the past ten years: an increase in the U.S.
demand for foreign goods, and an increase in the foreign demand for U.S.
assets. The path of the dollar since the late 1990s can be explained as the
reaction to these forces.
The shift in portfolio preferences toward U.S. assets manifested itself
first, in the late 1990s, in the form of high private demand for U.S. equi-
ties, and more recently in the form of high central bank demand for U.S.
bonds. The shift in demand away from U.S. goods is often attributed to
more rapid growth in the United States than in its trading partners. This
appears, however, to have played only a limited role: the performance of
import and export equations in macroeconometric models shows that
activity variables and exchange rates explain only about 60 percent of the
increase in the U.S. trade deficit, with unexplained time trends and resid-
uals accounting for the rest. We interpret this as evidence of a shift in the
U.S. trade balance relation.
Either shift could have induced the observed paths of the dollar and the
U.S. current account only in a world where financial assets are imperfect
substitutes. A shift in asset preferences could not account for these paths,
because it would be meaningless in a world where assets are perfect sub-
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 43

stitutes. Nor can the shift in preferences for goods explain these paths,
because with perfect substitutability such a shift—provided it were per-
ceived as long lasting—would have induced a quicker and sharper depre-
ciation of the exchange rate and a smaller increase in the current account
than we have observed.
As a way of organizing our thoughts about the U.S. current account
deficit and the dollar, we have studied a simple model characterized by
imperfect substitutability both among goods and among assets. The model
allows for valuation effects, whose relevance has recently been empha-
sized in a number of papers. The explicit integration of valuation effects in
a model of imperfect substitutability is, we believe, novel.
We find that the degree of substitutability between assets does not affect
the steady state. In other words, the eventual dollar depreciation induced
by either shift is the same no matter how closely U.S. and foreign assets
substitute for each other. But the degree of substitutability does play a cen-
tral role in the dynamics of adjustment.
In contrast to the case of perfect substitutability between assets, an in-
crease in U.S. demand for foreign goods leads to a limited depreciation ini-
tially, a potentially large and long-lasting current account deficit, and a
slow and steady depreciation over time. An increase in foreign demand for
U.S. assets leads to an initial appreciation, followed by a slow and steady
depreciation.
The slow rate of dollar depreciation implied by imperfect substitutabil-
ity contrasts with predictions by others of much more abrupt falls in the
dollar in the near future. We show that, in the absence of anticipated port-
folio shifts, the anticipated rate of depreciation depends on the change in
the ratio of U.S. net debt to U.S. assets: the faster the increase in net debt,
the faster the decrease in the relative demand for U.S. assets, and there-
fore the higher the rate of depreciation needed to maintain portfolio bal-
ance. If we take the annual increase in the ratio of net debt to U.S. GDP to
be 5 percent, we derive an upper bound on the anticipated annual rate of
depreciation of 2.7 percent a year.
If, however, shares in U.S. assets in the portfolios of either U.S. or for-
eign investors are expected to decline, the anticipated depreciation can be
much larger. If, for example, we anticipate that central banks will diversify
their reserves away from dollars and, as a result, that the share of U.S.
assets in foreign portfolios will decline by 2 percent over the coming year,
then the anticipated depreciation may be as large as 8.7 percent. This is
 44 Brookings Papers on Economic Activity, 1:2005

obviously an upper bound on the size of the depreciation, derived by

assuming that private investors are willing to keep a constant share of their
wealth in U.S. assets despite a high expected negative rate of return
between now and then. (If, in anticipation of this high negative rate of
return, private investors decide to decrease their share of dollar assets, then
some of the depreciation will take place now, rather than at the time of the
shift in composition of reserves, and so the anticipated depreciation will be
smaller.)
On the other hand, a further shift in investors’ preferences toward dol-
lar assets would slow down, or even reverse, the path of dollar deprecia-
tion. The relief, however, would only be temporary. It would lead to an
initial appreciation, but the accompanying loss of competitiveness would
speed up the accumulation of foreign debt. The long-run value of the dol-
lar would be even lower. The argument that the United States, thanks to the
attractiveness of its assets, can keep running large current account deficits
with no effect on the dollar appears to overlook the long-run consequences
of a large accumulation of external liabilities.
For basically the same reason, an increase in interest rates would be
self-defeating. It might temporarily strengthen the dollar, but the deprecia-
tion eventually needed to restore equilibrium in the current account would
be even larger—because (as in the case of a shift in portfolio preferences)
the accumulation of foreign liabilities would accelerate, and eventually the
United States would need to finance a larger flow of interest payments
abroad. A better mix would be a decrease in interest rates and a reduction
in budget deficits to avoid overheating. (To state the obvious: tighter fiscal
policy is needed to reduce the current account deficit, but it is not a substi-
tute for the dollar depreciation. Both are needed.)
The same will happen so long as China keeps pegging the exchange rate.
One should think of the People’s Bank of China as a special investor whose
presence has the effect of raising the portfolio share of the world outside the
United States invested in dollar assets. The longer the Chinese central bank
intervenes, the larger this share. Sooner or later, however—as in the case of
Korea in the late 1980s—the People’s Bank of China will find it increas-
ingly difficult to sterilize the accumulation of reserves. Eventually, when
the peg is abandoned, the depreciation of the dollar will be larger, the longer
the peg will have lasted, because in the process the United States will have
accumulated larger quantities of foreign liabilities. Thus, if China is wor-
ried about a loss of competitiveness, pegging may be a myopic choice.
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 45

What would abandonment of the Chinese peg imply for the euro and the
yen? Contrary to a commonly heard argument, if the renminbi were allowed
to float, both currencies would be likely to appreciate further against the
dollar. The reason is that, when the People’s Bank of China stops interven-
ing, the market effectively loses an investor with extreme dollar preferences,
to be replaced by private investors with less extreme preferences. A similar
argument holds if the People’s Bank of China diversifies its reserves away
from dollar assets. For Europe and Japan, however, what matter are effec-
tive exchange rates, and their currencies may well depreciate in effective
terms even if they appreciate relative to the dollar in bilateral terms.
We end with one more general remark. A large fall in the dollar would
not by itself be a catastrophe for the United States. It would lead to higher
demand for U.S. goods and higher output, and it would offer the opportunity
to reduce budget deficits without triggering a recession. The danger is more
serious for Japan and Europe, which suffer from slow growth already and
have little room to use expansionary fiscal or monetary policy at this stage.

APPENDIX A

Dynamics of the Model

THE DYNAMICS OF the system composed of equations 2 and 3 are more eas-
ily characterized by taking the continuous time limit. In continuous time the
portfolio and current account balance equations become, respectively,
 E e    E e    X * 
X = α 1 + r − r * + , s ( X − F ) +  1 − α *  1 + r − r * + , s   + F .
E E E

  E e   E e
F = rF +  1 − α  1+ r − r * + , s  ( X − F ) + D ( E, z ).
E E
Note the presence of both expected and actual appreciation in the current
account balance equation. Expected appreciation determines the share of
the U.S. portfolio invested in foreign assets; actual appreciation deter-
mines the change in the value of that portfolio, and in turn the change in
the U.S. net debt position.
We limit ourselves to a characterization of the equilibrium and local
dynamics, using a phase diagram. (The global dynamics are more complex.
The nonlinearities imbedded in the equations imply that the economy is
 46 Brookings Papers on Economic Activity, 1:2005

likely to have two equilibriums, only one of which is potentially saddle-

point stable. This is the equilibrium we focus on.) We do so here under the
additional assumption that r = r*. The extension to differences in interest
rates, which we used to construct figure 8, is straightforward.
The locus (Ė = Ėe = 0) is obtained from the portfolio balance equation
and is downward sloping. In the presence of home bias, an increase in net
debt shifts wealth abroad, decreasing the demand for U.S. assets and re-
quiring a depreciation.
The locus (Ḟ = 0) is obtained by assuming (Ė e = Ė) in the current
account balance equation and replacing (Ė e) with its implied value from
the portfolio balance equation. This locus is also downward sloping: a
depreciation leads to a smaller trade deficit and thus allows for a larger net
debt position consistent with current account balance.
Note that the locus (Ḟ = 0) is not the same as the current account balance
locus in figure 1; that locus is derived under the assumption that both Ḟ and
Ė are zero. Using that locus makes for a simple graphical characterization
of the equilibrium but is not appropriate for studying stability or dynamics.
The derivatives αR and α*R do not affect the slope of the locus (Ė = 0)
but do affect that of the locus (Ḟ = 0). The smaller these derivatives are
(that is, the lower the degree of substitutability between assets), the closer
the locus (Ḟ = 0) is to the locus (Ė = 0). In the limit, if the degree of sub-
stitutability between U.S. and foreign assets is zero, the two loci coincide.
The larger these derivatives are (that is, the higher the degree of substi-
tutability between assets), the closer the (Ḟ = 0) locus is to the current
account balance locus, 0 = rF + D(E ,z).
The condition for the equilibrium to be saddle-point stable is that the
locus (Ė = 0) be steeper than the locus (Ḟ = 0); this turns out to be the
same as the condition given in the text, that the portfolio balance locus be
steeper than the current account balance locus. For this to hold, the fol-
lowing condition must be satisfied:
r α + α* −1
< .
EDE (1 − α*) X * E
The interpretation of this condition was given in the text. It is more
likely to be satisfied the lower the interest rate, the larger the home bias,
and the larger the response of the trade balance to the exchange rate. If the
condition is satisfied, the dynamics are as shown in figure A-1. The saddle
path is downward sloping, implying that the adjustment to the steady state
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 47
Figure A-1. Adjustment of Exchange Rate and Net Debt in Continuous Time

Exchange rate (E)

Current account balance

rF + D(E,z) = 0

F=0

E=0

Net debt (F)

Source: Authors’ model described in the text.

from below (in terms of F) is associated with an expected depreciation,

and the adjustment from above with an expected appreciation. Valuation
effects imply that unexpected shifts in z or s are associated with initial
changes in F, according to
∆E
∆ F = (1 − α ) (1 + r *) ( X − F ) .
E
The effect of the degree of substitutability on the dynamics is as follows.
The smaller are αR and α*R, the closer the locus (Ḟ= 0) is to the locus (Ė = 0),
and so the closer the saddle-point path is to the locus (Ė = 0). In the limit, if
the degree of substitutability between U.S. and foreign assets is zero, the
two loci and the saddle-point path coincide, and the economy remains on
and adjusts along the (Ė = 0) locus, the portfolio balance relation.
 48 Brookings Papers on Economic Activity, 1:2005

The larger αR and α*R, the closer the (Ḟ = 0) locus is to the locus given
by 0 = rF + D(E ,z), and the closer the saddle-point path is to that locus as
well. Also, the larger are αR and α*R, the slower is the adjustment of F and
E over time. The slow adjustment of F comes from the fact that the cur-
rent account is close to balance. The slow adjustment of E comes from the
fact that, the larger the elasticities, the smaller is Ė for a given distance
from the Ė = 0 locus.
The limiting case of perfect substitutability is degenerate. The rate of
adjustment to an unexpected, permanent shift in z goes to zero. The econ-
omy is then always on the locus 0 = rF + D(E,z). For any level of net debt,
the exchange rate adjusts so that net debt remains constant, and, in the
absence of shocks, the economy stays at that point. There is no unique
steady state, and where the economy is depends on history.

APPENDIX B

Construction of Portfolio Shares

DATA ON THE country allocation of gross portfolio investment are from the
International Monetary Fund’s Coordinated Portfolio Survey for 2002. Data
for the country allocation of direct investment are from the Organization for
Economic Cooperation and Development and likewise refer to 2002. Finan-
cial wealth for the United States, the euro area, and Japan, which we need to
compute the home bias of portfolios, are from official flow of funds data.39
From these data we construct the aij elements in two steps. First, we
compute the geographical allocation of net foreign investment positions
by weighting the shares of portfolio assets and foreign direct investment
allocated to country j by the relative importance of portfolio ( pf ) and
direct investment ( fdi) in country i’s total investment abroad. We then
scale these shares by the share of total foreign investment (1 − aii), so that
aij = {[ pf ( pf
i i
+ fdii )] aij , p + [ fdii ( pf i
+ fdii )] aij , fdi } × (1 − aii ) .
Table B-1 presents the results.

39. For the United States, see footnote 8. The source for Japan is the Bank of Japan
flow of funds data (www.boj.or.jp/en/stat/sj/stat_f.htm), and that for the euro area is the
ECB Economic Bulletin (released February, 2005 and available at www.ecb.int/pub/html/
index.en.html).
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 49

Table B-1. Calculated Portfolio Shares by Investment Destinationa

Investing country
Destination United States Euro area Japan
United States 0.77 0.19 0.17
Euro area 0.08 0.53 0.12
Japan 0.04 0.02 0.63
Rest of the world 0.11 0.27 0.08
Sources: Authors’ calculations using data from the International Monetary Fund, the Organization for Economic Cooperation
and Development, and national central banks.
a. Investment includes both portfolio investment and foreign direct investment. Shares may not sum to 1.00 because of rounding.

To perform the simulation described in the text, we then allocate the

shares invested in the “rest of the world” to foreign holdings so as to keep
the relative shares in the remaining foreign assets the same. For the United
States, for example, we increase the foreign shares in euro and yen assets
to approximately 0.15 and 0.08, respectively. This gives us the numbers
reported in table 1.
The simulation presented in figure 9 uses these values, together with
asset levels of $36.8 trillion for the United States, $23.0 trillion for the
euro area, and $8.0 trillion for Japan. Trade is assumed to be bilateral
between the United States and each of the other regions, with elasticities
of the trade balance all being equal to the elasticity used in our earlier
two-country model.
Comments and
Discussion

Ben S. Bernanke: Olivier Blanchard, Francesco Giavazzi, and Filipa Sa

have produced a gem of a paper. They introduce a disarmingly simple model,
which nevertheless provides a number of crucial insights about the joint
dynamics of the current account and the exchange rate, in both the short and
the long run. Their analysis will undoubtedly become a staple of graduate
textbooks.
The authors’ model has two features that deserve special emphasis. First,
following an older and unjustly neglected literature, the model dispenses
with the usual interest rate parity condition in favor of the assumption that
financial assets may be imperfect substitutes in investors’ portfolios; that
is, the model allows for the possibility that the demand for an asset may
depend on features other than its rate of return, such as its liquidity or its
usability as a component of international reserves. In focusing on imper-
fect asset substitutability and its implications, the authors identify an issue
that has taken on great practical significance for policymakers in recent
years. At least two contemporary policy debates turn in large part on the
extent (or the existence) of imperfect asset substitutability. One is whether
so-called nonstandard monetary policies—such as large purchases of gov-
ernment bonds or other assets by central banks—can stimulate the economy
even when the policy interest rate has hit the zero lower bound. The other
is whether sterilized foreign exchange interventions, like those recently
undertaken on a massive scale by Japan and China, can persistently alter
exchange rates and interest rates.1 The authors’ analysis explores yet another
important implication of imperfect substitutability: that, if assets denomi-
nated in different currencies are imperfect substitutes, then agents may ratio-
nally anticipate the sustained depreciation of a currency even in the absence
1. Bernanke, Reinhart, and Sack (2004) present empirical evidence relevant to both of
these debates.

50
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 51

of cross-currency interest rate differentials. Thus, by invoking imperfect

substitutability, the authors are able to show that expected dollar depreci-
ation is not necessarily inconsistent with the currently low level of U.S.
long-term nominal interest rates and the evident willingness of foreigners
to hold large quantities of U.S. assets.
The assumption that financial assets of varying characteristics are imper-
fectly substitutable in investor portfolios seems quite reasonable. (Almost
as I write these words, an announcement by the U.S. Treasury that it is
contemplating the reinstatement of the thirty-year bond seems to have trig-
gered a jump in long-term bond yields, suggestive of a supply effect on
returns.) However, both the theoretical and the empirical literatures on asset
substitutability are exceedingly thin, which is a problem for assessing the
quantitative implications of the authors’ analysis. In particular, as they them-
selves note, in their model the speed of adjustment of the exchange rate
and the current account depends importantly on the elasticities of foreign
and domestic asset demands with respect to expected return differentials,
numbers that are difficult to pin down with any confidence. Further com-
plications arise if, as is plausibly the case, the degree of asset substitutability
is not a constant but varies over time or across investors. For example, if
private investors view assets denominated in different currencies as more
substitutable than central banks do, which seems likely, then changes in
the share of assets held by each type of investor will have implications for
exchange rate dynamics. Finding satisfying microfoundations for the phe-
nomenon of imperfect asset substitutability, and obtaining persuasive esti-
mates of the degree of substitutability among various assets and for different
types of investors, should be high on the profession’s research agenda.
The second feature of the authors’ analysis worth special note is its atten-
tion to the long-run steady state. By integrating short-run and long-run
analyses, the authors obtain some useful insights that a purely short-run
approach does not deliver. Notably, they demonstrate that factors affect-
ing the value of the dollar or the size of the U.S. current account deficit
may have opposite effects in the short and in the long run. For example, an
increased appetite for dollars on the part of foreign central banks is typi-
cally perceived by market participants as positive for the dollar in the short
run, and the model supports this intuition. However, the authors show that,
because the short-term appreciation of the dollar may delay necessary
adjustment, in the long run the result of an increased preference for dol-
lars may be more rather than less dollar depreciation. Thus developments
 52 Brookings Papers on Economic Activity, 1:2005

that are “good news” for the currency in the short run may be “bad news” at
a longer horizon.
One point that I take from the paper’s analysis, however, is that the par-
ticular assumptions one makes about the nature of the steady state may sig-
nificantly affect one’s predictions about short-run dynamics and the speed
of adjustment. For example, the authors assume in most of their analysis
that, in the long run, the U.S. current account must return to balance. One
might reasonably assume instead that, in the long run, the current account
will remain in deficit at levels consistent with long-run stability in the ratio
of external debt to GDP. This apparently innocuous change in the steady-
state assumption may have quantitatively important implications for the
medium-term pace of adjustment. In particular, to the extent that foreigners
are willing to accept a long-run U.S. debt-to-GDP ratio that is somewhat
higher than the current level of about 25 percent, the authors’ model predicts
that the period of current account adjustment could be extended for a num-
ber of years. Because we know little about the quantity of U.S. assets that
foreigners may be willing to hold in the long run, the model suggests that
one cannot forecast the speed of the adjustment process with any confidence.
Although the authors’ model is extraordinarily useful, like any simple
model it leaves out important factors. From my perspective, the model’s
most important omissions are related to its treatment of asset values and
interest rates. Except for the exchange rate itself, the model takes asset val-
ues and interest rates as exogenous, thereby excluding what surely must
be an important source of current account dynamics, namely, the endoge-
nous evolution of wealth and expected returns. For example, I doubt that
the recent decline in U.S. household saving, a major factor (arithmetically
at least) in the rise in the U.S. current account deficit, can reasonably be
treated as exogenous, as is done in the paper. Instead, at least some part of
the decline in saving likely reflects the substantial capital gains that U.S.
households have enjoyed in the stock market (until 2000, and to some extent
since 2003) and in the values of their homes. Capital gains have allowed
Americans to feel wealthier without saving out of current income.
Where did these capital gains come from? In my view an important driver
of the rise in U.S. wealth is the rapid increase over the past decade or so in
the global supply of saving, which in turn is the product of both the strong
motivation to save on the part of other aging industrial societies and a reluc-
tance of emerging economies to import capital since the financial crises of
the 1990s. Increased global saving has produced a striking decline in real
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 53

interest rates around the world, a decline that has contributed to the increased
valuation of stocks, housing, land, and other assets.2 Because of its open-
ness to foreign capital, its financial sophistication, and its relatively strong
economic performance, the United States has absorbed the lion’s share of
this increment to global saving; however, other industrial countries (includ-
ing France, Italy, Spain, and the United Kingdom) have also experienced
increased asset values (house prices, for the most part), increased consump-
tion, and corresponding movements in their current account balances toward
deficit. An implication of this story is that an endogenous moderation of
the U.S. current account deficit may be in store, even without major changes
in exchange rates and interest rates, as a diminishing pace of capital gains
slows U.S. consumption growth.3 This story, or any explanation that relies
heavily on endogenous changes in asset prices and the ensuing wealth and
spending dynamics, cannot be fully captured by the current version of the
authors’ model.
How might endogenous wealth dynamics change the authors’ conclu-
sions? One way of developing an intuition about the effects of wealth
dynamics in the context of their model is to use that model to consider the
implications for the current account and the dollar of an exogenous change
in the value of U.S. assets, X. Although this approach yields at best a sim-
ple approximation of the effect of making wealth endogenous, examining
model outcomes when one drops the authors’ assumption of unchanging
wealth should provide some insight.
To carry out this exercise, I write the key equations of the model as
follows:
 X* 
(1) F+1 = [1 − α * ( R, s )]  + F  (1 + r )
E
[1 − α ( R, s ) ( X − F )] (
− 1 + r ) + D ( E, X − F )
Rrealized

(1 + r ) E+e1 (1 + r ) E+1
where R = and Rrealized = (
(1 + r *) E 1 + r*)E

2. Bernanke (2005).
3. Recent experience in the United Kingdom shows that a stabilization of house prices
after a period of rapid increases may damp consumer spending and increase saving rates.
 54 Brookings Papers on Economic Activity, 1:2005

 X* 
(2) X = [ α ( R, s )] ( X − F ) + [1 − α * ( R, s )]  + F .
E

I use the authors’ notation, except that I find it useful to distinguish between
the anticipated relative return on U.S. assets, R, and the realized relative
return on U.S. assets, Rrealized. I also suppress the shock terms z and s, which
I will not use here.
Equation 1 is the current account equation, which describes the evolu-
tion of U.S. net foreign debt, F. The first term on the right-hand side of
equation 1 captures the idea that, all else equal, foreign debt grows at the
U.S. real rate of interest. The second term, which I have chosen to write in
a slightly different form than the authors do, is the valuation effect associated
with unanticipated changes in the exchange rate. In particular, when the
value of the dollar is less than expected, Rrealized < R, and the dollar value
of U.S. gross foreign assets rises. This valuation effect serves to reduce
U.S. net dollar liabilities. The third term in equation 1 is the trade deficit,
which adds directly to net foreign liabilities. I extend the authors’ model here
by including U.S. domestic wealth, X − F, as a determinant of the trade
deficit. I assume that the derivative of the trade deficit with respect to U.S.
wealth is positive; higher wealth induces U.S. households to spend more,
increasing the trade deficit.
Equation 2, the portfolio balance equation, is the same as in the paper.
This equation requires that the supply of U.S. assets X equal the sum of
U.S. and foreign demands for those assets.
The steady-state equations corresponding to equations 1 and 2 are

(3) rF = − D ( E, X − F )

(4) X = [ α (1, s )] ( X − F ) + [1 − α * (1, s )] ( X * E + F ) .

Equation 3 is the steady-state version of the current account equation, mod-

ified to allow U.S. wealth to affect the trade balance. Here I retain the authors’
assumption that the current account must be in balance in the long run (as
opposed to assuming a constant ratio of external debt to GDP in the long run).
Equation 4 is the steady-state version of the portfolio balance, exactly as
in the paper. Like the authors, I assume that the foreign real interest rate
equals the domestic rate, so that R = Rrealized = 1 in the steady state.
My figure 1, which is analogous to the figures in the paper, graphs the
steady-state equations 3 and 4. Because foreign debt F is included as a
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 55
Figure 1. Adjustment of the Exchange Rate and the Net Debt Position to an Increase
in U.S. Assets

Exchange rate (E)

Portfolio balance

C
Current account balance

Net debt (F)

Source: Authors’ model described in the text.

determinant of the trade deficit (more foreign debt reduces U.S. wealth and
thus the trade deficit), the current account line in my figure is flatter than its
analogue in the authors’ model, all else equal; under reasonable assump-
tions, however, it is still downward sloping. The portfolio balance line is
the same as in the authors’ analysis.
Consider now the effects of an exogenous increase in X. A first issue is
whether this increase is expected to be temporary or permanent. If con-
sumers have a target wealth-to-income ratio, which is not an unreasonable
supposition, the increase in X might be thought of as largely transitory.
In this case it is straightforward to show that the steady state will be
unaffected by the increase in U.S. assets, so that the current account and
the exchange rate will return to their original values in the long run; that is,
although it would imply a short-run depreciation, a temporary increase in
 56 Brookings Papers on Economic Activity, 1:2005

the value of U.S. assets would have no lasting effect on the dollar or the
U.S. net international position. Since this case, although possibly relevant,
is not very interesting, I consider instead the case in which the increase in
the value of U.S. assets is expected to be permanent.
Figure 1 shows the graphical analysis of a permanent increase in U.S.
assets. I assume that the economy is initially in the steady state defined by
point A. Inspection of equation 3 shows that an increase in X shifts the
current account line down, as greater U.S. wealth worsens the steady-state
trade balance at any given exchange rate. Conceptually, this downward shift
is analogous to the effect of an exogenous increase in the U.S. demand for
foreign goods, as analyzed by the authors. Absent any change in the port-
folio balance condition, this shift would imply both dollar depreciation and
increased foreign debt in the long run, exactly as in the paper’s analysis of
an exogenous shift in demand.
However, the portfolio balance line is not unchanged in my scenario but
instead is shifted downward by the increase in X, as foreigners are willing to
hold their share of the increase in U.S. assets only if the dollar depreciates.
(The depreciation implies an unanticipated reduction in the dollar share of
foreigners’ portfolios, for which they are assumed to compensate by buy-
ing additional dollar assets.) With the shifts in both the current account and
the portfolio balance relations taken into account, the new steady-state
position is shown as point C in figure 1. As indicated, and under plausible
assumptions, the economy adjusts by jumping immediately from point A
to point B, as the dollar depreciates and U.S. net foreign debt declines.
Over time the economy moves from point B to point C, as the dollar
depreciates further and foreign debt accumulates.
A key point is that, all else equal, the steady-state outcome described by
point C involves less dollar depreciation and less accumulation of foreign
debt than the scenario (analyzed by the authors) in which U.S. demand for
foreign goods increases exogenously (that is, a scenario in which only the
current account line shifts down). Economically, the unexpected depreci-
ation induced by the requirement of portfolio balance assists the U.S. cur-
rent account adjustment process in two ways: First, the depreciation reduces
the initial dollar value of U.S. net foreign debt directly, by means of the
valuation effect. Second, the early depreciation of the dollar associated
with the portfolio balance requirement mitigates the trade impact of the
rise in wealth. Note also that U.S. domestic wealth (that is, net of foreign
liabilities) is very likely to be higher in the long run than initially, reflecting
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 57

the capital gains enjoyed at the beginning of the process. This analysis
is overly simple, as already noted, but it suggests to me that inclusion of
endogenous wealth dynamics might give different and possibly less wor-
risome predictions about U.S. current account adjustment than those pre-
sented in the paper.
My final observations bear on the authors’ analysis of the case with more
than two currencies. I found this part of the paper quite enlightening, par-
ticularly the discussion of the likely effects of a revaluation of the Chinese
currency on the value of the euro. One occasionally hears the view expressed
that yuan revaluation would “take the pressure off” the euro (that is, allow
it to depreciate); the underlying intuition appears to be that the effective
dollar exchange rate must fall by a certain amount, and so, if it cannot fall
against the yuan, it will fall against the euro. The authors show that this intu-
ition is likely misguided, in that a stronger yuan probably implies a stronger
euro as well. Their argument can be understood either in terms of portfolio
balance or in terms of trade balance. From a portfolio perspective, a yuan
revaluation presumably would shift Chinese demand away from dollar assets
and toward euro assets, strengthening the exchange value of the euro. From
a trade perspective, if Chinese goods become more expensive for Ameri-
cans, U.S. demand may shift toward euro-zone goods, again implying euro
appreciation.
I see much merit in this analysis but would note that these results may not
generalize to cases with many countries and variable patterns of substitution
and complementarity among goods and among currencies. To illustrate, sup-
pose that Chinese goods and European goods are viewed as complements
by potential buyers in other nations. Then, in the same way that a rise in
the price of teacups lowers the price of saucers, a Chinese revaluation might
reduce the global demand for European exports to an extent sufficient to
cause the euro to depreciate. This example is probably not realistic (others
could be given), but it shows that drawing general conclusions about how
changes in the value of one currency affect that of another may be difficult.
Even if a revaluation of the yuan did lead to an appreciation of the euro,
however, one should not conclude that yuan revaluation is against the
European interest. A yuan revaluation might well lead to both an increase
in the demand for European exports (as U.S. demand is diverted from
China) and a reduction in European interest rates (reflecting increased
Chinese demand for euro assets). Yuan revaluation might therefore stim-
ulate the European economy even though the euro appreciates.
 58 Brookings Papers on Economic Activity, 1:2005

Hélène Rey: Olivier Blanchard, Francesco Giavazzi, and Filipa Sa have

given us a very clear and elegant framework within which to discuss some
complex and important questions. The U.S. current account deficit has been
at the center of the economic policy debate for some time. The deficit stood
at more than 6 percent of GDP in 2004, and in dollar terms it has reached
historically unprecedented levels.
A country can eliminate an external imbalance either by running trade
surpluses, or by earning favorable returns on its net foreign asset port-
folio, or both. The first of these, the trade channel of adjustment, has been
traditionally emphasized in studies of current account sustainability. The
valuation channel has received attention only lately, but with the recent
upsurge in cross-border asset holdings, its quantitative significance has
greatly increased. When the securities in which external assets and liabil-
ities are held are imperfectly substitutable, any change in asset prices and,
in particular, any change in the exchange rate create international wealth
transfers, which can be sizable. These transfers significantly alter the
dynamics of net foreign assets.
The following example illustrates the power of the valuation channel
to smooth the U.S. adjustment process. Following Cédric Tille,1 assume
that U.S. external liabilities, which amounted to about $10.5 trillion in
December 2003, are all denominated in dollars, whereas 70 percent of the
$7.9 trillion in U.S. external assets are in foreign currency. Then a mere
10 percent depreciation of the dollar, by increasing the dollar value of the
foreign-currency assets while leaving the dollar value of the liabilities
constant, would create a wealth transfer from the rest of the world to the
United States equal to 0.1 × 0.7 × 7 trillion, or about $553 billion, which
is approximately 5 percent of U.S. GDP and on the order of the U.S. cur-
rent account deficit in 2003. The exchange rate thus has a dual stabilizing
role for the United States. A dollar depreciation helps improve the trade
balance and increases the net foreign asset position, and this has to be
taken into account when assessing the prospects of the U.S. external
deficit and the future path of the dollar.
The authors have set out to do just that. They use a portfolio balance
model (drawing on the work of Pentti Kouri, Stanley Black, Dale Henderson
and Kenneth Rogoff, and William Branson in the 1980s) to model jointly
the dynamics of the current account and of the exchange rate, allowing for

1. Tille (2003).
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 59

imperfect substitutability between assets and for (some) valuation effects.

In such a framework, a negative shock to preferences for U.S. goods, say,
leads immediately to a depreciation of the dollar. This immediate, unexpected
depreciation does not, however, fully offset the shock. If it did, there would
be excess demand for U.S. assets, as the supply of those assets is taken to
be fixed and the dollar value of the rest of the world’s wealth rises. Instead
there is a less than fully offsetting drop in the dollar, and foreigners’ demand
for U.S. assets is kept in check by a further, expected depreciation of the
dollar toward its long-run steady-state value. Along the path of this depre-
ciation, the United States accumulates more debt, so that the long-run level
of the dollar will be below that which would have been needed to offset the
entire negative shock immediately. The dollar is expected to depreciate at
a decelerating rate in order for foreigners to keep accumulating U.S. assets.
A remarkable prediction thus emerges from this simple model: foreigners
continue to purchase U.S. assets and finance the U.S. current account deficit
even though they expect a further dollar depreciation, which implies capital
losses on their portfolio.
This result stems entirely from the imperfect degree of substitutabil-
ity between U.S. and foreign assets. If assets were perfect substitutes, the
exchange rate would jump immediately to the steady-state level that would
be compatible with the change in preferences for goods. Pierre-Olivier
Gourinchas and I present strong evidence that assets are imperfect substi-
tutes.2 We find that current external imbalances have substantial predictive
power on net asset portfolio returns and, in particular, on exchange rates.
Using a newly constructed database on U.S. external imbalances since 1952,
we show that negative external imbalances imply future expected depreci-
ations of the dollar. We find that a 1-standard-deviation increase in the
imbalance leads to an expected annualized depreciation of around 4 percent
over the next quarter. These empirical results are fully supportive of the port-
folio balance approach and of Blanchard, Giavazzi, and Sa’s model. We
also find, however, that the trade channel of adjustment kicks in at longer
horizons, so that the valuation effects operate in the short to medium run
whereas the trade balance effects operate in the longer run. In the authors’
model, in contrast, valuation and trade channels operate contemporaneously.
There is no lag in the adjustment dynamics of the trade flows. If there
were, the dynamics of the debt accumulation would be different. But I think

2. Gourinchas and Rey (2005).

 60 Brookings Papers on Economic Activity, 1:2005

it is reasonable to conjecture that this would not change any of the quali-
tative results of the paper.
A more important point is that the authors model rates of return using
(exogenous) interest rates only. In reality, U.S. assets and liabilities include
both equity and debt and indeed have a very asymmetric composition. The
external assets of the United States consist mainly of foreign direct invest-
ment and equities, whereas U.S. external liabilities contain a larger share of
bank loans and other debt. As a consequence, the returns on U.S. external
assets and liabilities differ substantially. The United States, as the world’s
banker, has traditionally enjoyed higher returns on its assets, which are dom-
inated by long-term risky investments, than it has had to pay on its mostly
liquid liabilities. (This explains in part why the income on U.S. net foreign
assets is still positive even though the United States’ liabilities exceed its
assets by about 30 percent.) Hence the net foreign asset dynamic is highly
dependent on differences in relative returns on portfolio equity, FDI, and so
forth, and is mischaracterized if one considers only the risk-free interest rate.
The authors’ framework also ignores the joint determination of exchange
rates, bond prices, and equity returns on asset markets. A more complete
model would feature endogenous valuation effects on the stock of assets
and liabilities, both in the current account equation and in the portfolio
balance equation. This also means that the steady-state condition of the
authors’ model, which equates the interest to be paid on the U.S. net for-
eign debt to the trade balance, may be significantly altered when one takes
into account the composition of the net debt. If it is dominated by contin-
gent claims such as equities, the equilibrium steady-state exchange rate
necessary to generate the required trade balance may differ considerably
from what their model assumes. The exogeneity of the rate of return (the
interest rate) is a clear limitation. In principle, the interest rate should be
determined by the reaction of the Federal Reserve and by endogenous
changes in world supply and demand for capital. Proponents of the “global
savings glut” theory see no mystery in persistently low long-term U.S.
interest rates. As it stands, the model has nothing to say on these issues.
The authors make a very natural extension of their model to a three-
country setting, and they demonstrate that putting pressure on China to intro-
duce more flexibility in its exchange rate regime would be counterproductive
if the objective is a less depreciated dollar against the euro. Indeed, by
forcing China out of the business of buying dollars, one effectively bans
from the market the agent with the stronger bias for dollars. Since the cur-
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 61

rency demand of the other agents is more diversified, this decreases the
demand for dollars and increases the pressure on the euro to appreciate.
Hence, at least in the short run, the dynamic is perverse. I think this is an
excellent insight that should be discussed in policy circles.
One of the messages of this very rich paper is that, as then-U.S. Treasury
secretary John Connolly put it in the 1970s, “the dollar is our currency but
your problem.” Indeed, the paper makes a very strong case that, to return
to the steady state after a negative shock to the U.S. current account, one
needs the dollar to depreciate in a predictable way at a moderate speed for
a long period. Along the adjustment path, foreign investors incur capital
losses as wealth is transferred to the United States. The adjustment is smooth
and relatively painless for the U.S. economy, but the rest of the world suf-
fers not only the capital loss but also a loss in competitiveness for the export
sector (but increased purchasing power). I have two comments on this point.
The first is that, within the model, the speed of the predicted depreciation
of the dollar can be computed only with considerable uncertainty. It depends
on several difficult-to-measure quantities such as world wealth, the degree
of home bias in U.S. and foreign portfolios, and the future change in that
bias. So it would not be surprising if the speed of depreciation turned out
to be quicker than the upper bound of 2.7 percent a year (or even 8.4 percent
a year) predicted by the authors. We just do not know.
My second comment is that the assumptions implicit in these results are
that bond prices are exogenous and that no run on dollar assets occurs. In the
authors’ model, whatever happens to the exchange rate does not affect the
U.S. interest rate. That is surely too extreme an assumption. Without making
any predictions, I would like to suggest that a less rosy scenario be put on
the table as well, in which turmoil occurs in both the bond and the foreign
exchange markets simultaneously. One can imagine that some Asian central
banks that are at least partly accountable to the citizens of their countries
(such as the Korean central bank) might start diversifying out of dollar assets
in order to decrease their exposure to exchange rate risk. To the extent that
such a move creates jitters in financial markets and private investors follow
suit, the U.S. interest rate could go up at the same time that the dollar is
going down, which could lead to a further unwinding of positions. We had
a small taste of such an event in early 2005, when the Korean central bank
announced that it would diversify its future accumulations of reserves (that
is, its flows, not even its stocks) out of the dollar, and U.S. interest rates rose
sharply for a short period. This scenario could be particularly damaging if
 62 Brookings Papers on Economic Activity, 1:2005

Asian central banks were to dump ten-year U.S. Treasuries, which consti-
tute the backbone of the U.S. mortgage market. Since we do not have precise
information on the maturity structure of the debt held by the Asian central
banks, or precise estimates of the degree of substitutability between the ten-
year bond and bonds at the short end of the yield curve, such a scenario
would be sure to be full of surprises. In the end much would depend on the
willingness of the Federal Reserve to tighten monetary policy aggressively.
If U.S. interest rates jumped sharply, the whole world economy could be in
for a hard landing.
To conclude, this paper is a remarkable achievement, and I am sure it will
prove to be an invaluable pedagogical tool. After almost three decades dur-
ing which the portfolio balance approach was largely neglected, this paper
and some other recent work point toward its renewed relevance. The authors
provide a perfect example of how powerful it can be to gain clear insights
on the very complex questions posed by the dynamics of the U.S. current
account deficit and the dollar. The next, very important step in this line of
research is to develop a more convincing model of asset prices and wealth
dynamics. Until we endogenize international portfolio flows in different
assets, the wealth dynamics, and the joint determination of the exchange
rate, equity prices, and interest rates, we will not be able to fully compre-
hend the nature of the international adjustment process and will have to
shy away from specific policy recommendations.

General discussion: Gian Maria Milesi-Ferretti observed that foreigners

own relatively little of U.S. housing wealth. As a consequence, any fall in
home prices due to rising interest rates would have a relatively small valua-
tion effect on foreign wealth, and therefore little effect on foreign demand for
U.S. assets. By the same token, it would have a relatively large effect on U.S.
wealth, saving, and the current account. He also pointed out that the large
increase in world saving over the past decade has come mainly from China,
where both saving and investment have risen spectacularly. Outside of China
saving rates have mostly declined. Indeed, the rise in current account sur-
pluses in other East Asian economies reflects a sharp decline in domestic
investment rather than an increase in saving. Sebastian Edwards added that
every region in the world outside North America, including Africa, has a cur-
rent account surplus, and most emerging economies are purchasing U.S.
assets. He reasoned that it will be difficult for these countries to grow rapidly
if their saving continues to go abroad rather than into domestic investment.
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 63

Richard Cooper suggested that alternative assumptions about the char-

acter of financial markets and the distribution of world saving could alter
some of the authors’ model results, quantitatively and possibly qualitatively.
For example: Saving in the rest of the world is roughly three times U.S.
saving, but more than half of the world’s easily marketable assets are located
in the United States, making it the preferred destination for foreign invest-
ment. Even with home bias, as long as rest-of-world wealth is growing
faster than U.S. wealth, net investment flows into U.S. assets are likely to
continue, and with them U.S. current account deficits. William Nordhaus
remarked that the situation Cooper described is changing: as Europe opens
its capital markets, the large U.S. share of the world’s marketable assets
should gradually fall.
Michael Dooley argued that Cooper’s analysis, and the imperfect substi-
tutability in the authors’ portfolio model, did not capture the growing risk
that private agents would perceive as U.S. net indebtedness continues to
grow. A counter to this constraint on private asset demand is provided when
foreign official sectors invest in U.S. assets the way several Asian central
banks are doing today. Peter Garber added that central banks of emerging
economies are readily buying these assets because they provide the collat-
eral that encourages outside investors to undertake gross investment flows
into these economies. He believed the exchange rate movements of the past
few years were mainly due to these official interventions, which underwrite
the U.S. capital market at low interest rates. At these low rates, private sec-
tor investors have shifted their demand toward European securities, causing
the euro to strengthen and reducing Europe’s current account surplus.
Edmund Phelps explained that his own model projected a much lower
dollar and a shift to U.S. current account surpluses, and he addressed the
macroeconomic implications for the United States of such a move. He dis-
agreed with the more optimistic experts who see such a transition as not
affecting aggregate output and employment in any important way. On that
scenario, the investment decline that accompanies lower business asset values
in his model would be smoothly offset by rising exports and a move toward
current account surpluses. Phelps, however, believed that the needed shift
in resources would be incomplete to the extent that the investment-type
activities are relatively labor intensive in production. He thus expected the
needed adjustment to have a significant macroeconomic impact, and he saw
the U.S. economy heading into a decade or more of slower growth and
weakening employment.
 64 Brookings Papers on Economic Activity, 1:2005

References
Baldwin, Richard E., and Paul R. Krugman. 1987. “The Persistence of the U.S.
Trade Deficit.” BPEA, no. 1: 1–43.
Bernanke, Ben S. 2005. “The Global Saving Glut and the U.S. Current Account
Deficit.” Homer Jones Lecture, Federal Reserve Bank of St. Louis, April 14.
Available on the Internet at www.federalreserve.gov/boarddocs/speeches/2005/
20050414/default.htm.
Bernanke, Ben S., Vincent R. Reinhart, and Brian P. Sack. 2004. “Monetary Policy
Alternatives at the Zero Bound: An Empirical Assessment.” BPEA, no. 2: 1–78.
Branson, William H., and Dale W. Henderson. 1985. “The Specification and Influ-
ence of Asset Markets.” In Handbook of International Economics II, edited by
Ronald W. Jones and Peter B. Kenen. Amsterdam: Elsevier Science Publishers.
Caballero, Ricardo J., Emmanuel Farhi, and Mohamad L. Hammour. 2004. “Spec-
ulative Growth: Hints from the U.S. Economy.” Massachusetts Institute of
Technology.
Chinn, Menzie D. 2004. “Incomes, Exchange Rates and the U.S. Trade Deficit,
Once Again.” International Finance 7, no. 3: 451–69.
Cooper, Richard N. 1986. “Dealing with the Trade Deficit in a Floating Rate
System.” BPEA, no. 1: 195–207.
Debelle, Guy, and Gabriele Galati. 2005. “Current Account Adjustment and Capital
Flows.” BIS Working Paper 169. Basel: Monetary and Economic Department,
Bank for International Settlements.
Dooley, Michael P., David Folkerts-Landau, and Peter M. Garber. 2004. “The U.S.
Current Account Deficit and Economic Development: Collateral for a Total
Return Swap.” Working Paper 10727. Cambridge, Mass.: National Bureau of
Economic Research.
Dornbusch, Rudiger. 1976. “Expectations and Exchange Rate Dynamics.” Journal
of Political Economy 84: 1161–76.
_________. 1987. “External Balance Correction: Depreciation or Protection?”
BPEA, no. 1: 249–69.
Gourinchas, Pierre-Olivier, and Hélène Rey. 2005. “International Financial Adjust-
ment.” Working Paper 11155. Cambridge, Mass.: National Bureau of Economic
Research (February).
Henderson, Dale, and Kenneth Rogoff. 1982. “Negative Net Foreign Asset Positions
and Stability in a World Portfolio Balance Model.” Journal of International
Economics 13: 85–104.
Houthakker, Hendrik S., and Stephen P. Magee. 1969. “Income and Price Elasticities
in World Trade.” Review of Economics and Statistics 51, no. 2: 111–25.
Kouri, Pentti. 1976. “Capital Flows and the Dynamics of the Exchange Rate.”
Seminar Paper 67. Stockholm: Institute for International Economic Studies.
_________. 1983. “Balance of Payments and the Foreign Exchange Market: A
Dynamic Partial Equilibrium Model.” In Economic Interdependence and Flexible
Olivier Blanchard, Francesco Giavazzi, and Filipa Sa 65

Exchange Rates, edited by Jagdeep S. Bhandari and Bulford H. Putnam. MIT

Press.
Krugman, Paul R. 1991. “Introduction.” In International Adjustment and Financing,
edited by C. Fred Bergsten and Paul R. Krugman. Washington: Institute for
International Economics.
Lane, Philip R., and Gian Maria Milesi-Ferretti. 2002. “Long-Term Capital Move-
ments.” In NBER Macroeconomics Annual 2001, edited by Ben S. Bernanke
and Kenneth S. Rogoff. MIT Press.
_________. 2004. “Financial Globalization and Exchange Rates.” CEPR Discussion
Paper 4745. London: Centre for Economic Policy Research.
Lawrence, Robert Z. 1990. “U.S. Current Account Adjustment: An Appraisal.”
BPEA, no. 2: 343–82.
Marquez, Jaime. 2000. “The Puzzling Income Elasticity of U.S. Imports.” Wash-
ington: Federal Reserve Board.
Masson, Paul. 1981. “Dynamic Stability of Portfolio Balance Models of the
Exchange Rate.” Journal of International Economics 11: 467–77.
Obstfeld, Maurice. 2004. “External Adjustment.” Review of World Economics 140,
no. 4: 541–68.
Obstfeld, Maurice, and Kenneth Rogoff. 2004. “The Unsustainable U.S. Current
Account Position Revisited.” Working Paper 10869. Cambridge, Mass.: National
Bureau of Economic Research.
Roubini, Nouriel, and Brad Setser. 2005. “Will the Bretton Woods 2 Regime
Unravel Soon? The Risk of a Hard Landing in 2005–2006.” Paper presented at
a symposium on the “Revived Bretton Woods System: A New Paradigm for
Asian Development?” organized by the Federal Reserve Bank of San Francisco
and the University of California, Berkeley, San Francisco, February 4, 2005.
Sachs, Jeffrey D. 1988. “Global Adjustments to a Shrinking U.S. Trade Deficit.”
BPEA, no. 2: 639–67.
Tille, Cédric. 2003. “The Impact of Exchange Rate Movements on U.S. Foreign
Debt.” Issues in Economics and Finance 9, no. 1: 1–7.
MAURICE OBSTFELD
University of California, Berkeley
KENNETH S. ROGOFF
Harvard University

Global Current Account Imbalances

and Exchange Rate Adjustments

THIS IS THE third in a series of papers we have written over the past five years
about the growing U.S. current account deficit and the potentially sharp ex-
change rate movements any future adjustment toward current account bal-
ance might imply.1 The problem has hardly gone away in those five years.
Indeed, the U.S. current account deficit today is running at around 6 per-
cent of GDP, an all-time record. Incredibly, the U.S. deficit now soaks up
about 75 percent of the combined current account surpluses of Germany,
Japan, China, and all the world’s other surplus countries.2 To balance its
current account simply through higher exports, the United States would
have to increase export revenue by a staggering 58 percent over 2004 levels.
And, as we argue in this paper, the speed at which the U.S. current account
ultimately returns toward balance, the triggers that drive that adjustment,
and the way in which the burden of adjustment is allocated across Europe

Eyal Dvir, José Antonio Rodriguez-Lopez, and Jón Steinsson provided dedicated and
excellent research assistance, for which we are extremely grateful. We also thank Philip
Lane and Gian Maria Milesi-Ferretti for discussions and data. Jane Trahan’s technical sup-
port was outstanding.
1. See Obstfeld and Rogoff (2000a, 2004). From an accounting perspective, a country’s
current account balance essentially adds net interest and dividend payments to its trade bal-
ance. As we discuss below, the United States presently receives about the same amount of in-
come on its foreign assets as it pays out to foreign creditors. Hence, for the United States (and
indeed many countries), the current account balance and the trade balance are quantitatively
very similar. As we later emphasize, however, the current account does not include capital
gains and losses on existing wealth. Thus the overall change in a country’s net foreign asset
position can, in principle, be less than or greater than its current account deficit or surplus.
2. Calculated from the World Economic Outlook database of the International Monetary
Fund, using current account data from 2004.

67
 68 Brookings Papers on Economic Activity, 1:2005

and Asia all have enormous implications for global exchange rates. Each
scenario for returning to balance poses, in turn, its own risks to financial
markets and to general economic stability.
Our assessment is that the risks of collateral damage—beyond the risks
to exchange rate stability—have grown substantially over the five years
since our first research paper on the topic, partly because the U.S. current
account deficit itself has grown, but mainly because of a mix of other fac-
tors. These include, not least, the stunningly low U.S. personal saving rate
(which, driven by unsustainable rates of housing appreciation and record
low interest rates, fell to 1 percent of disposable personal income in 2004).
But additional major risks are posed by the sharp deterioration in the U.S.
federal government’s fiscal trajectory since 2000, rising energy prices, and
the fact that the United States has become increasingly dependent on Asian
central banks and politically unstable oil producers to finance its deficits.
To these vulnerabilities must be added Europe’s conspicuously inflexible
economy, Japan’s continuing dependence on export-driven growth, the sus-
ceptibility of emerging markets to any kind of global financial volatility,
and the fact that, increasingly, the counterparties in international asset
transactions are insurance companies, hedge funds, and other relatively
unregulated nonbank financial entities. Perhaps above all, geopolitical
risks and the threat of international terror have risen markedly since Sep-
tember 2001, confronting the United States with open-ended long-term
costs for financing wars and homeland security.
True, if some shock (such as a rise in foreign demand for U.S. exports)
were to close up these global imbalances quickly without exposing any con-
comitant weaknesses, the damage might well be contained to exchange rates
and to the collapse of a few large banks and financial firms—along with, per-
haps, mild recession in Europe and Japan. But, given the broader risks, it
seems prudent to try to find policies that will gradually reduce global imbal-
ances now rather than later. Such policies would include finding ways to
reverse the decline in U.S. saving, particularly by developing a more credi-
ble strategy to eliminate the structural federal budget deficit and to tackle the
country’s actuarially insolvent old-age pension and medical benefit pro-
grams. More rapid productivity growth in the rest of the world would be par-
ticularly helpful in achieving a benign adjustment, but only, as the model we
develop in this paper illustrates, if that growth is concentrated in nontraded
(domestically produced and consumed) goods rather than the export sector,
where such productivity growth could actually widen the U.S. trade deficit.
Maurice Obstfeld and Kenneth S. Rogoff 69

It is also essential that Asia, which now accounts for more than one-third
of global output on a purchasing power parity basis, take responsibility
for bearing its share of the burden of adjustment. Otherwise, if demand
shifts caused the U.S. current account deficit to close even by half (from
6 percent to 3 percent of GDP), while Asian currencies remain fixed against
the dollar, we find that European currencies would have to depreciate by
roughly 29 percent. Not only would Europe potentially suffer a severe
decline in export demand in that scenario; it would also incur huge losses
on its net foreign asset position: Europe would lose about $1 trillion if the
U.S. current account deficit were halved, and twice that sum if it went to
zero.
We do not regard our perspective as particularly alarmist. Nouriel
Roubini and Brad Setser make the case that the situation is far grimmer
than we suggest, with global interest rates set to skyrocket as the dollar loses
its status as the premier reserve currency.3 Olivier Blanchard, Francesco
Giavazzi, and Filipa Sa present an elegant and thoughtful analysis sug-
gesting that prospective dollar exchange rate changes are even larger than
those implied by our model.4 William Cline argues that an unsustainable
U.S. fiscal policy has substantially elevated the risk of an adverse scenario.5
In our view, any sober policymaker or financial market analyst ought to
regard the U.S. current account deficit as a sword of Damocles hanging
over the global economy.
Others, however, hold more Panglossian views. One leading benevo-
lent interpretation, variously called the “Bretton Woods II” model or the
“Deutsche Bank” view, focuses on China; that view is forcefully exposited
in this volume by Michael Dooley and Peter Garber. This theory explains
the large U.S. current account deficit as a consequence of the central prob-
lem now facing the Chinese authorities: how to maintain rapid economic
growth so as to soak up surplus labor from the countryside. For China, a
dollar peg (or near peg) helps preserve the international competitiveness

3. Roubini and Setser (2004).

4. Blanchard, Giavazzi, and Sa (this volume).
5. Cline (forthcoming). Mann (2005), although not alarmist, also points to risks in the
adjustment process. Of course, similar discussions accompanied earlier U.S. adjustment
episodes, but the present situation is quite different in both scale and setting. Krugman
(1985, 1991) takes as dim a view as anyone on the sustainability of long-term twin (fiscal
and current account) deficits. His views on the 1980s experience would seem to apply with
even greater force to the current scene.
 70 Brookings Papers on Economic Activity, 1:2005

of exports while attracting foreign direct investment and avoiding stress

on the country’s fragile banking system. Is this argument plausible? Set
aside the fact that China maintained its peg even through the Asian finan-
cial crisis of 1997–98 and as the dollar soared at the end of the 1990s (pre-
sumably making Chinese exports much less competitive), or that China
risks a classic exchange rate crisis if its fortunes ever turn, say, because of
political upheaval in the transition to a more democratic system. The real
weakness in the Bretton Woods II theory is that the Chinese economy is
still less than half the size of Japan’s, and less than three-quarters the size
of Germany’s, at market exchange rates. So, while running surpluses of
similar size to China’s relative to their GDP, Germany and Japan actually
account for a much larger share of global surpluses in absolute terms.
(After all, Germany, not China, is the world’s leading exporter.) And sur-
plus labor is hardly the problem in these aging countries.
U.S. Federal Reserve Chairman Alan Greenspan, in a 2003 speech at
the Cato Institute and in many subsequent speeches, offers an intriguing
argument.6 He agrees that the United States is unlikely to be able to con-
tinue borrowing such massive amounts relative to its income indefinitely,
and he recognizes that the U.S. current account deficit will therefore nar-
row substantially someday. Greenspan argues, however, that increasing
global financial integration is both what allows the United States to run
such large deficits and the saving factor that will greatly cushion the
process of unwinding those deficits.
We completely agree that increasing global financial integration can
explain larger current account deficits, particularly to the extent that greater
trade integration helps underpin financial integration, as in our original
analysis.7 Indeed, this was a major point of our first approaches to this
problem. A narrowing of the U.S. current account deficit must ultimately
be the result, however, of more balanced trade, because the trade account
is overwhelmingly the main component of the current account. And, as
seemingly open as the U.S. economy is to financial flows, international
product markets remain quite imperfectly integrated.
Thus any correction to the trade balance is likely to entail a very large
change in the real effective dollar exchange rate: our baseline figure,
which assumes a moderate speed of adjustment and that the world’s major

6. Greenspan (2004).
7. Obstfeld and Rogoff (2000a, 2000b).
Maurice Obstfeld and Kenneth S. Rogoff 71

regions all return to current account balance, is 33 percent. A much smaller

dollar devaluation is possible only if the adjustment is stretched over a
very long period (say, a decade), in which case labor and capital mobility
across sectors and economies can significantly reduce the need for rela-
tive price changes. On the other hand, should adjustment take place very
abruptly (say, because of a sudden collapse in U.S. housing prices leading
to an increase in saving, or a dramatic reallocation of global central bank
reserves toward the euro), the potential fall in the dollar is much larger
than our baseline estimate of 33 percent, primarily because sticky nomi-
nal prices and incomplete exchange rate pass-through hamper adjustment.
True, in a recent Federal Reserve study, Hilary Croke, Steven Kamin, and
Sylvain Leduc argue that sustained current account imbalances in indus-
trial countries have typically terminated in a relatively benign fashion.8
But their threshold for a current account “reversal”—the country must
have run a deficit of at least 2 percent of GDP for three years, and must
have improved its current account balance by at least 2 percent of GDP
and a third of the total deficit—is a very low bar compared with where the
United States stands today. (Croke, Kamin, and Leduc are forced to choose
a low threshold, of course, because current account deficits of the size,
relative to GDP, of the recent U.S. deficits, although far from unprece-
dented, are not the norm.) Most important, the United States accounts for
over 75 percent of global deficits today, as we have noted, and so any
comparison based on the experience of small countries, even small indus-
trial countries, is of limited value.
In addition to Chairman Greenspan, a number of academic researchers
have emphasized how some important changes in the global financial sys-
tem, particularly over the past ten years, have changed the nature of inter-
national financial adjustment. Philip Lane and Gian Maria Milesi-Ferretti,
in a series of papers, have documented the explosion of gross asset flows.9

8. Croke, Kamin, and Leduc (2005). Freund and Warnock (2005) survey current
account adjustment in industrial countries and find that deficits tend to be associated with
real depreciations, which are larger for consumption-driven deficits.
9. See especially Lane and Milesi-Ferretti (2005a, 2005b). In line with this develop-
ment, Cooper (2001) identifies ongoing international portfolio diversification as a driving
force behind the U.S. deficit. Diversification does not, however, require any net capital
flows: even with a balanced current account, foreigners and U.S. residents can still swap
assets. According to preliminary estimates by the Bureau of Economic Analysis, for exam-
ple, private foreign investors added $1.1 trillion in U.S. assets to their portfolios in 2004,
far more than that year’s U.S. current account deficit of $666 billion.
 72 Brookings Papers on Economic Activity, 1:2005

They and Cédric Tille, as well as Pierre-Olivier Gourinchas and Hélène

Rey, have shown that asset revaluation effects from dollar depreciation
can have a significant impact on U.S. net financial obligations to foreign-
ers.10 Gourinchas and Rey point out, in fact, that the historical extent of
such revaluations suggests that the United States might need to adjust its
trade balance by only two-thirds of the amount that would be needed to
fully repay its net external debt; even this, however, would still imply very
large dollar movements. We agree that the size and composition of gross
asset positions are increasingly important, and our model simulations in
this paper explicitly take account of the revaluation channel. We find,
however, that valuation effects mute the requisite exchange rate changes
only modestly.
The growing financial globalization that these authors and Chairman
Greenspan emphasize is, moreover, a two-edged sword. Enhanced global
financial integration may well facilitate gradual current account and
exchange rate adjustment, but it might also promote the development of
large, unbalanced financial positions that leave the world economy vul-
nerable to financial meltdown in the face of sharp exchange rate swings.
The net foreign asset revaluation channel might help modestly, but a rise
in U.S. interest rates could well wipe out the benefits. Because the United
States borrows heavily in the form of low-risk bonds, while lending heav-
ily in the form of equities and high-risk bonds, it is especially sensitive to
even a modest rise in the interest rates it pays on its foreign debt. Indeed,
we show that, in terms of exchange rate adjustments, the adverse effect of
a 1.25-percentage-point rise in the interest rate that the United States pays
on its short-term foreign debt is similar in magnitude to the benefits gained
via the valuation channel, even with a 20 percent dollar depreciation. More
generally, although increased global financial integration and leverage can
indeed help countries diversify risk, they also expose the system to other
vulnerabilities—such as counterparty risk—on a much larger scale than ever
before. All in all, although we believe that growing financial globalization
is largely a positive development, it does not justify excessive confidence
in a benign adjustment process.
This paper begins by trying to put the recent U.S. experience with current
account imbalances in historical perspective. We hope this first section
will provide a useful reference, although some readers will already be famil-

10. Tille (2004); Gourinchas and Rey (2005a, 2005b).

Maurice Obstfeld and Kenneth S. Rogoff 73

iar with the essential elements. One historical observation that is important
for our later analysis is that the United States (so far) has had the remark-
able ability to consistently pay a lower rate of interest on its liabilities than
it earns on its assets. Some component of this differential in returns has
been due to luck, another to huge central bank holdings of U.S. Treasury
bills, another perhaps to the unique and central role of the dollar in interna-
tional finance. Still another, which we have already emphasized, is the fact
that Americans hold a much larger share of their foreign assets in equities
and high-risk (equity-like) bonds than foreigners hold of U.S. assets (and
thus benefit more from the equity premium). An open question is whether
this advantage can continue in the face of large and persistent U.S. deficits.
We then provide a nontechnical summary of our core three-region
(Asia, Europe, and the United States) model. Readers interested in the tech-
nical details of our model can read the theoretical section that follows, and
the most adventurous can venture into appendix A, where we fully lay out
the structure. Our model simulations calibrate the requisite dollar decline
against European and Asian currencies under various scenarios. Most of
our analysis focuses on real exchange rates, but, by assuming that the
regions’ central banks target GDP or consumption deflators (or sometimes,
in the case of Asia, exchange rates against the dollar), we are able to extract
nominal exchange rate predictions (relative to the initial position) as well.
As noted earlier, our baseline simulation, in which Asia’s, Europe’s, and
the United States’ current accounts all go to zero, implies that the dollar
needs to depreciate in real effective terms by 33 percent (and in nominal
terms by a similar amount). Because the trade balance responds to an
exchange rate change only with a lag, this exercise slightly overstates the
necessary depreciation relative to today’s exchange rates. However, our cal-
ibration assumes flexible prices and does not allow for possible exchange
rate overshooting, which could significantly amplify the effect. A halving of
the U.S. deficit, with counterpart surplus reductions shared by Asia and
Europe in the same proportions as in the first simulation (arguably a more
likely scenario over the short term) of complete current account adjustment,
would lead to a depreciation of the real effective dollar of 17 percent. In our
base case the real value of Asian currencies would need to rise by 35 percent
and that of European currencies by 29 percent against the dollar.
If, however, Asia sticks to its dollar exchange rate peg as the U.S. cur-
rent account deficit narrows, the real effective value of the European cur-
rencies would have to rise by almost 60 percent. Indeed, to maintain its
 74 Brookings Papers on Economic Activity, 1:2005

dollar peg in the face of global demand shifts that fully restore U.S. current
account balance, Asia would actually have to better than double its already
massive current account surplus. Even halving these numbers (correspond-
ing, for example, to the case in which the U.S. current account deficit falls
only by half), one can still appreciate the enormous protectionist pressures
that are likely to emerge if Asia tries to stick to its dollar peg in the face of
a significant pullback in the United States’ voracious borrowing.
It is perhaps surprising that, despite Asia’s current account surplus being
several times that of Europe (which we define broadly here to include the
euro zone and the other largest non-Asian, non-U.S. economies), the
required rise in the Asian currencies relative to the European currencies is
not even larger in the global rebalancing scenario. As we shall see, a cou-
ple of factors drive this result: one is that Asia’s economies are relatively
more open than Europe’s to the rest of the world, so that a given exchange
rate change has a bigger impact on trade; the other is that a large, unantic-
ipated dollar depreciation inflicts brutal damage on Asia’s net foreign
asset position, a factor we explicitly incorporate in our calibrations.
The analysis highlights two important but widely misunderstood points
about the mechanism of U.S. current account deficit reduction. First, real
dollar depreciation is not a substitute for policies that raise U.S. saving,
such as reductions in the federal fiscal deficit. Instead, depreciation and sav-
ing increases are complements: exchange rate changes are needed to bal-
ance goods markets after a change in global consumption patterns, whereas
dollar depreciation that is not accompanied by U.S. expenditure reduction
will lead to inflationary pressures that, over time, will offset the initial gains
in U.S. competitiveness. The second, and related, point is that it makes little
sense to ask how much dollar depreciation is needed to reduce the current
account deficit by 1 percent of GDP. Exchange rates and current account
balances are jointly determined endogenous variables. As the simulations
in this paper illustrate, there are numerous different scenarios in which the
U.S. external deficit might be erased, all with different implications for
the dollar’s foreign exchange value.
Although our model is considerably richer than those previously advanced
in the literature (including our own earlier studies), it remains subject to a
wide range of qualifications and interpretations; we try to emphasize the
most important ones. Nevertheless, we view the simulations as quite use-
ful. The paper’s final section highlights the main conclusions that we draw
from the technical analysis.
Maurice Obstfeld and Kenneth S. Rogoff 75

The U.S. Current Account and Foreign Wealth Position,

1970–2005 and Beyond

The main analytical contribution of the paper is its modeling and numer-
ical calibration of exchange rate and net foreign asset valuation adjustments
under alternative scenarios for reducing the U.S. current account deficit.
Our framework is intended as a tool for assessing risks and evaluating pol-
icy options. At some level, however, the exercise must entail an assessment
of how unstable the current trajectory of external payments imbalances
really is, along with the likelihood of adjustment taking place in the next
few years. In order to think about this overarching issue, it is helpful to
understand the history of the problem.

Perspectives on the U.S. Deficit

Figure 1 traces the U.S. current account balance as a percent of GDP from
1970 to the present. After fluctuating between +1 and −1 percent of GDP
during the 1970s, the current account began to go into deep deficit during
the mid-1980s, reaching 3.4 percent of GDP in 1987. After recovering
temporarily at the end of the 1980s and actually attaining a slight surplus

Figure 1. Current Account, 1970–2005a

Percent of GDP

–1

–2

–3

–4

–5

–6

1972 1976 1980 1984 1988 1992 1996 2000 2004

Source:––Bureau of Economic Analysis, National Income and Product Accounts, International Transactions Accounts.
a.––Data for 2005 are projections.
 76 Brookings Papers on Economic Activity, 1:2005

in 1991 (propped up by a large, one-time transfer from foreign govern-

ments to help pay for the Gulf War), the U.S. current account balance
began a slow, steady deterioration throughout the 1990s, which continues
today. As already noted, U.S. international borrowing in 2004 accounted
for about 75 percent of the excess of national saving over investment of all
the world’s current account surplus countries.
What are the proximate causes of this profound deterioration in the U.S.
external balance? That, of course, is the $666 billion (and rising) question.
Since, in principle, the current account balances of all countries should add
up to zero, the U.S. current account deficit—equal to the excess of U.S.
investment over national saving—has to be viewed as the net result of the
collective investment and saving decisions of the entire world. German
demographics, OPEC oil revenue investment decisions, depressed invest-
ment in Asia—all these factors and many others impinge on global interest
rates and exchange rates and, in turn, on U.S. investment and saving. We
do not believe there is any simple answer.
Nevertheless, U.S. fiscal policy clearly has played a dominant role in
some episodes. The current account balance equals, by definition, the sum
of government saving less investment plus private saving less investment.
Because the Ricardian equivalence of public debt and taxes does not seem
to hold in practice, the big Reagan tax cuts of the 1980s almost certainly
played a role in the U.S. current account deficits of that era. Similarly, the
Bush II tax cuts of the 2000s have likely played a role over the past few
years, preventing the current account deficit from shrinking despite the
post-2000 collapse in U.S. investment. Currency over- and undervaluations
also loomed large in both episodes, usually operating with a lag of one to
two years. For example, the peak of the U.S. current account deficit in 1987
lagged by two years the peak of the real trade-weighted dollar exchange
rate (figure 2). The weak dollar of the mid-1990s was matched by a pause
in the U.S. current account’s decline, and the dollar peak in early 2002
was followed again, with some lag, by a sharp worsening in the external
balance. Admittedly, both correlations with the current account deficit—
of fiscal deficits and dollar appreciation—are fairly loose. As figure 3
illustrates, U.S. fiscal deficits have expanded massively in recent years
compared with those of the rest of the world. But, as the figure also illus-
trates, Japan has run even larger fiscal deficits relative to its GDP than the
United States, yet at the same time it has consistently run the world’s
largest current account surplus in absolute terms.
Maurice Obstfeld and Kenneth S. Rogoff 77

Figure 2. Real Effective Exchange Rate of the Dollar, 1973–2004a

Index, March 1973 = 100

130
120
110
100
90
80
70

1975 1979 1983 1987 1991 1995 1999 2003

Source:–Federal Reserve data.

a.–Broad currency index.

Figure 3. Fiscal Balances in Selected Major Economiesa

Percent of GDP

–2

–4

–6
1999–2001
–8 2002–03
2004–05

Canada United France Germany Euro United Japan

Kingdom area States

Source:–2005 OECD Factbook, OECD Economic Outlook No. 77.

a.–The data are on a national accounts basis, averaged across years indicated.
 78 Brookings Papers on Economic Activity, 1:2005
Figure 4. U.S. Current Account, Saving, and Investment, 1970–2004

Percent of GDP Percent of GDP

Private saving-investment balance (left scale)

6 Government saving-investment balance (left scale)
20
4
Private investment 15
2 (right scale)
10
0
5
–2
0
–4

–6 Current account balance –5

(left scale)

1972 1976 1980 1984 1988 1992 1996 2000 2004

Source:–
–Bureau of Economic Analysis data.

Indeed, during the 1990s the major proximate drivers of the U.S. cur-
rent account balance were a declining rate of private saving and rising rate
of investment. The U.S. personal saving rate, which had been stable at
around 10 percent of disposable personal income until 1985, has steadily
declined since, reaching a mere 1 percent in 2004. The declining private
saving rate has apparently been driven first by the stock price boom of the
1990s and then by the still-ongoing housing price boom.11 Were the U.S.
personal saving rate simply to rise to 5 percent of disposable personal
income, or halfway toward its level of two decades ago, more than half of
today’s current account deficit could be eliminated.
During the late 1990s U.S. investment was robust, as shown in figure 4,
so that the United States’ high external borrowing really was, in principle,
financing future growth. Today, however, the picture has changed. As fig-
ure 4 also shows, the main proximate driver of recent U.S. current account
deficits has been low private and government saving rather than high

11. Obviously, if one measures saving taking into account capital gains and losses on
wealth, the trend decline in saving is much less, although housing wealth is largely not
internationally tradable and both housing and securities wealth can evaporate quickly.
Maurice Obstfeld and Kenneth S. Rogoff 79

investment. So much for the prominent view of former Treasury secretary

Paul O’Neill, who argued that the U.S. external deficit was driven mainly
by foreigners’ desire to invest in productive U.S. assets. The more sophis-
ticated analysis of Jaume Ventura is also inconsistent with declining U.S.
investment.12
Another important factor contributing to the U.S. current account deficit
since the late 1990s has been the persistently low level of investment in
Asia since the region’s 1997–98 financial crisis. Indeed, today, sluggish
investment demand outside the United States, particularly in Europe and
Japan but also in many emerging markets, is a major factor holding global
interest rates down. Low global interest rates, in turn, are a major driver in
home price appreciation, which, particularly in the United States with its
deep, liquid home-equity loan markets, contributes to high consumption.

International Assets, Liabilities, and Returns

Naturally, this sustained string of current account deficits has led to a
deterioration in the United States’ net foreign asset position, as illustrated
in figure 5. In 1982 the United States held net foreign assets equal to just
over 7 percent of GDP, whereas now the country has a net foreign debt
amounting to about 25 percent of GDP. Accompanying this growth in net
debt has been a stunning increase in gross international asset and liability
positions, as figure 5 also shows. From 29.5 percent and 22.3 percent of
GDP in 1982, U.S. gross foreign assets and liabilities, respectively, had
risen to 71.5 percent and 95.6 percent of GDP by the end of 2003. This
process of increasing international leverage—borrowing abroad in order
to invest abroad—characterizes other industrial country portfolios and is
in fact much further advanced for some smaller countries such as the
Netherlands and primary financial hubs such as the United Kingdom; see
table 1 for some illustrative comparative data.13
The implications of the reduction in U.S. net foreign wealth would be
darker but for the fact that the United States has long enjoyed much better

12. Ventura (2001).

13. See also Lane and Milesi-Ferretti (2005a, 2005b) and Obstfeld (2004). The BEA
applies market valuation to foreign direct investment holdings starting only in 1982. Gour-
inchas and Rey (2005b) construct U.S. international position data going back to 1952. In
1976, with foreign direct investment valued at current cost rather than at market value, U.S.
gross foreign assets amounted to 25 percent of GDP, and gross foreign liabilities were 12.6
percent of GDP.
 80 Brookings Papers on Economic Activity, 1:2005
Figure 5. U.S. Foreign Assets and Liabilities, 1982–2003

Percent of GDP

80
Liabilities
60 Assets

40

20

Net foreign assets

–20

1984 1986 1988 1990 1992 1994 1996 1998 2000 2002

Source:–
–Bureau of Economic Analysis data.

investment performance on its foreign assets than have foreign residents

on their U.S. assets. This rate-of-return advantage, coupled with the
expansion in foreign leverage documented in figure 5, has so far allowed
the United States to maintain a generally positive balance of net interna-
tional investment income even as its net international investment position
has become increasingly negative. Figure 6 shows two measures of U.S.

Table 1. International Investment Positions of Selected Industrial Countries, 2003

a
Percent of GDP
Country Gross assets Gross liabilities Net position
Canada 75 93 −18
Euro area 107 118 −10
France 179 172 7
Germany 148 141 6
Italy 95 100 −5
Japan 87 48 39
Switzerland 503 367 135
United Kingdom 326 329 −2
Source: International Monetary Fund, International Financial Statistics.
a. Gross assets may differ from the sum of gross liabilities and the net position because of rounding.
Maurice Obstfeld and Kenneth S. Rogoff 81
Figure 6. U.S. Net Foreign Investment Income and Total Net Return on Foreign
Assets, 1983–2003

Billions of dollars

500 Total net return on

foreign assets
400

300

200

100

–100
Net foreign
–200 investment income

–300

1985 1987 1989 1991 1993 1995 1997 1999 2001

Source: Bureau of Economic Analysis, National Income and Product Accounts, International Transactions Accounts.

net international investment income.14 The first, net foreign investment

income (income receipts on U.S. assets owned abroad less income pay-
ments on foreign-owned assets in the United States), is taken from the
U.S. balance of payments accounts and comprises transactions data only,
that is, actual income earned on assets. Interestingly, this balance has not
yet entered negative territory, although it could do so soon. Over 1983–
2003 the income return on U.S.-owned assets exceeded that on U.S. lia-
bilities by 1.2 percentage points a year on average.
A more comprehensive investment income measure adds the capital gains
on foreign assets and liabilities, reflecting price changes that could be due to
either asset price movements (such as stock price changes) or exchange rate
changes. The Bureau of Economic Analysis (BEA) incorporates estimates

14. Gourinchas and Rey (2005b) present a similar graph covering a much longer
period. The estimates in the text are consistent with those found by Obstfeld and Taylor
(forthcoming) using a different methodology. For a complementary discussion of returns
on foreign assets and liabilities, see Lane and Milesi-Ferretti (2005b).
 82 Brookings Papers on Economic Activity, 1:2005

of these gains into its updates of the U.S. international investment position,
although they do not appear in the international transactions or national
income accounts. As one would expect, figure 6 shows this net income
measure to be much more volatile than that based on investment income
alone. Although it is negative in some years, cumulatively this balance is
even more favorable for the United States than the smoother transactions
measure. On average over 1983–2003, the total return on the United
States’ foreign investment, inclusive of capital gains, exceeded that on U.S.
liabilities to foreigners by a remarkable 3.1 percentage points a year.15
To understand better the implications of the U.S. rate-of-return advan-
tage, let rW be the rate of return on foreign assets, rU the rate of return on
liabilities, F the stock of net foreign assets, and L gross liabilities. Then
the net total return on the international portfolio is rW F + (rW − rU)L. This
expression shows that, even when F < 0 as it is for the United States, total
investment inflows can still easily be positive when rW > rU and the stock
of gross liabilities is sufficiently large. The expression also reveals, how-
ever, that the leveraging mechanism generating the U.S. surplus on
investment returns also heightens the risk associated with a possible
reversal. An unresolved but critical question is whether the United States’
favorable position in international markets will be sustained in the face of
a large and growing external debt. Should the United States at some point
be forced to pay a higher rate on its liabilities, the negative income effect
will be proportional to the extent of leverage, L.
Part of the historical U.S. international investment advantage is a matter
of chance and circumstance. Japanese investors famously bought trophy
properties like Pebble Beach golf club, Rockefeller Center, and Columbia
Pictures at premium prices, only to see those investments sour. Europeans
poured money into the U.S. stock market only at the end of the 1990s, just
as the technology bubble was about to burst. However, a deeper reason why
the United States’ net debt position has accumulated only relatively slowly
is that Americans hold a considerably larger fraction of their foreign assets

15. The broad rate-of-return measures for gross assets and liabilities are constructed by
adding to the investment income flow the total capital gain on the previous end-of-period
assets (or liabilities) and then dividing this total return by the previous value of assets (or
liabilities). Thus, in 2003, a year in which the dollar depreciated, the rate of return on U.S.
foreign assets was 19 percent, and that on liabilities 8.4 percent. Total capital gains are cal-
culated by subtracting the change in U.S.-owned assets abroad (change in foreign-owned
assets in the United States), as reported in the financial account, from the change in U.S. for-
eign assets (liabilities) at market value, as reported in the BEA international position data.
Maurice Obstfeld and Kenneth S. Rogoff 83
Figure 7. Foreign Exchange Reserves, Selected Countries, Various Years

Billions of dollars

Dec 1997
800 Dec 2001
700 May 2005

600
500
400
300
200
100

Japan China India NIEsa ASEAN4b

Sources:–The Economist, Ministry of Finance of Japan.

a.–Newly industrializing economies (Brazil, Hong Kong, Korea, Singapore, Taiwan).
b.–Four members of the Association of Southeast Asian Nations (Indonesia, Malaysia, Philippines, Thailand).

in equities (both portfolio equity and foreign direct investment) than do

foreigners of their U.S. assets. At the end of 2003, Americans held almost
$7.9 trillion in foreign assets, of which 60 percent was in equities, either
foreign stocks or foreign direct investment (here measured at market value).
Foreigners, by contrast, held only 38 percent of their $10.5 trillion in U.S.
assets in the form of equity. Given that equity has, over long periods, con-
sistently paid a significant premium over bonds, it is not surprising that
U.S. residents have remained net recipients of investment returns even
though the United States apparently crossed the line to being a net debtor
in the late 1980s.
A major reason why foreigners hold relatively more U.S. bonds than
Americans hold foreign bonds is that the dollar remains the world’s main
reserve and vehicle currency. Indeed, of the $3.8 trillion in international
reserves held by central banks worldwide, a very large share is in dollars,
and much of it is in short-term instruments.16 Figure 7 illustrates the bur-
geoning reserves of Asia, now in excess of $2 trillion. According to the

16. See the survey in Central Banking, “The Rise of Reserve Management,” March
2005, p. 14.
 84 Brookings Papers on Economic Activity, 1:2005

BEA, over 45 percent of the $700 billion stock of dollar currency is held
abroad, and this is probably an underestimate.17 (Note that, when one speaks
of the United States enjoying rents or seigniorage from issuing a reserve
currency, the main effects may come from foreigners’ relative willingness to
hold cash or liquid short-term Treasury debt, rather than from any substantial
inherent U.S. interest rate advantage.) In any event, our empirical analysis
will take account of the systematically lower return on U.S. liabilities than
on assets elsewhere, and will ask what might happen should that advan-
tage suddenly disappear in the process of current account reversal.18
At present, as we have noted, the net U.S. foreign debt equals about
25 percent of GDP. This ratio already roughly equals the previous peak of
26 percent, reached in 1894. A simple calculation shows that if U.S. nom-
inal GDP grows at 6 percent a year and the current account deficit remains
at 6 percent of nominal GDP, the ratio of U.S. net foreign debt to GDP
will asymptotically approach 100 percent. Few countries have ever reached
anywhere near that level of indebtedness without having a crisis of some
sort.19
If large, sudden exchange rate movements are possible, the greater depth
of today’s international financial markets becomes a potential source of
systemic stress. As we have documented, the volume of international asset
trading is now vast. Although many participants believe themselves to be
hedged against exchange rate and interest rate risks, the wide range of lightly
regulated or unregulated nonbank counterparties now operating in the
markets raises a real risk of cascading financial collapse. In a world where
a country’s current account may adjust abruptly, bringing with it large
changes in international relative prices, a persistently large U.S. deficit
constitutes an overhanging systemic threat.
A sober assessment of present global imbalances suggests the need for
a quantitative analysis of how a U.S. current account adjustment would
affect exchange rates. We take this up next.

17. See Porter and Judson (1996).

18. Of course, multinationals’ practice of income shifting in response to differing
national tax rates on profits distorts reported investment income flows, making an accurate
picture of the true flows difficult to obtain. See, for example, Grubert, Goodspeed, and
Swenson (1993) and Harris and others (1993). The expansion of gross international posi-
tions over the past decade may have worsened this problem.
19. Obstfeld and Rogoff (2000a).
Maurice Obstfeld and Kenneth S. Rogoff 85

Summary of the Analytical Framework

Here we summarize the main features and mechanisms in our analysis.

After reading this section, readers who are primarily interested in our
exchange rate predictions can skip the following section, which presents
the details of the model, and proceed directly to the discussion of our
numerical findings.
We work within a three-region model of a world economy consisting
of the United States, Europe, and Asia. These regions are linked by trade
and by a matrix of international asset and liability positions. Each region
produces a distinctive export good, which its residents consume along with
imports from the other two regions. In addition, each region produces non-
traded goods, which its residents alone consume.
A key but realistic assumption is that each country’s residents have a
substantial relative preference for the traded good that is produced at home
and exported; that is, consumption of traded goods is intensive in the home
export, creating a home bias in traded goods consumption. This feature builds
in a “transfer effect” on the terms of trade, which provides one of the key
mechanisms through which changes in the international pattern of current
account balances change real and nominal exchange rates. A reduction in
the U.S. current account deficit, if driven by a fall in U.S. spending and a
matching rise in U.S. saving, represents a shift in world demand toward
foreign traded goods, which depresses the price of U.S. exports relative to
that of imports from both Asia and Europe. (The international terms of
trade of the United States deteriorate.) Because the U.S.-produced export
good has a larger weight in the U.S. consumer price index (CPI) than that
of foreign imports, whereas foreign export goods similarly have larger
weights in their home countries’ CPIs, the result is both a real and a nom-
inal depreciation of the dollar.
This terms-of-trade effect of current account adjustment has been promi-
nent in the literature, but it is potentially less important quantitatively than
is a second real exchange rate effect captured in our model. That effect is
the impact of current account adjustment on the prices of nontraded
goods. The CPI can be viewed as made up of individual sub-CPIs for
traded and nontraded goods, with the latter empirically having about three
times the weight of the former in the overall CPI, given the importance of
nontraded service inputs into the delivery even of traded products to con-
sumers. The real exchange rate between two currencies is the ratio of the
 86 Brookings Papers on Economic Activity, 1:2005

issuing countries’ overall CPIs, both expressed in a common currency.

Thus a fall in a country’s prices for nontraded goods, relative to the same-
currency price of nontraded goods abroad, will depress its relative price
level just as a terms-of-trade setback does, causing both a real and a nomi-
nal depreciation of its currency. Because nontraded goods are so impor-
tant a component of the CPI, ignoring effects involving their prices would
omit much of the effect of current account adjustment on exchange rates.
Hence this additional mechanism, absent from much of the policy discus-
sion, is critical to include.
When the U.S. external deficit falls as a result of a cut in domestic con-
sumption, part of the reduction in demand falls on traded goods (exports
as well as imports), but much of it falls on U.S. nontraded goods. The con-
sequent fall in the nontraded goods’ prices reinforces the effect of weaker
terms of trade in causing the dollar to depreciate against the currencies of
Europe and Asia. As noted, in our calibration this second effect receives
more than twice the weight that terms-of-trade effects receive in explain-
ing exchange rate movements.
We consider several scenarios for U.S. current account adjustment,
involving different degrees of burden sharing by Europe and Asia and the
resulting effect on those regions’ bilateral and effective exchange rates. For
example, if Europe’s deficit rises to offset a fall in America’s deficit, while
Asia’s surplus remains constant, the dollar will depreciate more against
Europe’s currencies, and less against Asia’s, than if Asia and Europe shared
in the burden of accommodating the U.S. return to external balance. In terms
of its trade-weighted effective exchange rate, the dollar depreciates more
under the second of these two scenarios. Because Asia trades more with the
United States than Europe does, bilateral depreciation against Asia’s curren-
cies plays the more important role in determining the effective depreciation.
We also consider the effect of dollar exchange rate changes in revalu-
ing gross foreign asset positions, thus redistributing the burden of interna-
tional indebtedness, as well as the possibility that the adjustment process,
especially if disorderly, could entail higher interest payments abroad on U.S.
short-term foreign obligations. Finally, key parameters in our model gov-
ern the substitutability in consumption among various traded goods and
between traded and nontraded goods. In general, the lower these substitu-
tion elasticities, the greater the relative price changes caused by current
account adjustment and the greater, therefore, the resulting terms-of-trade
and exchange rate responses. Because the values of these elasticities are
Maurice Obstfeld and Kenneth S. Rogoff 87

quite uncertain and can differ between the short and the long run, we
quantitatively examine their role in generating our numerical estimates.

The Model

The three-country endowment model we develop here extends our earlier

small-country and two-country frameworks.20 We label the three coun-
tries (or regions), whose sizes can be flexibly calibrated, U (for the United
States), E (for Europe), and A (for Asia). The model distinguishes both
between home- and foreign-produced traded goods and between traded
and nontraded goods (with the latter margin, largely ignored in many dis-
cussions of the U.S. current account deficit, turning out to be the more
important of the two quantitatively in our simulations). Our focus here
will be on articulating the new insights that can be gained by going from
two countries to three, particularly in understanding different scenarios of
real exchange rate adjustment across regions as the current account deficit
of the United States falls to a sustainable level.
Four features of our model are of particular interest. First, by assuming
that endowments are given exogenously for the various types of outputs,
we implicitly assume that capital and labor are not mobile between sec-
tors in the short run. To the extent that global imbalances close only
slowly over long periods (which experience suggests is not the most likely
case), factor mobility across sectors will mute any real exchange rate
effects.21 Second, we do not allow for changes in the mix of traded goods
produced or for the endogenous determination of the range of nontraded
goods, two factors that would operate over the longer run and could also
mute the effects on real exchange rates of current account movements.
Third, our main analysis assumes that nominal prices are completely flexi-
ble. That assumption—in contrast to our assumption on factor mobility—
almost surely leads us to understate the likely real exchange rate effects of
a current account reversal. As we discuss later, with nominal rigidities
and imperfect pass-through from exchange rates to prices, the exchange
rate will need to move more, and perhaps much more, than in our base

20. See Obstfeld and Rogoff (2000a and 2004, respectively).

21. Obstfeld and Rogoff (1996).
 88 Brookings Papers on Economic Activity, 1:2005

case in order to maintain employment stability. Fourth, we do not explicitly

model the intertemporal allocation of consumption, but rather focus on
the intratemporal price consequences of alternative patterns of production-
consumption imbalances.

The Core Model

Although notationally intricate, our core three-region model is concep-
tually quite simple. We assume that consumers in each of the three regions
allocate their spending between traded and nontraded goods. Within the
category of traded goods, they choose among goods produced in each of
the three regions. The equilibrium terms of trade and the relative price of
traded and nontraded goods (and thus both bilateral and effective real
exchange rates) are determined endogenously. Given assumptions about
central bank policy (depending, for example, on whether the central bank
aims to stabilize the CPI deflator, the GDP deflator, or a bilateral exchange
rate), the model can also generate nominal exchange rates.
We begin by defining Cij ≡ country i consumption of good (or good
category) j. The comprehensive country i consumption index depends on
U.S., European, and Asian traded goods consumption (T), as well as con-
sumption of domestic nontraded goods (N). It is written in the following
nested form:
θ

C i =  γ (CTi ) + (1 − γ ) (C Ni )  , i = U , E , A,
1 θ −1 1 θ −1 θ −1
θ θ θ
(1) θ

with
η
η −1 η −1 η −1

CTU = α (CUU ) + (β − α ) (CEU ) + (1 − β ) (C UA ) 

1 1 1 η −1
η η η η η
(2) η

η
η −1 η −1 η −1

CTE = α (CEE ) + (β − α ) (CUE ) + (1 − β ) (C AE ) 

1 1 1 η −1
η η η η η η

 
1 1 η −1
η −1
1 − δ η η −1
1 − δ η η −1

CTA = δ (C AA ) (CEA ) (CUA )  .

1
η η
+ η
+ η

 2  2  

We do not assume identical preferences in the three countries. On the

contrary, we wish to allow, both in defining real exchange rates and in
Maurice Obstfeld and Kenneth S. Rogoff 89

assessing the effects of shocks, for a realistic home bias in traded goods
consumption, such that each country has a substantial relative preference
for the traded good that it produces and exports abroad.22 Home consump-
tion bias gives rise to a “transfer effect,” whereby an increase in relative
national expenditure improves a country’s terms of trade, that is, raises
the price of its exports relative to that of its imports.
In the equations above, the United States and Europe are “mirror sym-
metric” in their preferences for each other’s goods, but each attaches the
same weight to Asian goods. Asia weights U.S. and European imports
equally but may differ in openness from the United States and Europe.
Specifically, we assume that 1 > β > α > 1⁄2. We also assume that δ > 1⁄2. For
example, if β = 0.8 and α = 0.7, then the U.S. traded goods consumption
basket has a weight of 0.7 on U.S. exports, 0.1 on European exports, and 0.2
on Asian exports. (A very similar—and for many exercises isomorphic—
model arises if one assumes that all countries have identical preferences,
but that international trading costs are higher than domestic trading costs.)23
The values of the two parameters θ and η are critical in our analysis.
Parameter θ is the (constant) elasticity of substitution between traded and
nontraded goods. Parameter η is the (constant) elasticity of substitution
between domestically produced traded goods and imports from either for-
eign region. The two parameters are important because they underlie the
magnitudes of price responses to quantity adjustments. Lower substitution
elasticities imply that sharper price changes are needed to accommodate a
given change in quantities consumed.

Price Indexes and Real Exchange Rates

Using standard methods, we derive exact consumption-based price
indexes.24 Define Pji ≡ the country i exact price index for consumption cat-
egory j. The corresponding overall CPIs, in dollars, are
1
1− θ

(3) P =  γ ( PTi )
C
i
1− θ
+ (1 − γ ) ( P
N
i
)
1− θ
 , i = U , E , A,

22. Warnock (2003) also takes this approach.

23. Obstfeld and Rogoff (2000b).
24. See, for example, Obstfeld and Rogoff (1996).
 90 Brookings Papers on Economic Activity, 1:2005

where subscript C denotes the comprehensive consumption basket. (Our

main analysis is in terms of real prices and exchange rates, so all prices can
be expressed in terms of the common numeraire.) In equation 3,

PTU = [ αPU1− η + (β − α ) PE1− η + (1 − β ) PA1− η ]

1
1− η
(4)
PTE = [ αPE1− η + (β − α ) PU1− η + (1 − β ) PA1− η ]
1
1− η

  1 − δ  1− η  1 − δ  1− η  1− η

P = δPA1− η + 
A
P + P .
T
 2  U 2  E 
Here Pi, i = U, E, A, is just the price of the differentiated traded good pro-
duced by country i.
We assume the law of one price for traded goods, so that the price of any
given country’s traded good is the same in all regions. (In practice, of course,
the law of one price holds mainly in the breach, partly because of the dif-
ficulties in separating out the truly tradable component of “traded” goods.)
Because of the home export consumption bias we have assumed, the price
indexes for traded goods PTi can differ across countries even when the law
of one price holds, reflecting the asymmetric consumption weightings. As
a result, changes in the terms of trade, through their differential effects on
countries’ price levels for traded goods, affect real exchange rates.
There are three bilateral terms of trade, three bilateral real exchange
rates, and three real effective exchange rates. The terms of trade are
PE P P τ
(5) τU , E = , τU , A = A , τ E , A = A = U , A .
PU PU PE τU , E

Here, for example, a rise in τU,E is a rise in the price of European traded
goods in terms of U.S. traded goods, that is, a deterioration in the U.S. terms
of trade. Bilateral real exchange rates are
PCE PCA PCA q
(6) qU ,E = , qU ,A
= , q E ,A
= = U ,A .
PCU
PCU
PCE
qU ,E

A rise in qU,E, for example, is a rise in the price of the European consump-
tion basket in terms of the U.S. consumption basket, that is, a real depre-
ciation of the dollar.
As we have noted, asymmetric preferences over traded goods allow the
terms of trade to affect traded goods price indexes. The United States’
price index places a comparatively high weight on U.S. exports, whereas
Maurice Obstfeld and Kenneth S. Rogoff 91

Europe’s does the same for its own exports. Thus the U.S. traded goods
price index falls relative to Europe’s when Europe’s bilateral terms of
trade against the United States improve. Denoting a percent change with a
caret, we can logarithmically approximate the evolution of the relative
European-to-American traded goods price ratio as

(7) PˆTE − PˆTU = ( 2α − β ) τˆ U ,E .

(Exact formulas for relative price indexes, which we use to generate the
numerical results reported below, are given in appendix A.) This expres-
sion equates the difference between European and U.S. price inflation in
traded goods to the European consumption weight on its own exports, α,
less the U.S. consumption weight on imports from Europe, β − α, all
multiplied by the percentage increase in Europe’s terms of trade against the
United States. Observe that the terms of trade against Asia do not enter this
expression. Given the bilateral Europe-U.S. terms of trade, changes in the
terms of trade against Asia enter the European and U.S. traded goods price
indexes symmetrically (that is, with identical consumption weights of 1 − β)
and therefore drop out in computing their log-difference change.
Similarly, the evolution of the Asian price level for traded goods rela-
tive to that of the United States also reflects terms-of-trade movements.
But because, under our assumptions, Asia trades more extensively with
Europe than the United States does, the prices of European exports have a
relatively bigger impact on Asia’s average import prices. This is shown
by the following logarithmic approximation:

 1 − δ  
(8) PˆTA − PˆTU = [ δ − (1 − β )] τˆ U , A +   − (β − α )  τˆ U ,E .
 2  

The weights on the terms-of-trade changes here simply reflect relative con-
sumption weights, as before. Now, however, given the bilateral Asia-U.S.
terms of trade, an improvement in Europe’s terms of trade vis-à-vis the
United States raises Asia’s price index for traded goods relative to that in the
United States when, as we assume in our simulations, the Asian consump-
tion weight on European imports, (1 − δ)/2, exceeds the weight attached
by U.S. consumers, β − α. Such third-country asymmetries cannot be cap-
tured, of course, in a two-country framework.
Bilateral real exchange rate movements follow immediately from the
expressions above. For Europe and the United States, for example, the log
 92 Brookings Papers on Economic Activity, 1:2005

change in the bilateral real exchange rate is simply the consumption

weight on traded goods times the log change in relative traded goods price
indexes, plus the consumption weight on nontraded goods times the log
change in relative nontraded goods price indexes:
(9) qˆU ,E = γ ( 2α − β ) τˆ U ,E + (1 − γ ) ( PˆNE − PˆNU ) .

Analogously, between the United States and Asia we have

 1 − δ  
(10) qˆU , A = γ [ δ − (1 − β )] τˆ U , A + γ   − (β − α )  τˆ U ,E
 2 
+ 1 − γ P − P ).
( ) ( ˆ
N
A ˆ
N
U

We emphasize one key aspect of these expressions. The weight on

nontraded goods is likely to be quite large because of the large component
of nontradable services included in the consumer prices of goods gener-
ally classified as entirely tradable. In our simulations we therefore take
the weight on nontraded goods above, 1 − γ, to be 0.75. An implication is
that, although the terms of trade certainly are an empirically important
factor in real exchange rate determination given home consumption bias,
relative prices for nontraded goods potentially play an even larger quanti-
tative role.

Solution Methodology
The methodology we use to calculate the effects of current account shifts
on relative prices is essentially the same as that in our earlier papers,
extended to a three-region setting.25 Given fixed output endowments, an
assumed initial pattern of current account imbalances, an assumed initial
pattern of international indebtedness, and a global interest rate, relative
prices are determined by the equality of supply and demand in all goods
markets. Changes in the international pattern of external imbalances, whether
due to consumption shifts or other changes (including changes in produc-
tivity), shift the supply and demand curves in the various markets, result-
ing in a new set of equilibrium prices. These are the price changes we
report below, under a variety of current account adjustment scenarios. (The
global sums of external imbalances and of net international asset positions
are both constrained to be zero.)

25. The methodology is specified in appendix A and further online at www.economics.

harvard.edu/faculty/rogoff/papers/BPEA2005.pdf.
Maurice Obstfeld and Kenneth S. Rogoff 93

There are six market-clearing conditions, covering the three regional

nontraded goods markets and the three global markets for traded goods
(although one of these is redundant by Walras’ Law). The five indepen-
dent equilibrium conditions allow solutions for
—the U.S. terms of trade against Europe, τU,E
—the U.S. terms of trade against Asia, τU,A
—the price of nontraded goods in terms of traded goods in the United
States, PNU/PUT
—the price of nontraded goods in terms of traded goods in Europe,
PNE/PET
—the price of nontraded goods in terms of traded goods in Asia, PNA/PTA.
One can then calculate the three bilateral real exchange rates, for which
these five relative prices are the critical inputs. Because of the asymmetric
preferences over traded goods, there is, as we have noted, a transfer effect
in the model (wealth transfers feed into the terms of trade and through that
channel into real exchange rates), although it is more complex than would be
the case with only two countries in the world. Finally, we will also want
to define and analyze real effective (loosely speaking, trade-weighted)
exchange rates:
β−α 1− β

(P ) (P )
E 1− α A 1− α

(11) qU = C C

PCU
β−α 1− β

(P ) (P )
U 1− α A 1− α

q =
E C C

PCE

(P ) (P )
1 1
U 2 E 2

q = A C C
.
PCA

Three extensions to the analysis add to its relevance and realism.26

First, we ask how real exchange rate changes translate into nominal
exchange rate changes; this depends on central bank policy. In general,
this turns out not to be a critical issue empirically; the other two exten-
sions are potentially far more important. One of these is to take into
account how exchange rate changes affect the net foreign asset positions
of the different regions, because of currency mismatches between gross

26. Details can be found in appendix A and online at www.economics.harvard.edu/

faculty/rogoff/papers/BPEA2005.pdf.
 94 Brookings Papers on Economic Activity, 1:2005

assets and liabilities.27 This valuation effect is significant, but its impact on
aggregate demand is of secondary importance compared with the primary
demand shifts emphasized in our preceding analysis. Finally, our third
extension takes into account the effect of a rise in relative U.S. interest
rates (due, say, to concern about government deficits or erosion of the dol-
lar’s reserve currency status). This effect, which works to worsen rather
than ease the adjustment problem, is also significant, although again it is
less important (at least over the range of interest rates we consider) than
the primary effects of a rebalancing of global demand.

Model Predictions

With these critical behavioral parameters in hand, we are now ready to

explore the model’s quantitative predictions for global exchange rates and
the terms of trade under various scenarios for rebalancing the U.S. current
account. We first need to think about parametrizing the model.

Choosing Parameters
As we have already observed, the critical parameters in the model are θ,
the elasticity of substitution in consumption between traded and nontraded
goods, and η, the elasticity of substitution in consumption among the traded
goods produced by the three regions. The lower are these elasticities, the
greater the exchange rate and price adjustments needed to accommodate
any interregional shifts in aggregate demand. Most of our simulations will
be based on a value of θ = 1, which is high relative to some estimates sug-
gested in the literature.28 We will also report results, however, for an even
higher elasticity of θ = 2.
Our baseline choice of η = 2 as a representative aggregate trade elas-
ticity is a compromise between two sets of evidence. Estimates based on
trade flows within disaggregated product categories cover a wide range

27. As noted above, this effect has recently been emphasized by Tille (2004), Lane and
Milesi-Ferretti (2005a, 2005b), and Gourinchas and Rey (2005a, 2005b).
28. Mendoza’s (1991) point estimate is 0.74, Ostry and Reinhart (1992) report esti-
mates in the range 0.66 to 1.28 for a sample of developing countries, and Stockman and
Tesar (1995) use an estimate of 0.44. Using a different approach, Lane and Milesi-Ferretti
(2004) derive estimates as low as 0.5. Indeed, for larger and relatively closed economies
(such as the United States, Europe, and Japan), they suggest that the value should be even
lower.
Maurice Obstfeld and Kenneth S. Rogoff 95

but typically include many values much higher than η = 2.29 On the other
hand, conventionally estimated aggregate trade equations, as well as cali-
brations of dynamic general equilibrium models, tend to indicate much
smaller values for η, typically 1 or even lower.
A number of mechanisms have been suggested to explain this discrep-
ancy, some echoing Guy Orcutt’s classic skepticism about the low elastic-
ities seemingly implied by macro-level estimators.30 Aggregation bias
lowers estimated macroelasticities because the price movements of low-
elasticity goods tend to dominate overall movements in import and export
price indexes.31 Another issue is that macroeconomic estimates of business-
cycle frequency correlations tend to confound permanent and temporary
price movements, in contrast to micro-level cross-sectional or panel stud-
ies centered on trade liberalization episodes.32 In taking η = 2, we try, in a
crude way, to address these biases while also recognizing the empirically
inspired rules of thumb that inform policymakers’ forecasts. We also include
an illustrative simulation of the case η = 100 (in which all traded goods
are essentially perfect substitutes). That simulation shuts down the terms-
of-trade effects and thereby shows how large a role is being played by
substitution between traded and nontraded goods, the channel we have
emphasized elsewhere.33
We set both α and δ equal to 0.7; these are the consumption weights
that Americans and Europeans, on the one hand, and Asians, on the other,
attach to their own domestic products within their traded goods consump-
tion baskets. That choice is plausible based on our discussion in an earlier

29. Examples are the estimates of Feenstra (1994) and the more recent figures of Broda
and Weinstein (2004).
30. Orcutt (1950).
31. For an excellent example of this bias in action, see Hooper, Johnson, and Marquez
(2000), who report that, because oil and tourism demand are relatively price-inelastic, trade
equations based on aggregates that include oil and services imply apparently much lower
price elasticities than equations for nonoil manufactures only. For the Group of Seven
countries, Hooper, Johnson, and Marquez report short-run price elasticities for imports and
exports (including oil and services) that in most cases do not satisfy the Marshall-Lerner
condition. We view the elasticities implied even by aggregated estimates that exclude oil
and services as unreasonably low; but, if they are accurate, they imply larger terms-of-trade
and real exchange rate effects of international spending shifts.
32. See Ruhl (2003). Our model omits not only dynamics of the type suggested by
Ruhl, but also those resulting from the introduction of new product varieties, which would
act over the longer run to dampen the extent to which a rise in a country’s relative produc-
tivity lowers its terms of trade. See, for example, Krugman (1989) and Gagnon (2003).
33. Obstfeld and Rogoff (2000a).
 96 Brookings Papers on Economic Activity, 1:2005

paper.34 We set β = 0.8, implying that Europe and the United States alike
place weights of β − α = 0.1 on each other’s traded goods, and twice that
weight (0.2) on Asian goods. Asia, by assumption, distributes its demand
evenly across the other two regions (placing a weight of 0.15 on the exports
of each). So, in our model, Europe and the United States both trade more
with Asia than with each other. We assume that all three regions produce
the same number of units of tradable goods output.
Appendix A discusses in detail our assumptions regarding gross liabil-
ities and assets for each region, as well as the currencies of denomination
of these stocks. The point we stress here is that, to a first approximation,
the United States is a net debtor (to the tune of 25 percent of its GDP, or
100 percent of its exportable GDP), and greater Europe has approximately
a zero net international position. Our model’s third region, Asia, therefore
is left as a net international creditor in an amount equal to 100 percent of
U.S. tradable GDP. U.S. gross foreign liabilities are almost all in dollars,
but U.S. gross foreign assets are only about 40 percent in dollars. We
assume that greater Asia’s gross liabilities are equally divided among U.S.,
European, and Asian currencies (because Japan borrows in yen), whereas
Asian gross foreign assets are 80 percent in dollars and 20 percent in
European currencies. For Europe we assume that gross foreign assets are
32 percent in dollars, 11 percent in Asian currencies, and 57 percent in
European currencies. In our model, 80 percent of European gross liabili-
ties are denominated in European currencies, and the balance in dollars.
These numbers are very rough approximations, based in some cases on
fragmentary or impressionistic data, but portfolio shares can shift sharply
over time, and so there is little point in trying too hard to refine the esti-
mates. As we shall see, these shares do imply large potential international
redistributions of wealth due to exchange rate changes, but those redistri-
butions themselves have only a secondary impact on the exchange rate
implications of current account adjustment.
For nominal interest rates we take a baseline value of 3.75 percent a year
for U.S. liabilities but 5 percent a year for all other countries’ liabilities.
This assumption captures the “exorbitant privilege” the United States has
long enjoyed of borrowing from the world more cheaply than it lends.35

34. Obstfeld and Rogoff (2000b).

35. The phrase “exorbitant privilege” is commonly but wrongly attributed to French
president Charles de Gaulle. For its true origin, see the interesting historical note provided
by Gourinchas and Rey (2005b).
Maurice Obstfeld and Kenneth S. Rogoff 97

Turning to current accounts, we place the U.S. external deficit at 20 per-

cent of U.S. tradable GDP.36 This is consistent with a U.S. current account
deficit of 5 percent of total GDP, a reasonable baseline if part of the 2004
deficit is due to temporarily high oil prices. Because we find our simula-
tion results to be approximately linear within the parameter space we are
considering, it is easy to adjust the prediction to the case in which the
2004 deficit of 6 percent of GDP persists. In any event, what matters most
for our calibration is how much the current account balance adjusts (for
example, from 6 to 3 percent of GDP). We assume an initial position with
Europe’s current account surplus at 5 percent of U.S. tradable GDP and
Asia’s at 15 percent.37
A final benchmark to establish is our initial reference value for measur-
ing subsequent exchange rate adjustments. This issue was less critical in our
earlier two papers, because trade-weighted effective exchange rates move
more slowly than the bilateral exchange rates that we consider below. In our
basic model prices are flexible and economic responses to them are imme-
diate. In practice, however, there are considerable lags: Michael Mussa, for
example, posits the rule of thumb that the U.S. trade balance responds with
a two-year lag to dollar exchange rate changes.38 In that case, if today’s cur-
rent account balances reflect averages of exchange rates over the past two
years, it would be more accurate to think of our simulations as giving
exchange rate changes relative to two-year average reference rates rather
than current rates. Table 2 presents some resulting reference exchange
rates. (The Chinese and Malaysian currencies have been pegged over the
past two years, and so their current and average rates are the same.)

Simulations
With the model and our parameter assumptions in hand, we are ready to
consider alternative simulations. Underlying much of our analysis is the
assumption that demand shocks (such as a rise in U.S. saving) are driving
the redistribution of global imbalances. This seems by far the most realis-
tic assumption, given the magnitude of the external financing gaps.

36. As noted earlier, we estimate tradable GDP to be at most 25 percent of total GDP.
37. It would be interesting and useful to extend the model to include emerging markets
and OPEC as a composite fourth region, as suggested by our discussant T. N. Srinivasan.
38. Mussa (2005).
 98 Brookings Papers on Economic Activity, 1:2005
Table 2. Recent and Two-Year-Average Exchange Rates of Selected Currencies
Currency units per dollar except where noted otherwise
Exchange rate

Currency As of June 1, 2005 Two-year average

a
U.K. pound sterling 1.81 1.79
Canadian dollar 1.25 1.23
Euroa 1.22 1.23
Korean won 1,010 1,129
New Taiwan dollar 31.30 33.21
Singapore dollar 1.67 1.69
Japanese yen 108.4 109.3
Source: Federal Reserve data.
a. In dollars per indicated currency unit.

Tables 3 through 6 lay out the results of three scenarios under which the
U.S. current account balance might improve by 20 percent of tradable
GDP or, equivalently, 5 percent of total GDP. (All simulations include the
effect of exchange rate changes in revaluing the regions’ foreign assets and
liabilities.) In the “global rebalancing” scenario (the first column in each
table), all regions’ current account balances go to zero (with trade balances
adjusting as needed to service interest flows on the endogenously deter-
mined stocks of net foreign assets). Looking first at bilateral real exchange
rates, in table 3, we see that Asia’s exchange rate with the United States
rises by 35.2 percent, and Europe’s rises by 28.6 percent (we define the
real exchange rate such that these changes indicate real depreciations of
the dollar). Europe sees an improvement in its terms of trade against the
United States (a rise in the price of Europe’s exports relative to its U.S.
imports) of 14.0 percent, and Asia sees an improvement of 14.5 percent.
What are the implications for nominal exchange rates? To answer this
question we must specify monetary policies. We consider two possibili-
ties: that central banks stabilize the domestic CPI, and that they stabilize
the domestic GDP deflator. Table 4 reports the results. Under CPI target-
ing, the monetary authorities hold overall price levels constant, so that the
only source of real exchange rate change is nominal exchange rate change.
As a result, nominal and real exchange rate changes are equal, as can be
seen by comparing table 4 with table 3.39 Because none of the three regions
is extremely open to trade, movements in CPIs and in GDP deflators are

39. We provide a detailed account of nominal exchange rate determination under GDP
deflator targeting at www.economics.harvard.edu/faculty/rogoff/papers/BPEA2005.pdf.
Maurice Obstfeld and Kenneth S. Rogoff 99

Table 3. Changes in Real Exchange Rates and Terms of Trade Following U.S.
Current Account Adjustment under Baseline Assumptionsa
Log change × 100
Adjustment scenario

Real exchange rate Global Bretton Europe and United

or terms of trade rebalancingb Woods IIc States trade placesd
Real exchange rate
United States/Europe 28.6 58.5 44.6
United States/Asia 35.2 −0.5 19.4
Europe/Asia 6.7 −59.0 −25.2
Terms of trade
United States/Europe 14.0 29.4 22.0
United States/Asia 14.5 7.2 11.1
Europe/Asia 0.5 −22.2 −10.8
Source: Authors’ calculations using model described in the text.
a. Exchange rates are defined such that an increase represents a real depreciation of the first region’s currency against the sec-
ond’s; terms of trade are defined such that an increase represents a deterioration for the first region (that is, a fall in the price of
the first region’s export good against the second). Assumed parameter values are as follows: substitution elasticity between traded
and nontraded goods θ = 1; substitution elasticity between traded goods of different regions η = 2; share of traded goods in total
consumption γ = 0.25.
b. Current account balances of all three regions go to zero.
c. Asia’s current account surplus rises to keep its exchange rate with the dollar fixed. Europe’s current account absorbs all
changes in the U.S. and Asian current accounts.
d. Europe absorbs the entire improvement in the U.S. current account balance while Asia’s current account balance remains
unchanged.

Table 4. Changes in Nominal Exchange Rates Following U.S. Current Account

Adjustment under Alternative Inflation Targetsa
Log change × 100
Adjustment scenario

Nominal Global Bretton Europe and United

exchange rate rebalancing Woods II States trade places
Target is consumer price indexb
United States/Europe 28.6 58.5 44.6
United States/Asia 35.2 −0.5 19.4
Europe/Asia 6.7 −59.0 −25.2
Target is GDP deflator
United States/Europe 30.0 61.4 46.8
United States/Asia 36.9 0.0 20.6
Europe/Asia 6.9 −61.4 −26.3
Source: Authors’ calculations using model described in the text.
a. See table 3 for definitions of exchange rates, scenarios, and parameter assumptions.
b. With flexible prices and CPI targeting by central banks, nominal exchange rate changes are equal to the real exchange rate
changes reported in table 3.
 100 Brookings Papers on Economic Activity, 1:2005
Table 5. Changes in Real and Nominal Effective (Trade-Weighted) Exchange Rates
Following U.S. Current Account Adjustment under Baseline Assumptionsa
Log change × 100
Adjustment scenario
Effective Global Bretton Europe and United
exchange rateb rebalancing Woods II States trade places
U.S. real −33.0 −19.1 −27.8
U.S. nominal −34.6 −20.5 −29.3
Europe real 5.1 58.9 31.7
Europe nominal 5.4 61.4 33.1
Asia real 20.9 −29.8 −2.9
Asia nominal 21.9 −30.7 −2.9
Source: Authors’ calculations using model described in the text.
a. See table 3 for definitions of scenarios and parameter assumptions. An increase is an appreciation of the indicated currency
against foreign currencies.
b. Nominal exchange rate changes are calculated under the assumption of GDP deflator targeting; see appendix A for details.

fairly close, and, as a result, nominal exchange rate changes when the GDP
deflator is stabilized differ very little from those under CPI stabilization.
The appreciation of Europe’s currencies against the dollar is smaller
than that of Asia’s under the first scenario, because Asia starts out in our
simulation with a much larger external surplus than Europe does, and so it
has more adjusting to do. But the Asian currencies’ appreciation against the
dollar is mitigated somewhat by the fact that Asia trades more with the United
States than Europe does.40 We see in table 5 that Europe’s real effective
currency appreciation—represented, as is traditional for such multilateral
indexes, by a positive number—is much smaller than Asia’s: only 5.1 percent
versus 20.9 percent. Again, this reflects the greater weight of the dollar in
Asia’s trade-weighted real exchange rate than in Europe’s. Notice that, as
in table 4, nominal (under GDP deflator targeting) and real effective
exchange rate changes are again quite close numerically.
Another factor underlying the equilibrium exchange rate responses is that
dollar depreciation implies a much bigger reduction in Asia’s net foreign
asset position than in Europe’s. (Table 6 shows the impacts under GDP

40. Indeed, if one recalibrates the model so that β = 0.85 (in which case all countries’
preferences are completely symmetric, so that Europeans and Americans no longer prefer
Asian goods to each other’s), then, in the global rebalancing scenario, Asia’s currency
appreciates in real terms against the dollar by 37.8 percent and against European currencies
by 12.2 percent. These numbers exceed the 35.2 percent and 6.7 percent reported in table 3.
Maurice Obstfeld and Kenneth S. Rogoff 101

Table 6. Net Foreign Assets by Region Following U.S. Current Account Adjustmenta
Ratio to value of U.S. traded goods output
Adjustment scenariob

Baseline net
foreign asset Global Bretton Europe and United
Region position rebalancing Woods II States trade places
United States −1.0 −0.3 −0.1 −0.2
Europe 0.0 −0.1 −0.7 −0.4
Asia 1.0 0.4 0.8 0.6
Source: Authors’ calculations using model described in the text.
a. See table 3 for definitions of exchange rates, scenarios, and parameter assumptions.
b. Net asset positions taking into account valuation effects of changes in nominal exchange rates under GDP deflator targeting.

deflator targeting.) Asia has 80 percent of its assets, but only 34 percent of
its liabilities, in dollars. Thus, under the global rebalancing scenario, dol-
lar depreciation raises Asia’s gross liabilities relative to its gross assets,
pushing its net foreign assets down (as a fraction of U.S. tradable GDP)
by 60 percent. Europe, by contrast, has only 32 percent of its assets and
20 percent of its liabilities in dollars. The fact that Asia loses so much on
the asset side implies that its trade surplus shrinks by less than its current
account surplus does. Because trade surpluses are what drive the constel-
lation of real exchange rates, the real appreciation of the Asian currencies
is mitigated. In sum, thanks to Asia’s greater openness and to the fact that
Asia suffers particularly large capital losses on foreign assets when the
dollar falls, Asian exchange rates do not need to change quite as much as
a model-free, back-of-the-envelope calculation might suggest.
The tables cover two other possible scenarios. The second column in
tables 3 through 6 analyzes a “Bretton Woods II” scenario, in which Asia
clings to its dollar peg.41 We calibrate this case by setting the U.S. current
account balance to zero and endogenously varying Asia’s and Europe’s
current account balances in a way that both maintains Asia’s bilateral
nominal exchange rate with the United States (assuming GDP deflator tar-
geting) and absorbs the fall in U.S. borrowing. (Of course, nonmonetary
policy instruments such as fiscal policy would have to be used to attain
just the right constellation of current account balances.) In this case the
bilateral real exchange rates of the European currencies against the dollar

41. Dooley, Folkerts-Landau, and Garber (2004a, 2000b).

 102 Brookings Papers on Economic Activity, 1:2005

must rise spectacularly, by 58.5 percent, and they would rise against the
Asian currencies by 59 percent. This result also is approximately linear in
the change in the U.S. current account balance. Thus, under the Bretton
Woods II scenario, eliminating only half the U.S. current account deficit
would raise the real value of the European currencies against the dollar by
as much as would occur in a global rebalancing scenario that eliminates
the U.S. current account deficit entirely.
For Asia to maintain its nominal exchange rate peg in the face of a bal-
anced U.S. current account, it must drive its own current account balance
significantly further into surplus, from 15 percent to 31 percent of U.S.
tradable GDP. And Europe would have to move from a surplus equal to
5 percent of U.S. tradable GDP to a 31 percent deficit! (See the footnotes
to table 3.) When Asia pegs its currencies to a falling dollar, its own
traded goods become more competitive and its imports more expensive
relative to domestic nontraded goods. Both factors shift world demand away
from Europe, which, by assumption, is passively absorbing the blow, and
toward Asia. These calibrations make patently clear why sustaining Asia’s
dollar peg is likely to be politically unpalatable for many of its trading
partners if the U.S. current account deficit ever shrinks. Asia would be
extremely vulnerable to a protectionist backlash.
As table 6 shows, the sharp appreciation of Europe’s currencies in the
Bretton Woods II scenario also decimates its external asset position,
which declines from balance to −70 percent of the value of U.S. tradable
production. Asia suffers somewhat, and the U.S. net asset position is the
major beneficiary, because U.S.-owned foreign assets are concentrated
in European currencies. Europe is thus hammered both by a sharp
decline in its competitiveness and by a loss on its net foreign assets of
about $2 trillion.
The third scenario reported in tables 3 through 6 is a muted version of
the Bretton Woods II scenario. Here, instead of maintaining its dollar cur-
rency peg, Asia maintains its current account surplus unchanged in the
face of U.S. adjustment to a balanced position. That is, rather than increas-
ing its current account surplus, it allows enough exchange rate adjustment
to keep the surplus constant. In this case, as table 5 shows, Europe’s real
effective exchange rate rises by much less than in the Bretton Woods II
scenario (31.7 percent versus 58.9 percent), and the Asian currencies expe-
rience a real effective depreciation of only 2.9 percent, versus 29.8 per-
cent in Bretton Woods II. Still, because the U.S. current account balance
Maurice Obstfeld and Kenneth S. Rogoff 103

improves dramatically while Asia’s holds steady, the Asian currencies rise
in real terms by 19.4 percent against the dollar (table 3). This exercise
reveals a fallacy in the argument that Asia cannot allow its dollar peg to
move without losing the ability to absorb its surplus labor. To the extent
that European demand increases, Asia can retain its external surplus while
releasing its dollar peg.
In table 7 we revisit the global rebalancing scenario but vary the critical
substitution elasticities in the model. (Only real exchange rate changes,
which equal nominal changes under CPI inflation targeting, are listed.) In
the first column we assume an elasticity of substitution between traded
and nontraded goods, θ, of 2 instead of 1. As we have already argued, the
limited evidence in the empirical macroeconomics literature suggests that
this estimate is well on the high side, but it allows us to incorporate a more
conservative range of potential exchange rate adjustments alongside our
baseline estimates. Under this assumption the real dollar exchange rate
with the European currencies rises by only 19.3 percent, instead of 28.6
percent as in the first column of table 3, and the Asian currencies rise
against the dollar by 22.5 percent instead of 35.2 percent. The dollar falls
in real effective terms (results not shown) by 21.5 percent rather than
33 percent. These calculations show that, even with a relatively high value
for θ, the required adjustment of exchange rates is quite significant even
if, as here, prices are flexible.

Table 7. Changes in Real Exchange Rates and Terms of Trade in the Global
Rebalancing Scenario under Alternative Calibrationsa
Log change × 100
Higher elasticity of Very high elasticity of
substitution between traded substitution between
Real exchange rate and nontraded goods regions’ traded goods
or terms of trade (θ = 2, η = 2) (θ = 1, η = 100)
Real exchange rate
United States/Europe 19.3 16.5
United States/Asia 22.5 23.5
Europe/Asia 3.3 7.0
Terms of trade
United States/Europe 14.6 0.0
United States/Asia 15.1 0.0
Europe/Asia 0.5 0.0
Source: Authors’ calculations using model described in the text.
a. In the global rebalancing scenario all regions’ current account balances go to zero. See table 3 for definitions of exchange
rates and other parameter assumptions.
 104 Brookings Papers on Economic Activity, 1:2005

The second column in table 7 examines the case in which θ = 1 but η =

100, so that the various countries’ tradable outputs are almost perfect sub-
stitutes. This exercise, which essentially eliminates terms-of-trade adjust-
ments as a factor in moving real exchange rates, allows us to see how much
of the change in exchange rates is due to within-country substitution
between traded and nontraded goods. This variation mutes the exchange
rate changes by an amount roughly similar to those found in the previous
exercise. The real effective dollar exchange rate (again not shown) falls by
21 percent. According to this calibration, roughly two-thirds of the needed
dollar adjustment is driven by substitution between traded and nontraded
goods, and only one-third is driven by the terms-of-trade channel typi-
cally emphasized in the literature. This should not be surprising, given that
(according to our previously cited calibration) roughly 75 percent of GDP
is nontraded. With more conservative assumptions about international trade,
however (either greater home bias in consumption or lower substitutability
of countries’ traded outputs, such that η = 1), the terms-of-trade channel
would become more important.
At present the United States is absorbing traded goods (domestic and
foreign) equivalent to roughly 30 percent of its GDP. This demand needs to
adjust downward while avoiding a reduction in nontraded goods absorption
if full employment is to be maintained; such a shift will therefore require
a significant change in the relative price of nontraded goods. Still, terms-
of-trade changes do account for about one-third of the overall adjustment,
a proportion slightly larger than that found in our two-country model,
where we did not allow for trade or terms-of-trade adjustments between
non-U.S. economies.
Given the United States’ leveraged international portfolio, with gross
debts mostly in dollars and assets significantly in foreign currencies, an
unexpected dollar depreciation reduces the U.S. net foreign debt. The first
two columns of table 8 report the results of simulations, within the global
rebalancing scenario, that illustrate the quantitative importance of such
asset valuation effects. Gourinchas and Rey have recently estimated that
nearly one-third of the settlement of the U.S. net foreign debt has histori-
cally been effected by valuation changes, with the remaining two-thirds
covered by higher net exports.42 The first column in table 8 shows results

42. Gourinchas and Rey (2005a).

Maurice Obstfeld and Kenneth S. Rogoff 105
Table 8. Changes in Real Exchange Rates and Terms of Trade in the Global
Rebalancing Scenario with and without Valuation and Interest Rate Effectsa
Log change × 100
With valuation Without valuation With valuation
Real exchange rate effects and without effects or interest effects and interest
or terms of trade interest rate effectsb rate effects rate effectsc
Real exchange rate
United States/Europe 28.6 33.7 30.1
United States/Asia 35.2 40.7 37.2
Europe/Asia 6.7 7.0 6.3
Terms of trade
United States/Europe 14.0 16.5 15.1
United States/Asia 14.5 16.5 15.3
Europe/Asia 0.5 0.0 0.2
Source: Authors’ calculations using the model described in the text.
a. In the global rebalancing scenario all regions’ current account balances go to zero. See table 3 for definitions of exchange
rates and other parameter assumptions.
b. Same as the baseline scenario reported in first column of table 3.
c. Interest rates on U.S. short-term liabilities held by foreigners are assumed to rise 1.25 percentage points, to the same level as
the return earned by U.S. residents abroad.

for the global rebalancing scenario with valuation effects taken into account
(identical to the first column in table 3). The second column shows the
changes in bilateral exchange rates that would be required if there were no
valuation effects (or, equivalently, if exchange rate changes were accu-
rately anticipated and nominal returns adjusted fully to compensate). All
relative price changes against the United States are larger in this case,
because the United States does not get the benefit of a sharp reduction in
its net dollar liabilities. Correspondingly, the U.S. trade balance needs to
adjust more for any given adjustment in the current account deficit. The
real exchange rate between the dollar and the European currencies needs
to move by 33.7 percent, rather than 28.6 percent when valuation effects
are taken into account, and the real value of the Asian currencies needs to
rise by 40.7 percent against the dollar instead of 35.2 percent. The real
effective dollar exchange rate falls by 37.8 percent instead of 33.0 per-
cent (results not shown). According to these numbers, asset revaluation
effects will mute the required movement in exchange rates as the U.S.
current account closes up, but the trade balance has to do the heavy lift-
ing, since 87 percent (33.0 ÷ 37.8) of the necessary real exchange rate
adjustment remains. That valuation effects have only a secondary effect
on equilibrium relative price changes is not surprising: big valuation
effects can only come from big exchange rate movements.
 106 Brookings Papers on Economic Activity, 1:2005

Our calculations so far do not take into account the likelihood of an

accompanying rise in global interest rates, which would hurt the United States
(a net debtor) and help Asia (a net creditor). A broad range of scenarios
are possible here; we examine only a single very simple one. (Appendix A
gives details of the calculation.) In the third column of table 8, we assume
that annual interest rates on short-term U.S. debt rise from 3.75 percent to
5 percent, the same level assumed for all other liabilities. In other words,
perhaps because of heightened risk perceptions, the United States simply
loses its historical low borrowing rate and is put on a par with other debtors.
This change wipes out a good deal of the effect of the valuation changes
(and would wipe out even more if it applied to all U.S. external liabilities,
not just the roughly 30 percent consisting of short-maturity debt). As our
introductory discussion suggested, the United States, as an important issuer
of bonds relative to equity, is extremely vulnerable to increases in interest
rates, even when all global bond rates rise together.
Until now we have been concentrating on demand shocks. Productivity
shocks may make the adjustment process more or less difficult, depending
on their source. Higher productivity in foreign traded goods production
can actually result in an even greater real depreciation of the dollar as
equilibrium is reestablished in world markets. If, on the other hand, it is
nontraded goods productivity in Asia and Europe that rises, the exchange
rate effects of global rebalancing will be muted. As table 9 illustrates, a

Table 9. Changes in Real Exchange Rates and Terms of Trade in Global

Rebalancing Scenario with Higher Productivity in Non-U.S. Nontraded Goodsa
Log change × 100
With 20 percent increase in
Real exchange rate Without increase productivity in European
or terms of trade in productivityb and Asian nontraded goods
Real exchange rate
United States/Europe 28.6 17.0
United States/Asia 35.2 23.6
Europe/Asia 6.7 6.6
Terms of trade
United States/Europe 14.0 15.0
United States/Asia 14.5 15.3
Europe/Asia 0.5 0.2
Source: Authors’ calculations using model described in the text.
a. In the global rebalancing scenario all regions’ current account balances go to zero. See table 3 for definitions of exchange
rates and other parameter assumptions.
b. Same as the baseline scenario reported in the first column of table 3.
Maurice Obstfeld and Kenneth S. Rogoff 107

20 percent rise in nontraded goods productivity outside the United States

implies notably smaller real exchange rate changes, although the terms-of-
trade shifts are similar. A large rise in U.S. traded goods productivity would
also facilitate a softer landing. In this case, however, although the extent
of real dollar depreciation is somewhat reduced, the U.S. terms of trade
fall much more sharply (results not reported).

Some Further Considerations

We believe our model offers many useful insights, but of course there are
many caveats to its interpretation. Some of these suggest that our results
understate the dollar’s potential decline, and some that they overstate it.

Intersectoral Factor Mobility

A critical implicit assumption of our model is that capital and labor
cannot quickly migrate across sectors, so that prices rather than quantities
must bear the burden of adjustment in response to any sudden change in
relative demands for different goods. This assumption seems entirely rea-
sonable if global current account adjustment (full or partial) takes place
moderately quickly, say, over one to two years. In the short run, workers
cannot change location easily, worker retraining is expensive, and a great
deal of capital is sector-specific. Over much longer periods, however (say,
ten to twelve years), factor mobility is considerable. If, for example,
prices rise dramatically in the U.S. traded goods sector, new investment
will be skewed toward that sector, as will new employment. Thus, in prin-
ciple, a gradual closing of the U.S. current account deficit would facilitate
much smoother adjustment with less exchange rate volatility. Unfortu-
nately, our model is not explicitly dynamic.43 One can, however, artifi-
cially approximate gradual current account adjustment by allowing for
progressively higher elasticities of substitution. We do this in table 10,
where we reconsider our central scenario (which assumed θ = 1 and η = 2)
by comparing it with two cases in which substitution elasticities are much
higher. As the table shows, in the case with “gradual” unwinding (proxied

43. For an example of a dynamic approach see the small-country q-model analysis in
Obstfeld and Rogoff (1996, chapter 4).
 108 Brookings Papers on Economic Activity, 1:2005
Table 10. Changes in Real Exchange Rates under Alternative Assumed Speeds of
Global Rebalancinga
Log change × 100
Speedb
Moderate Gradual Very gradual
Real exchange rate (1–2 years)c (5–7 years) (10–12 years)
United States/Europe 28.6 13.4 6.5
United States/Asia 35.2 17.3 8.5
Europe/Asia 6.7 3.9 2.0
Source: Authors’ calculations using model described in the text.
a. In the global rebalancing scenario all regions’ current account balances go to zero. See table 3 for definitions of exchange
rates and other parameter assumptions.
b. Proxied by varying elasticities of substitution: moderate, θ = 1, η = 2; gradual, θ = 2, η = 4; very gradual, θ = 4, η = 8.
c. Same as the baseline scenario reported in the first column of table 3.

by θ = 2 and η = 4), which we loosely take to capture a five- to seven-year

adjustment horizon, the bilateral exchange rate changes involving the dol-
lar are only about half as big as in our central global rebalancing scenario.
For a “very gradual” unwinding (which we take to occur over ten to twelve
years, with θ = 4 and η = 8), the same real exchange rate changes are less
than a quarter as large as in the central scenario.

Sticky Prices
Factor mobility kicks in to smooth current account adjustment if the
adjustment is slow and relatively well anticipated. If, on the other hand,
current account imbalances have to close up very quickly (say, because of
a collapse in U.S. housing prices), the bias in our estimates would point in
the other direction. Nominal rigidities in prices would then play a large
role, and actual exchange rate movements would likely be two or more
times as large as in our central scenario, for several reasons.44
For one thing, our model assumes that the law of one price holds for
traded goods, whereas in fact at most half of an exchange rate adjustment
typically passes through to traded goods prices even after one year.45 Thus,
in order to balance supply and demand for the different categories of goods

44. See the discussion in Obstfeld and Rogoff (2000a).

45. P. Goldberg and Knetter (1997); Campa and L. Goldberg (2002). For recent evi-
dence suggesting a substantial decline in pass-through to U.S. import prices, see Marazzi
and others (2005).
Maurice Obstfeld and Kenneth S. Rogoff 109

while maintaining full employment, central banks would have to allow

much larger exchange rate movements—possibly double those suggested
by the model. These larger movements would be “overshoots” in the sense
that they would unwind over time as domestic prices adjust.
The nominal prices of nontraded goods are typically even stickier than
those of traded goods; this further amplifies the overshooting effect. In
general, both sticky prices and slow factor mobility point toward the like-
lihood that a slow unwinding of the U.S. current account deficit will lead
to smaller changes in real exchange rates than would a relatively abrupt
correction.

Rising U.S. Interest Rates and the Dollar

Another qualification to our results is that our model does not account
for financial factors, and in particular for the possibility of temporarily
high real interest rates in the United States muting the dollar’s decline.
Using the Federal Reserve’s macroeconomic model, David Reifschneider,
Robert Tetlow, and John Williams estimate that a 1-percentage-point rise
in the federal funds rate (presumably unmatched by the rest of the world)
leads to a 2.2 percent appreciation of the dollar after one year, and a 4.9 per-
cent appreciation after two years.46 Therefore the fact that, over the past
year, U.S. short-term interest rates have been rising relative to Europe’s is
a countervailing consideration to those discussed above (although our cal-
culations suggest that it is likely to be far less important quantitatively). In
addition, Europe and Asia can always choose to lower their interest rates
to further mute the dollar’s decline. Of course, interest rate policy can
only affect the dollar’s real value temporarily, and so long-term global
rebalancing will still require a combination of real exchange rate adjust-
ment and factor reallocation across sectors.

The Fundamental Unpredictability of Exchange Rates

Our model suggests that the gaping U.S. current account deficit is a
very large negative factor in assessing the future prospects of the dollar. It

46. Reifschneider, Tetlow, and Williams (1999). A back-of-the-envelope calculation

based on the Dornbusch overshooting model (Dornbusch, 1976) yields a similar result.
 110 Brookings Papers on Economic Activity, 1:2005

is well known, however, that it is extremely difficult to explain exchange

rate swings between major currencies, much less forecast them, at least at
horizons up to eighteen months.47 Although a number of small qualifica-
tions must be made to this result,48 it remains broadly true. How, then, can
one be concerned about a medium-term dollar decline if a rise is equally
likely? There are two broad answers to this question. First, even the most
cheery U.S. current account optimist would have to concede that an abrupt
reversal is a potential risk, particularly while federal government deficits
remain less than fully tamed. Reversal need not result from what Guillermo
Calvo, in the context of emerging markets, has called a “sudden stop” of
capital inflows;49 as we have noted, it could follow, for example, from a
rise in U.S. saving due to a purely domestic asset price collapse. Our cal-
ibrations are useful in laying out the exchange rate consequences and in
illuminating how the burden of adjustment might be shared among the
major economies.
Second, and more fundamentally, there is some evidence that nonlin-
earities are also important, so that, when exchange rates are particularly
far out of line with one or more fundamentals, some predictability emerges.
Obstfeld and Alan Taylor, for example, argue that convergence to pur-
chasing power parity is much more important quantitatively when a cur-
rency is relatively heavily over- or undervalued compared with its long-term
real exchange rate.50 Gourinchas and Rey argue that, contrary to the canon-
ical Meese-Rogoff result, there is a forecastable component to trade-
weighted dollar exchange rate movements when net foreign assets or
debts are large relative to the United States’ net export base.51 Their work
supports much earlier work by Peter Hooper and John Morton suggesting
that net foreign assets may be important in explaining dollar move-
ments.52 As we argued in the introduction, the U.S. current account deficit
today is so large and unprecedented that it is difficult to project its future
path and the consequences thereof simply by extrapolating from past data.

47. Meese and Rogoff (1983).

48. See the survey in Frankel and Rose (1995), for example.
49. Calvo (1998).
50. Obstfeld and Taylor (1997).
51. Gourinchas and Rey (2005a).
52. Hooper and Morton (1982).
Maurice Obstfeld and Kenneth S. Rogoff 111

Conclusions

We have developed a simple stylized model that can be used to cali-

brate exchange rate changes in response to various scenarios under which
the U.S. current account deficit might be reduced from its unprecedented
current level. Aside from its quantitative predictions, the model yields a
number of important qualitative insights.
First, Asia’s greater openness to trade implies that the requisite exchange
rate adjustments for that region are not all that much greater than Europe’s.
This appears true despite the fact that Asia starts from a much larger current
account surplus than Europe.
Second, we find that, if Asia tries to stick to its dollar peg in the face of,
say, a rise in the U.S. saving rate that closes up the U.S. current account gap
even partly, Asia will actually have to run significantly larger surpluses
than it does now. Europe would bear the brunt of this policy, ending up with
a current account deficit even larger than that of the United States today,
while at the same time suffering a huge loss on its net foreign assets.
Third, although dollar depreciation does tend to improve the U.S. net
foreign asset position (because virtually all of its gross foreign liabilities,
but less than half of its gross foreign assets, are denominated in dollars), this
effect only slightly mitigates the requisite exchange rate change. Valuation
effects will not rescue the United States from a huge trade balance adjust-
ment. Indeed, if relative interest rates on U.S. short-term debt rise even
moderately during the adjustment process, this adverse effect could easily
cancel out any gain due to valuation effects.
Fourth, our model suggests that the need for deficit countries to shift
demand toward nontraded goods (and for surplus countries to shift demand
away from them) is roughly twice as important quantitatively as the much
more commonly stressed terms-of-trade channel (which involves substi-
tution between the traded goods produced by different countries). The
importance of the terms of trade would be greater with lower international
trade elasticities than we have assumed, or with a greater degree of home
bias in consumption.
We have only scratched the surface of the possible questions that can be
asked within our framework. To that end, we have tried to make our approach
as transparent as possible so that other researchers can easily investigate
alternative scenarios using the model. Clearly, it would be interesting to
extend the model in many dimensions, in particular to allow for sticky
 112 Brookings Papers on Economic Activity, 1:2005

prices and for dynamic adjustments, such as factor movement across sec-
tors. It would also be interesting to extend the framework to allow for
more regions of the world economy, for example, oil producers, non-Asian
emerging markets, and Asian subregions. Nonetheless, in a literature that
is often long on polemics and short on analysis, we hope it is useful to
have a concrete model on which to base policy evaluation.

APPENDIX A

Equilibrium Prices, Revaluation Effects, and Interest

Rate Effects

Equilibrium Prices

Here we show how real exchange rates depend on equilibrium relative

prices, and we spell out the relevant equilibrium conditions for our three-
region world economy. By definition, real exchange rates depend on rela-
tive international prices for both traded and nontraded goods. We take up
relative traded goods prices first.
As the text noted, notwithstanding the law of one price, the assumed
internationally asymmetric preferences over tradables permit relative
regional price indexes for tradable consumption to vary over time. Instead
of being fixed at unity, these ratios are given in our model by

[ατ1U−,Eη + (β − α ) + (1 − β ) τ1U−,ηA ]
1
1− η
PTE
( A1) =
[α + (β − α ) τU1−,Eη + (1 − β ) τU1−,ηA ]
1
PTU 1− η

 1− η  1 − δ   1 − δ  1− η  1− η

PTA δτU , A +  2  +  2  τU ,E 
=
[α + (β − α ) τ1U−,Eη + (1 − β ) τ1U−,ηA ]
1
PTU 1− η

 1− η  1 − δ   1 − δ  1− η  1− η

PTA δτU , A +  2  +  2  τU ,E 
= .
[ατ1U−,Eη + (β − α ) + (1 − β ) τ1U−,ηA ]
1
PTE 1− η
Maurice Obstfeld and Kenneth S. Rogoff 113

Thus shifts in interregional real exchange rates q reflect both shifts in

the relative prices of traded and nontraded goods and shifts in the relative
prices of exports and imports:
1

 γ + (1 − γ ) ( PNE PTE ) 
1− θ 1− θ

= TU × 
PE
( A2) qU ,E 1
PT  γ + (1 − γ ) ( PNU PTU ) 
1− θ 1− θ

[ατ + (β − α ) + (1 − β ) τ ]
1
1− η 1− η 1− η

= U ,E U ,A

[α + (β − α ) τ + (1 − β ) τ ]
1
1− η 1− η 1− η
U ,E U ,A

 γ + (1 − γ ) ( PNE PTE ) 
1− θ 1− θ

× 1

 γ + (1 − γ ) ( PNU PTU ) 
1− θ 1− θ

 γ + (1 − γ ) ( PNA PTA ) 
1− θ 1− θ

= TU × 
PA
qU , A 1
PT  γ + (1 − γ ) ( PNU PTU ) 
1− θ 1− θ

 1− η  1 − δ   1 − δ  1− η  1− η

δτU , A +  2  +  2  τU ,E 
= 
[α + (β − α ) τ1U−,Eη + (1 − β ) τ1U−,ηA ]
1
1− η

 γ + (1 − γ ) ( PNA PTA ) 
1− θ 1− θ

× 1 .
 γ + (1 − γ ) ( PNU PTU ) 
1− θ 1−θ

Having defined relative price indexes, one can easily derive global market-
clearing conditions for each region’s tradable output, again using very
standard techniques for constant elasticity of substitution models such as
the one we have here.53 For real U.S. tradable goods output, the market-
clearing condition is given by
−η −θ −η −θ
 P   PU   P   PE 
( A3) YTU = γ α  UU   TU  C U + γ (β − α )  UE   TE  C E
 PT   PC   PT   PC 
−η −θ
 1 − δ   PU   PT A

+ γ CA,
2   PTA   PCA 

53. As illustrated, for example, in Obstfeld and Rogoff (1996).

 114 Brookings Papers on Economic Activity, 1:2005

and that for real European traded goods output is given by

−η −θ −η −θ
 P   PE   P   PU 
( A4) YTE = γ α  EE   TE  C E + γ (β − α )  EU   TU  C U
 PT   PC   PT   PC 
−η −θ
 1 − δ   PE   PTA 
+ γ CA.
2   PTA   PCA 

Walras’ Law implies that the condition for Asian traded goods equilib-
rium is superfluous, given the two others. One can similarly derive the
market-clearing condition for U.S. nontraded goods as
−θ
 PU 
( A5) Y = (1 − γ )  NU  C U
U
N
 PC 

(which depends, of course, only on U.S. demand), as well as the two cor-
responding conditions for European and Asian nontraded goods.
We take output endowments as given, and we then use the market-
equilibrium conditions just stated to solve for relative prices as functions
of current account balances and initial net foreign asset positions. (In our
simulations we allow for currency revaluation effects on foreign assets and
liabilities, and for the feedback to trade balances needed to sustain any
given constellation of current accounts.)
To proceed, we first rewrite the equilibrium condition for the U.S.
export good’s market as
−η −η −η
P  P   1 − δ   PU 
( A6) YTU = α  UU  CTU + (β − α )  UE  CTE +  CTA ,
 PT   PT  2   PTA 

or, in nominal terms, as

1− η 1− η
P  P 
( A7) PU Y = α  UU 
U
P C + (β − α )  UE 
U U
PTE CTE
T
 PT  T T
 PT 
1− η
 1 − δ   PU 
+ PTACTA .
2   P A  T

If trade were balanced and international debts zero, then, of course, the
value of U.S. traded goods consumption would have to equal that of U.S.
Maurice Obstfeld and Kenneth S. Rogoff 115

traded goods production. Here we want to allow for international debt as

well as for trade and current account imbalances (which are the same in
the model except for net factor payments). The U.S. current account sur-
plus in dollars is given by
(A8) CAU = PU YTU + rF U − PTU CTU ,

where FU is the stock of U.S. net foreign assets (in dollars) and r is the
nominal (dollar) rate of interest. Similarly, for Europe (and again measuring
in dollars),
(A9) CA E = PE YTE + rF E − PTE CTE .

In the aggregate, of course (in theory if not in the actual data),

( A10) CAU + CAE + CA A = 0.

Similarly,

( A11) F U + F E + F A = 0.
Thus,

( A12) CA A = − (CAU + CAE ) = PAYTA − r ( F U + F E ) − PTACTA .

In this framework one can consider the effects of a variety of shocks

that change the current nexus of global current account imbalances into
one where, say, CAU = 0. (Other external balance benchmarks can be ana-
lyzed just as easily.)
To do so, we use the above current account equations (and the implied
trade balances) to substitute for dollar values of consumption of traded
goods in the goods-market equilibrium conditions. The results are
1− η
P 
( A13) PU Y = α  UU 
U
(P Y U
+ rF U − CAU )
T
 PT  U T

1− η
P 
+ (β − α )  UE  (P Y E
+ rF E − CAE )
 PT  E T

1− η
 1 − δ   PU 
+ [P Y A
− r ( F U + F E ) + CAU + CAE ]
2   PTA  A T
 116 Brookings Papers on Economic Activity, 1:2005

1− η
P 
PEYTE = α  EE  (P Y E
+ rF E − CAE )
 PT  E T

1− η
P 
+ (β − α )  EU  (P Y U
+ rF U − CAU )
 PT  U T

1− η
 1 − δ   PE 
+ [P Y A
− r ( F U + F E ) + CAU + CAE ].
2   P A  T
A T

Critically, current account imbalances also spill over into relative prices
for nontraded goods, to a degree that depends on the elasticity of substitution
between traded and nontraded goods. For the three nontraded goods mar-
kets, one can show that
1− θ
1 − γ  PNU 
( A14) PNU YNU = PTU CTU
γ  PTU 
1− θ
1 − γ  PNU 
= (P Y U
+ rF U − CAU )
γ  PTU  U T

1− θ
1 − γ  PNE 
PY =
E E
(P Y E
+ rF E − CAE )
N
γ  PTE 
N E T

1− θ
1 − γ  PNA 
PY =
A A
[P Y A
− r ( F U + F E ) + CAU + CAE ].
N
γ  PTA 
N A T

Revaluation of Gross Asset Stocks through Exchange Rate Changes

A key variable in the simulation analysis is f i, which is the ratio of net

foreign assets (in dollars), Fi, divided by the dollar traded goods income
of the United States, PUY TU. In reality, a country’s gross assets and liabili-
ties are often denominated in different currencies, so that focusing only
on the net position misses important revaluation effects that can occur as
the exchange rate changes. Here we show how we have modified our sim-
ulation analysis to take into account both the normalization of dollar net
foreign assets and the revaluation effects of exchange rate changes.54

54. Details can be found online at www.economics.harvard.edu/faculty/rogoff/papers/

BPEA2005.pdf.
Maurice Obstfeld and Kenneth S. Rogoff 117

Let Hi equal the gross assets of country i and Li its gross liabilities,
measured in dollars. Then
(A15) F i = H i − Li
and
H i − Li
( A16) fi = .
PU YTU
One can show that, under a monetary policy that targets the GDP deflator,
γ −1
 PU 
[α + (β − α ) τ + (1 − β ) τ1U−, ηA ] .
γ −1

(A17) PU =  NU  1− η 1− η

 PT  U ,E

The first step is to substitute this formula for PU into the denominators
of f U, f E, and f A. The second step is to consider how exchange rate
changes affect the numerators.
Let ωij be the share of region i gross foreign assets denominated in the
currency of region j, j = U, E, A, where the European and (especially) the
Asian regional currencies are composites. Similarly, define the portfolio
currency shares λij on the liability side. We will assume that central banks
target GDP deflators and that EU, j denotes the (nominal) dollar price of
currency j ( j = E, A) under the monetary rule. Then, after a change in
exchange rates, the new dollar values of net foreign assets (with values
after the change denoted by primes) are
 E´ − EU ,E  U U
( A18) F U´ = F U +  U ,E (ω H − λUE LU )
 EU ,E  E

 E´ − EU , A  U U
+  U ,A (ω H − λUA LU )
 EU , A  A

 E´ − EU ,E  E E
F E´ = F E +  U ,E (ω H − λ EE LE )
 EU ,E  E

 E´ − EU , A  E E
+  U ,A (ω H − λ EA LE )
 EU , A  A

 E´ − EU ,E  A A
F A´ = F A +  U ,E (ω H − λ EA LA )
 EU ,E  E

 E´ − EU , A  A A
+  U ,A (ω H − λ AA LA ) .
 EU , A  A
 118 Brookings Papers on Economic Activity, 1:2005

Note that the following two constraints must hold in a closed system:
( A19) ω UE H U + ω EE H E + ω EA H A = λUE LU + λ EE LE + λ EA LA
ω UA H U + ω EA H E + ω AA H A = λUA LU + λ EA LE + λ AA LA .

So we can eliminate the European asset shares by writing the preceding as

post-change net asset values:
 E´ − EU ,E  U U
( A20) F U´ = F U +  U ,E (ω H − λUE LU )
 EU ,E  E

 E´ − EU , A  U U
+  U ,A (ω H − λUA LU )
 EU , A  A

 E´ − EU ,E  U U
F E´ = F E +  U ,E ( λ L + λ EA LA − ωUE H U − ω EA H A )
 EU ,E  E

 E´ − EU , A  U U
+  U ,A ( λ L + λ AA LA − ωUA H U − ω AA H A )
 EU , A  A

 E´ − EU ,E  A A
F A´ = F A +  U ,E (ω H − λ EA LA )
 EU ,E  E

 E´ − EU , A  A A
+  U ,A (ω H − λ AA LA ) .
 EU , A  A

We also know that

(A21) H U + H E + H A = LU + LE + LA .
For our numerical findings we must posit estimated values for nominal
assets and liabilities. Given the well-known measurement problems, any
numbers are bound to be loose approximations at best. For the United
States, the numbers we use are for end-2003 (from the 2005 Economic
Report of the President) and show foreign-owned assets in the United
States to be $10.5 trillion and U.S.-owned assets abroad to be $7.9 tril-
lion. We take the current values to be $11 trillion and $8.25 trillion,
respectively, for purposes of our simulations. To a first approximation,
essentially all U.S. foreign liabilities are denominated in dollars, but only
about 40 percent of U.S. foreign assets are. (In principle, foreign assets
such as stocks and land are real, but in practice the dollar returns on these
Maurice Obstfeld and Kenneth S. Rogoff 119

assets are highly correlated with dollar exchange rate movements.) Of the
remaining 60 percent, we take 41 percent to be in European currencies
and 19 percent in Asian currencies. Following Tille (2004), and including
Canada, the United Kingdom, and Switzerland in region E, the United
States does have a very small share of its liabilities in foreign currencies.
The exact portfolio weights that we assume for the United States are
( A22) ω UE = 0.405, ω UA = 0.193,
λUE = 0.03, λUA = 0.006.
Drawing on the work of Lane and Milesi-Ferretti (but taking into
account the adding-up constraints that need to hold in our theoretical model),
we take Asia’s assets to be $11 trillion and its liabilities to be $8.25 trillion.55
As for portfolio shares, on the asset side, data from the International
Monetary Fund’s 2001 Coordinated Portfolio Investment Survey suggest
that most Asian countries hold predominantly U.S. dollars (and some yen),
but that Japan’s foreign assets are more evenly balanced between dollar and
euro holdings. If we assume that Japan owns about 40 percent of the
region’s gross foreign assets, we have the following approximation:

( A23) ω EA = 0.2, ω AA = 0.

On the liabilities side, Japan borrows in yen, but the other Asian
economies have equity liabilities (including foreign direct investment) in
local currencies, and extraregional debt liabilities predominantly in dollars
and euros (or sterling). We assume that
( A24) λ EA = 0.33, λ AA = 0.33.

We again base our portfolio estimates for the E zone in our model on
the latest data from Lane and Milesi-Ferretti, which indicate that assets
and liabilities at the end of 2003 were both approximately $11 trillion.
Thus we take HE = LE = $11 trillion. Most of greater Europe’s liabilities
are in domestic currencies; here we assume the share is 80 percent. We
take the remaining 20 percent to be entirely in U.S. dollars. On the asset
side, however, we derive from equation A19 that 32 percent of Europe’s
holdings are in dollar assets, and 11 percent in assets denominated in

55. Lane and Milesi-Ferretti (forthcoming).

 120 Brookings Papers on Economic Activity, 1:2005

Asian currencies, with the remaining 57 percent in assets denominated in

European currencies.56
In our simulations we take PUY TU = $(11/4) trillion, based on Obstfeld
and Rogoff (2000b), who argue that roughly one-quarter of U.S. GDP
may be regarded as traded.
Given our assumptions on each region’s gross assets and liabilities
and their currencies of denomination, our analysis will also tell us how
net foreign assets change across various scenarios for the current account
and the exchange rate, as well as allow for the feedback effect on interest
payments. We will see that, given the large size of gross stocks, large
changes in exchange rates can translate into large changes in net foreign
asset positions. Indeed, for many short-run and medium-run issues,
knowing the gross asset and liability positions is at least as important
as understanding the net positions. This conclusion is very much in line
with the empirical findings of Gourinchas and Rey (2005a) for the
United States.

Effects of Changing Interest Rates

It seems plausible that, in the process of U.S. current account adjust-

ment, global interest rates will shift. Such changes could come about sim-
ply as a result of the reequilibration of the global capital market, or they
could also reflect a shift in the portfolio preferences of foreign investors
such that, given the exchange rate of the dollar, higher dollar interest
rates are necessary to persuade them to maintain their existing dollar-
denominated portfolio shares. We adopt the latter perspective, allowing
the interest rate on U.S. short-term debt liabilities to rise as the dollar
adjusts, without a corresponding increase in the earnings on U.S. foreign
assets. Capital market shifts of this nature are likely to be quantitatively
more important for the dollar than more generalized, synchronized
increases in world interest rates (although the United States, as a debtor,
would naturally lose while its creditors would gain).
To illustrate this channel, we first, for simplicity, abstract from the effects
of nominal exchange rate changes on asset stocks (for the purpose of our

56. The European position assumptions are not needed to implement equation A20, but
they are necessary for assessing the effects of interest rate changes below.
Maurice Obstfeld and Kenneth S. Rogoff 121

simulations, this case is only a computation check). We focus on the sce-

nario under which, as the United States adjusts, it faces a sharp increase in
its borrowing rates. Thus there are two interest rates in the world econ-
omy: the rate rU that the United States pays on its liabilities, and the rate
rW > rU that all other countries pay on their liabilities and that all coun-
tries, including the United States, earn on assets outside the United States.
We focus on the implications of rU rising when the United States adjusts;
the increase in rU may itself have an effect on U.S. adjustment, although
that possibility does not affect our calculation.
There is also a long-run versus short-run distinction: in the short run only
U.S. short-term liabilities will pay higher interest (as these are rolled over).
According to U.S. Treasury data for September 2004 (from www.treas.
gov/tic/debta904.html), U.S. short-term liabilities were about 30 percent
of total liabilities (and thus about 30 percent of U.S. GDP). If the United
States were required to pay, for example, 200 basis points more on this
liability base, the result would be an additional drain of about 0.02 × 0.3 =
0.6 percent of total GDP.
Let ω∼ i represent the share of country i gross foreign assets invested in
j
country j.
To make the previous modeling consistent, we replace rFi everywhere
(for the United States, Europe, and Asia, respectively) by
( A25) r W H U − r U LU
[ω
r
E
U
U
+ (1 − ω
UE ) r W ] H E − r W LE
[ω
r
A
U
U
+ (1 − ω
UA ) r W ] H A − r W LA .

From estimates described in the last subsection, we have the dollar values
of Hi and Li. Asian currency shares probably exceed the Asian country
shares, because of Asian claims on offshore Eurodollars; we might assume
∼ A = 0.6. Since total U.S. liabilities equal the claims on the United
that ω U
States of Europe and Asia,
( A26) ω
UE H E + ω
UA H A = LU ,
and so, with HE, HA, and LU each equal to $11 trillion, we must have
∼ E = 0.4.
ω U
We now turn to the calibration of interest rates (or, rather, nominal rates
of return on asset and liability portfolios). We know that, for the United
States currently, rWHU − rULU ≈ 0. Since, also, HU/LU ≈ 0.75, rU/rW ≈ 0.75.
 122 Brookings Papers on Economic Activity, 1:2005

So we take rU = 3.75 percent initially,57 but we maintain the earlier baseline

assumption that rW = 5 percent. We ultimately wish to consider alternative
increases in rU, for example, of 125 basis points or more. These possibili-
ties range from a scenario in which the United States simply loses its priv-
ilege of borrowing at a favorable rate, to some in which there is an element
of loss of confidence in U.S. solvency absent ongoing dollar depreciation.
We will also assume that only the interest rate on short-term liabilities rises
in the short run. Suppose the share σ of short-term liabilities in total U.S.
foreign liabilities is 30 percent, or σ = 0.3. Then the investment income
account of the United States and the other two regions would change as
follows:

( A27) r W H U − r U LU → r W H U − ( r U + σ∆r U ) LU
[ω
E
U
r U + (1 − ω
UE ) r W ] H E − r W LE →
[ω
(r + σ∆r
E
U
U U ) + (1 − ω
UE ) r W ] H E − r W LE

[ω
A
U
rU + (1 − ω
) r ] H
A
U
W A
− r W LA →
[ω
(r + σ∆rr
A
U
) + (1 − ω
UA ) r W ] H A − r W LA .
U U

∼E + ω
The last two changes assume that, empirically, ω ∼ A ≈ 1 and that Europe
U U
and Asia hold equal proportions of short-term U.S. liabilities.
One might also consider a formulation where ∆rU = f (∆CAU), f ′ > 0. In
this case adjustment could be quite painful if the f function is too rapidly
increasing, LU is too big, or σ is too big (or any combination of these three).
We leave this possibility for future research.

Synthesis of Interest Rate Changes and Asset Revaluations

We are now ready to illustrate the techniques used to calculate the

results in the third column in table 8, in which asset revaluations and
interest rate changes occur simultaneously and interactively. We proceed
as in the last section but add the following equations:

57. This number is in line with the estimate given above of the excess return of U.S.
foreign assets over U.S. liabilities to foreigners.
Maurice Obstfeld and Kenneth S. Rogoff 123

 E ´ − EU ,E  U U  EU´, A − EU , A  U U
( A28) H U´ = H U +  U ,E ω H + ω H
 EU ,E  E  EU , A  A

 E ´ − EU ,E  E E  EU´, A − EU , A  E E
H E´ = H E +  U ,E ω H + ω H
 EU ,E  E  EU , A  A

 E ´ − EU ,E  A A  EU´, A − EU , A  A A
H A´ = H A +  U ,E ω H + ω H
 EU ,E  E  EU , A  A

and

 E´ − EU ,E  U U  EU´ , A − EU , A  U U
( A29) LU´ = LU +  U ,E λ L + λ L
 EU ,E  E  EU , A  A
 E´ − EU ,E  E E  EU´ , A − EU , A  E E
LE´ = LE +  U ,E λ L + λ L
 EU ,E  E  EU , A  A
 E´ − EU ,E  A A  EU´ , A − EU , A  A A
LA´ = LA +  U ,E λ L + λ L.
 EU ,E  E  EU , A  A

These equations, rather than the equations for net positions used in the
simpler revaluation exercise in which interest rates do not change, become
necessary because assets and liabilities can now pay different rates of
interest and therefore must be tracked separately.
Comments and
Discussion

Richard N. Cooper: This paper by Maurice Obstfeld and Kenneth Rogoff

is very much a “what if” exercise. What if demand behaves according to
constant elasticity of substitution functions? What if consumption is fixed,
apart from terms-of-trade effects? What if the U.S. current account deficit
is eliminated (or, in one of the authors’ simulations, halved) by an appro-
priate increase in the U.S. saving rate? Then we learn from the authors’
model what the implied changes in exchange rates and the terms of trade
must be.
The authors’ calibration is necessarily arbitrary, but it seems reasonable.
A major contribution of their model is to provide a general equilibrium
framework that includes two stylized regions outside the United States. It
therefore permits an exploration of differing effects by region, and that seems
very useful. The model also includes asset revaluation effects and not just
trade effects.
The model assumes flexible prices. As the authors note, it probably under-
states the exchange rate changes that would be required in a sticky-price
regime. The authors also usefully note the potentially ambiguous impact of
faster European or Japanese economic growth, which many people see as a
potential partial solution to the correction of the U.S. current account deficit.
The source of the growth makes a difference, and one should not assume that
more-rapid growth abroad will ease the problem. Productivity increases in
traded goods, which is where such increases have typically occurred, espe-
cially in Japan, could aggravate rather than mitigate the imbalances.
No doubt it is interesting to see how large are the exchange rate changes
required to close the U.S. current account deficit, but all the adjustment here
is done through prices, including the asset revaluation effect. Whether
that is helpful to policy is not at all clear to me. There is no explicit treat-
ment of output or employment in the model. The simulations are driven
124
Maurice Obstfeld and Kenneth S. Rogoff 125

by a postulated increase in U.S. saving and a corresponding decline in or,

in most of the simulations, elimination of the U.S. current account deficit,
so that presumably U.S. output is unaffected.
The paper’s simulations are, however, unclear about what is happening
to saving in the other two regions of the world as the U.S. external deficit
is reduced. Their current account surpluses at the outset reflect excessive
saving in those two regions, and they disappear in the simulation in which
all three current accounts go to zero. But how do they disappear?
Consumption is assumed to be fixed, except for the terms-of-trade effect.
The fall in saving in the paper’s simulated Europe and Asia therefore
implies a corresponding fall in output and income in order to get these
results. But a decline in output will surely affect employment, hence income,
hence consumption and saving, raising the question of how the initial level
of consumption is sustained in Asia and Europe. That in turn leaves me
wondering whether the paper provides any useful lessons for addressing the
issue of global imbalances in, say, the coming decade.
Let me offer, in sketchy terms, my own view of the issue. The discus-
sion of the U.S. current account deficit has focused largely on how much
adjustment must occur in the United States. The authors’ model is properly
a general equilibrium model, and so it includes the rest of the world. I will
focus on adjustment in the rest of the world.
In 2004 the large current account surpluses in the world were in Japan,
at $172 billion, and in Germany and the Netherlands (which can be consid-
ered a satellite of the German economy) at $116 billion. Thus these three
countries together account for nearly half the U.S. current account deficit.
Russia and China together add another $130 billion. These are where the big
numbers are. Adding up all of the rest of East Asia accounts for another
$110 billion, and OPEC another $100 billion. Then there is the statistical
discrepancy, which has grown to about $200 billion.
There is, as always in ex post accounting, a problem of attribution, but
if we stipulate that the U.S. current account deficit is to be eliminated, we
need to ask where the impact will fall in the rest of the world, and it has to
fall largely where the big surpluses are. (The present surpluses of emerging
economies could also shrink, or their deficits grow, but that would require
a willingness on the part of savers around the world to invest in those
countries on the required scale, which is not a given. If all the adjustment
occurred there, their current account deficits would have to rise far in excess
of generally accepted levels.)
 126 Brookings Papers on Economic Activity, 1:2005

Like the authors, I will put oil to one side. The increase in oil prices over
the past two years has added roughly $100 billion to the U.S. current account
deficit. I assume that, one way or another, either through a decline in oil
prices or through an increase in absorption by the oil-exporting countries,
their surpluses will decline significantly in the next few years. Instead I
will focus on Germany and Japan, which, again, are where the really big
surpluses are, and then I will comment on China.
Germany and Japan are rapidly aging, high-saving societies with limited
domestic investment. Saving rates have declined in Japan, but saving in the
corporate sector remains quite high. What has fallen in Japan is investment,
which remains low even after a revival in the last year.
A big absorber of capital in rich countries is the residential sector. Invest-
ing in housing does not look very attractive in rapidly aging societies, with
very low birth rates and low new household formation, which is the case
in both of these countries. If anything, Germany and Japan have a surplus
of housing in the aggregate, although it may not all be in quite the right
places. Housing construction is down essentially to replacement plus a little
bit to allow for mobility. Meanwhile rates of return on industrial investment
are low and, of course, very sensitive to what is happening to the export
sector.
I will now make some sweeping (perhaps too sweeping) national gen-
eralizations. For reasons having to do with their defeat in World War II, a
key question for the Germans and the Japanese was how to rebuild their
national self-esteem. Both countries built it on export performance. That
legacy continues sixty years later. The national psyche in both Germany
and Japan is heavily influenced by export performance. If exports are not
doing well, people feel badly about the economy and society. In my view,
that influences their saving behavior. If the economy is not performing
well, precautionary saving rises in these now-rich countries.
Given the aging of their society, as the Japanese have been saying for
some years, Japanese saving should decline and eventually become nega-
tive. That may be so, but it has been a much slower process than the life
cycle advocates forecast twenty years ago. Saving remains remarkably high
given Japan’s demographic structure, and the same is true of Germany.
That syndrome, in which German and Japanese saving is sensitive to
perceived economic performance, which in turn is remarkably sensitive to
export performance, is important when it comes to correcting the U.S.
Maurice Obstfeld and Kenneth S. Rogoff 127

current account. If, as Obstfeld and Rogoff suggest, there will be big changes
in exchange rates and big declines in the export competitiveness of key sur-
plus countries, we are likely to see an increase, not a reduction, in the
propensity to save in those countries. Whether that increase gets translated
into actual additional saving depends, of course, on what happens to income.
The conditions just described are, after all, the conditions under which a
recession could occur. An increase in the propensity to save with no obvious
vehicle for that saving leads to a fall in output and income.
In the textbooks the adjustment mechanism in this process is the interest
rate, which is assumed to reconcile ex ante differences in saving and invest-
ment. Suppose the long-term nominal interest rate is only 2 percent, as it
has been in Japan for several years, and not much higher in Germany. The
question then becomes, What sort of investment in Japan and Germany will
be stimulated by a 2 percent interest rate, given the demographics, in the
presence or even with the prospect of a significantly stronger yen and euro?
The sector most responsive to low interest rates in rich economies gener-
ally is the housing sector, not industrial investment. Firms will not invest
in increased capacity if they see poor sales prospects, no matter how low
the interest rate. Yet, for the demographic reasons already noted, demand
for housing will be limited, even at low long-term interest rates.
Hence I do not see the interest rate as being an effective adjuster here.
With a large appreciation of these surplus countries’ currencies, the adjuster
is more likely to be economic activity. Economic activity will decline, except
insofar as the authorities become so concerned about it that the Europeans
break all the rules they have imposed on themselves, through the Stability
and Growth Pact’s constraints on fiscal policy and the European Central
Bank’s primary focus on price stability, and pursue an aggressively stim-
ulative policy.
I therefore see a big problem with substantial current account adjust-
ment, mainly for Europe but also for Japan. Both already have large budget
deficits: the Japanese budget deficit is roughly 7 percent of GDP, and France,
Germany, and Italy, the core of Europe, have fiscal deficits expected to
exceed the 3 percent limit under the Pact.
Excess saving in these big rich countries manifests itself in budget
deficits and current account surpluses. Savers directly or indirectly buy
claims on their governments or claims on foreigners. In my judgment, fur-
ther reductions in the long-term interest rate are not going to produce
 128 Brookings Papers on Economic Activity, 1:2005

enough domestic investment to substitute for those two, particularly in the

face of a decline in competitiveness brought about through large appreci-
ations of the currency.
Japan and Germany are perhaps unusual because of their peculiar history
and their dependence on export performance. But I do not see how the
currency appreciations that the paper simulates will produce the changes
in saving required to eliminate, or even greatly reduce, the current account
surpluses of Asia and Europe. A decline in Asian and European output in
turn is likely to reduce the value of U.S. equity claims on those countries,
weakening and possibly even reversing the valuation effect arising from
dollar depreciation.
China is more complicated, and I will not discuss it in detail. Although I
do not subscribe to the whole of the Dooley-Garber thesis, I am sympathetic
to one of its main thrusts. China is a very-high-saving, high-investment
country. I believe there is a pent-up, latent demand in China for foreign
assets, which cannot be realized because violation of the foreign currency
rules, especially those regarding the export of resident funds, is severely
punished. China has a very weak capital market, and the central bank of
China is, in effect, making the foreign investments that the public is pre-
vented from making. Tighter fiscal policy, which might be called for at
the moment on domestic grounds, would actually increase national saving
in China. That moves in the wrong direction.
To conclude, I believe that the United States has comparative advantage
at producing marketable assets. We sell these marketable assets to the rest
of the world. As long as Americans use the proceeds of the sale of those
marketable assets productively—and that is an important qualification,
bearing on the desirability of the fiscal deficit—I do not see why that process
cannot go on indefinitely.
I am not saying the U.S. current account deficit will go on forever. My
crystal ball fades rapidly after fifteen or twenty years. But, for the next
decade, I do not see why the process whereby the United States generates
marketable assets and sells them to foreigners who are eager to buy them can-
not continue on the current scale, that is, roughly half a trillion dollars a year.
Will there be a dollar crisis? I don’t have any idea. It depends in large
measure on how markets react to debates such as the one we are having
here. Expectations in financial markets can be very fragile. Currency mar-
kets could start to run away rapidly from the dollar. My main point is that
there need not be a crisis.
Maurice Obstfeld and Kenneth S. Rogoff 129

T. N. Srinivasan:1 Had Yogi Berra, the great baseball player and savant,
been the discussant of this paper, he likely would have begun with his famous
words, “It is déjà vu all over again!” Fifteen years ago Stefan Gerlach and
Peter Petri published a collection of essays with a pompous title, The Eco-
nomics of the Dollar Cycle.2 They viewed the movements of the external
value of the dollar as cyclical: having appreciated by more than 40 percent
between late 1979 and February 1985, the dollar had then collapsed to a
new low by 1987, only to stabilize and fluctuate narrowly around the bottom
of the range experienced during the 1980s. Gerlach and Petri also made
the following astounding claim:
Unlike narrowly focused studies in a technical specialty, this book explores the
subject simultaneously from the viewpoints of exchange rate economics, empirical
trade analysis, the economics of international financial markets, and macro-
economic policy-making in the United States, Japan, Europe, and the developing
countries.3

Here we are, fifteen years later, exploring the same issues, except that the
buzzword of that day—“newly industrializing countries” or NICs—has been
replaced by another, “emerging markets.” Evsey Domar, when asked by a
graduate student at MIT why the questions in the macroeconomics exam-
ination do not change from year to year, is said to have replied, “Ah—but
the answers do!” Domar did not claim that the answers got better over
time, but one can hope that the papers in this volume will provide better
answers to the same questions covered in the Gerlach-Petri volume.
In preparing this comment, I found particularly useful the contributions
to that volume by my late colleague James Tobin, and the comments on
Tobin’s paper by his Yale student Ralph Bryant and my dear departed friend
Rudiger Dornbusch. In 1988, as now, the U.S. current account was in deficit,
albeit at a little more than 2 percent of GDP rather than 6 percent as now.
Then as now, the United States was running a fiscal deficit of around
3.7 percent of GDP, similar to today’s 4 percent.4 In 1988 nominal inter-
est rates in the United States had fallen from the dizzy Volckerian heights
of over 19 percent in the early 1980s to a low of less than 7 percent and

1. I thank Benjamin Friedman and Robert Solow for their comments.

2. Interestingly, some of the contributors to the Gerlach-Petri volume were participants
at this Brookings Panel meeting.
3. Gerlach and Petri (1990, p.2 ).
4. See the authors’ figure 4.
 130 Brookings Papers on Economic Activity, 1:2005

had again begun to rise. Nominal rates again reached a low of around 1 per-
cent (in terms of the federal funds rate) in mid-2004 and again have begun
to rise.
On the financing of the U.S. current account deficit in those days, Tobin
remarked, “We are being warned incessantly that we depend on foreigners—
mainly Japanese banks, insurance companies and pension funds—to buy
US Treasury bonds and other dollar assets. . . . Should they decide not to
buy dollar securities, we are told, [the result] would be calamitous.”5
Now, besides the Japanese, the financiers are the Asian central banks, par-
ticularly those of China, India, and Korea. (Even then, as Dornbusch noted
in his comment, “Central banks rather than private savers have been financ-
ing the US current account.”)6 But the dire warnings are being repeated.
Then, as Dornbusch put it, there was a “sharp shift in trade with the NICs.
The United States has experienced a $60 billion shift in its manufactures
trade with these countries since 1980.”7 Now China figures prominently
in U.S. and world manufacturing trade, not to mention the prominence of
India and other countries in services trade through offshoring. Related to
this shift in trade was the issue of the domestic price consequences of dol-
lar depreciation. To quote Tobin again, with any further depreciation of
the dollar, “certainly imports from Japan and Europe will be more costly
in dollars. So will imports from Asian ‘NICs’ if we induce them to let
their currencies rise against the dollar.”8 Today the inducement takes the
form of a demand by the secretary of the Treasury of the United States that
the Chinese revalue their renminbi by at least 10 percent.
At that time some held that the dollar was correctly valued, and hence
there was no need for policy-induced corrections.9 This view was ratio-
nalized in three ways. First was the J-curve story: the response of the
economy to the fall in the value of the dollar that had already taken place
had yet to manifest itself fully and would surely do so soon. Second, U.S.
current account deficits can be financed indefinitely by selling U.S. assets,
since, in the portfolio of the rest of the world, the share of U.S. assets was

5. Tobin (1990, p. 34).

6. Dornbusch (1990, p. 53).
7. Dornbusch (1990, p. 53).
8. Tobin (1990, p. 30, emphasis added).
9. In the Gerlach-Petri volume, these views are summarized by Dornbusch (1990,
pp. 53–54).
Maurice Obstfeld and Kenneth S. Rogoff 131

probably below its long-run equilibrium value. The contemporary version

of this argument holds that China (and probably other Asian countries,
including India, as well) undervalues its currency to sustain its export-led
growth, so as to provide employment to its huge stock of underemployed
labor. Until this stock is exhausted, undervaluation of its currency may
continue, with the result that any possible appreciation from capital inflows
is prevented by accumulation of reserves held in dollar-denominated
assets. This argument is forcefully put forward by Dooley, Folkerts-Landau,
and Garber.10 Third, the United States can wipe out the value of dollar-
denominated assets held by the rest of the world through inflation. Then
as now, the question was raised whether whatever transitional adjustments
(policy induced or otherwise) would inevitably take place to restore the
dollar to its long-run equilibrium would be orderly and smooth (a soft
landing), or delayed such that, when they eventually do take place, they
would be abrupt and large (a hard landing).
Brookings also held a workshop in that earlier period, in January 1987,
to discuss these issues, just as it is doing now. Then, however, it assembled
a group of multicountry-macroeconometric modelers and gave them alter-
native, but commonly specified, scenarios for the future. Starting from the
common actual history of U.S. and foreign prices and exchange rates, the
modelers (five in all) were asked to investigate the causes of the burgeon-
ing external deficit of the United States during 1980–86 and to study the
likely path of the deficit for the period 1987–91.
Bryant, in his comment on Tobin’s paper in the Gerlach-Petri volume,
reported on his updating of the conclusions of that workshop. His qualita-
tive and quantitative findings regarding policy are not that much different
from those of Obstfeld and Rogoff in the paper under discussion here,
once one adjusts for differences in initial conditions. Briefly, he found that,
first, in the short run (that is, until 1989), it was plausible to expect a sub-
stantial reduction in the U.S. external deficit.11 Second, in the medium and
long run (1990 and after), the prospects were much less encouraging, with
the improvement in the constant-price deficit leveling off after 1989 and
the current-price deficit (the deficit measured in nominal dollars) ceasing
to improve well before it reached an acceptably low level. Third, a fall in

10. Dooley, Folkerts-Landau, and Garber (2004a, 2004b).

11. Bryant (1990, pp. 42–43).
 132 Brookings Papers on Economic Activity, 1:2005

the dollar’s real exchange value can play a powerful role in reducing the
external deficit. Fourth, the deficit in nominal dollar terms was unlikely
to decline to an acceptable level ($30 billion to $40 billion, or around
0.75 percent of GDP—one-fifth of its value in 1987) without either a some-
what further depreciation of the dollar or markedly slower growth in the
United States than abroad.
Bryant’s own estimate of the real dollar depreciation needed to bring the
deficit down to an acceptable rate was between 7 percent and 15 percent;
he viewed a 20 percent depreciation as excessive. Obstfeld and Rogoff esti-
mate the depreciation in terms of the real trade-weighted exchange rate
required in order to eliminate the present U.S. deficit of about 5 percent of
GDP to be between 19 and 28 percent, in a scenario in which Asia neither
adjusts its exchange rate nor reduces its current account surplus (see their
table 4). Bringing the deficit down to 1 percent of GDP, or one-fifth of its
initial value (if one can make a linear interpolation from the authors’ esti-
mates), would call for a dollar depreciation in the range of 15 to 22 per-
cent, very close to the estimate of 20 percent, for a similar proportional
reduction of the deficit in 1987, that Bryant cited but found excessive.
Tobin’s contribution to the Gerlach-Petri volume is aptly titled, “Eight
Myths about the Dollar.” He encountered these myths regarding what
should be done to eliminate the U.S. external deficit “all too often in con-
temporary public and, yes, professional discussion.”12 In exploding the
myths, he also provided an analytical framework for thinking about poli-
cies for eliminating the deficit. Some of the myths seem to be still going
around, and, more important, Tobin’s analytical framework remains rel-
evant today. I therefore briefly summarize Tobin’s contribution in the
next section.
TOBIN’S ANALYTICAL FRAMEWORK AND THE EIGHT MYTHS. Unsurpris-
ingly, Tobin found that, “just as [Hicks’s] IS-LM [model], for all the hard
knocks it has received from theorists, remains a good general first approx-
imation, so its international application, Mundell-Fleming, has been a
good guide.”13 In the basic, two-country version of this model, there are
two goods, and each country specializes in producing and exporting part
of its output of one of the goods. There are two real assets, consisting of

12. Tobin (1990, p. 28).

13. Tobin (1990, p. 28).
Maurice Obstfeld and Kenneth S. Rogoff 133

the real money stocks of the two countries. Each country holds some of
the other country’s asset. The simplest way of incorporating a trade (or
current account) deficit in equilibrium in this model is through a capital
transfer from one country to another that allows the receiving country to
spend more on the two goods than its income, by the amount of the transfer.
With one country’s good as the numeraire, the price of the second coun-
try’s good (that is, its relative price in terms of the first country’s good) is
the real exchange rate in the model. Given the amount of the transfer in
numeraire terms, goods market equilibrium determines the real exchange
rate. The income of each country includes, besides the value of its goods
output, its asset income, which in this simple framework equals the inter-
est income on the part of its domestic real money stock held at home at its
domestic interest rate and the interest income (in numeraire terms) on the
foreign asset held by domestic residents at the foreign interest rate. Asset mar-
ket equilibrium requires that, given the portfolio choices (in which capital
transfers from one country to the other are incorporated), the demand for
each country’s asset equal its exogenous supply. With free capital mobility,
asset market equilibrium implies that the difference between the domestic
and the foreign interest rate satisfies the uncovered interest party condi-
tion: in other words, that it equals the rate of expected real depreciation.
Under perfect foresight (rational expectations) the expected rate of depre-
ciation is zero in equilibrium, so that the model solves for two prices: the
real exchange rate and the common interest rate.
Obviously, this real model cannot determine the nominal exchange rate
or any other nominal variables. Trivially, one could introduce nominal vari-
ables by viewing each country’s asset as its nominal currency stock. By
choosing units of measurement of the two goods such that the price of each
country’s good in its own currency is unity, nominal and real exchange
rates can be made to equal each other. Any other price normalization rule
will lead to a different nominal exchange rate corresponding to a given real
exchange rate, but has no consequence for the determination of equilibrium
real values.
Clearly, away from equilibrium and assuming away unwanted inventory
accumulation (for example, assuming that the two commodities are per-
ishable), the excess of expenditure over income by one country will equal
the capital transfer from the other country ex post—this is the identity by
which the current account deficit is matched by an equivalent capital inflow.
However, the excess expenditure and the corresponding capital inflows
 134 Brookings Papers on Economic Activity, 1:2005

are not equilibrium amounts. Thus, as Rachel McCulloch correctly pointed

out, from the ex post identities that national dissaving (the excess of expen-
diture over income or production) equals capital surplus in the balance of
payments, which in turn equals the current account deficit, no causal rela-
tionship among the variables can be inferred.14 Put another way, no infer-
ence about the policy or behavioral changes needed to restore equilibrium
can be drawn from ex post identities. Indeed, some of Tobin’s eight myths
illustrate this proposition very clearly.
Simply put, policies that claim to restore equilibrium in both markets
by operating only in one market cannot, in general, succeed. In other words,
both the real exchange rate and the interest rate will have to change to
restore equilibrium. Tobin’s first myth, that “eliminating the federal budget
deficit will automatically eliminate the deficit in the U.S. external current
account,” and the second myth, its corollary, that “correction of the federal
budget would solve the problem of external imbalance without further
depreciation of the dollar,”15 illustrate this proposition. I cannot resist
referring also to Tobin’s sixth myth: “depreciation will be counterproduc-
tive for the United States because it will cause recessions in Europe and
Japan and diminish their demands for US goods and services.”16 As Tobin
rightly said, even an undergraduate should be ashamed to fall for the chain
of arguments that lead to this myth, yet “it has been advanced with straight
faces by high financial officials”17—a statement that remains true today.
This palpably false reasoning follows, in effect, by reading causality from
ex post accounting identities. Myths three, four, five, and seven illustrate
other but similarly faulty reasoning. Myth eight, on the naïveté of faith in
macroeconomic policy coordination among major economies, lives on. At
every meeting of the governors of the International Monetary Fund or the
summit of the G-7 or G-8, such coordination for addressing “global
imbalances” is advocated. Tobin was right: attempts at coordination on a
possibly wrong policy can be very harmful, and such an event cannot be
ruled out as unlikely, given the differences among analysts as to what
constitutes the right policy in the first place.

14. McCulloch (1990).

15. Tobin (1990, pp. 28–29).
16. Tobin (1990, p. 33).
17. Tobin (1990, p. 33).
Maurice Obstfeld and Kenneth S. Rogoff 135

THE OBSTFELD-ROGOFF CALIBRATED MODEL AND ITS SIMULATIONS. The

Obstfeld-Rogoff model is, in some aspects, a more elaborate version of
the basic real model described above, and less elaborate in others. Each
country produces two goods, an internationally traded and a nontraded good,
instead of one in the basic model; each country consumes three goods—
the traded good that it produces, that produced by the other country, and,
of course, its own nontraded good—instead of two. However, theirs is a
static endowment model: each country has an exogenously determined
endowment of its two goods; there is no production (and there are no factors
of production) or investment, only consumption and trade. There are no
factor or asset markets. Appearances to the contrary, in the Obstfeld-
Rogoff model nominal exchange rates, the currency denomination of assets
held by U.S. and foreign residents, their valuation, and interest rate effects
are all add-ons: they do not form part of the behavioral specification of the
model. The model determines the equilibrium real exchange rates from
which Obstfeld and Rogoff then derive nominal exchange rates under alter-
native assumptions of the policy objectives of the central bank: to stabilize
the CPI deflator, the GDP deflator, or a bilateral exchange rate. These
assumptions, being unrelated to the behavioral variables of the model,
do not influence its equilibrium determination. Investment decisions, inter-
est rates, and portfolio choices obviously do not arise in an endowments
model. Obstfeld and Rogoff add the return from net asset holdings at an
exogenously specified interest rate to the value of endowment on the income
side. Although asset valuation effects and interest rates do influence the
equilibrium through their income effects, they are still add-ons, since
gross values of assets and liabilities, portfolio weights, and interest rates
are not endogenously determined by the model, and the fact that gross
values are in nominal terms does not matter, since nominal exchange rates
are mechanically linked to real rates, given the assumed behavior of the
central bank.
When I say that add-on assumptions are not part of the behavioral struc-
ture model, I do not mean to imply that the results of the authors’ numer-
ical simulations, such as the magnitudes for nominal exchange rates, are not
plausible. They might well be, but their plausibility or otherwise cannot be
inferred from the plausibility of the assumptions alone. I will not, therefore,
comment on the paper’s numerical results.
Obstfeld and Rogoff simulate the impact of specified changes in the inter-
national pattern of external imbalances, those changes being the consequence
 136 Brookings Papers on Economic Activity, 1:2005

of unspecified shocks to demand. In one scenario these shocks are assumed

to bring down initial current account deficits and surpluses in all three
regions to zero. In another the U.S. current account is set at zero, Asia pegs
its nominal exchange rate to the dollar, and so on. The model then deter-
mines the comparative-static changes in equilibrium real exchange rates and
other endogenous variables associated with the change in external imbal-
ances. Since, in an endowment model, the only way to influence international
balances is by affecting demand, analysis of changes in international balances
brought about through shocks to supply, investment, or saving behavior is
ruled out.
Policies are not explicitly modeled—they are whatever is needed to
induce the unspecified shocks to demand. For example, all fiscal policy
combinations that, through shocks to consumption, result in the specified
changes in external imbalances are equivalent from the perspective of the
model. The focus of the analysis is entirely on the real exchange rate impli-
cations of the specified changes in the pattern of initial external imbalances,
and not on the policies that are behind those changes in imbalances. Those
policy changes are outside the model and cannot be evaluated through the
model. Thus the various policy proposals currently being made, including
a policy of neglect, benign or otherwise, of the initial imbalances, cannot
be evaluated. Also, the policy-relevant question of whether the prevailing
U.S. current account deficit of 6 percent of GDP is indefinitely sustainable
cannot be meaningfully posed, let alone answered, by the model. The
implications for U.S. interest rates of a shift away from dollar-denominated
assets in the portfolios of foreigners, including central banks, or a reduc-
tion in saving propensities abroad, or a rise in saving propensities in the
United States, cannot be examined. Such questions as whether the end to
global imbalances will come smoothly, predictably, and at a modest cost,
or abruptly, unexpectedly, and at a heavy cost, cannot be analyzed either.
The analytics of the model are easily illustrated in a slightly simpler
version in which there are two countries, home and foreign, each producing
two traded goods. Each traded good produced by one country is a perfect
substitute for the corresponding good produced by the other country. This
simplification, of course, rules out home bias, as in the Obstfeld-Rogoff
model, in the consumption of traded goods, and it rules out having more
than one real exchange rate. However, it goes beyond the Obstfeld-Rogoff
model by allowing a production (supply) response to changes in relative
prices. Allowing both demand and supply responses to changes in global
Maurice Obstfeld and Kenneth S. Rogoff 137

balances will, in general, attenuate the change in relative prices required

to restore equilibrium. By distinguishing the short from the long run, such
that flexibility to shift resources from producing one good to producing
the other is limited in the short run relative to the long run, one can allow
for possible overshooting in equilibrium relative prices in the short run
without having to appeal to price rigidities. The model collapses to the
endowment model if there is complete inflexibility in the short run. In a
special case of the model, there is no change in the equilibrium real
exchange rate from its initial value as the economy adjusts to the elimina-
tion of the trade deficit.
The model is static, and there are no investment or capital flows. The
current account deficit is modeled as a pure income transfer from one
country to the other. For simplicity, preferences over three commodities
(the two traded goods and the nontraded good) in each country are assumed
to be homothetic. For simplicity, assume that the economy running a trade
deficit financed by an income transfer is a small open economy trading
with a large rest of the world, so that it is a price taker in the world market
for its traded goods. (This is the so-called dependent economy of W. Salter
and T. Swan,18 in which the relative price of one traded good in terms of
the other is set by the world market.) This means that the two traded goods
can be aggregated into a single composite traded good for the small open
economy, as long as conditions in the world market do not change.
Suppose there are two factors of production (say, capital and labor), and
suppose the technology of production of all three goods exhibits constant
returns to scale and factor intensity (capital-labor ratio) nonreversal, so
that the ranking of the goods in terms of cost-minimizing factor intensities
is independent of factor prices. Then, given the relative price of traded
goods, the factor prices can be uniquely solved from the two expressions
equating price to unit cost, if both goods are produced in positive amounts.
Given the unique factor prices, the minimum unit cost of production of
the nontraded good and hence its price, given that a positive amount of it
is produced, are determined. Suppose the factor endowments are such that
the production possibility frontier (PPF) includes a nonempty subset S in
which all three goods are produced in positive amounts. In S the marginal
rate of transformation between the traded composite good (T ) and the
nontraded good is constant.

18. Salter (1959); Swan (1963).

 138 Brookings Papers on Economic Activity, 1:2005
Figure 1. Small Open Economy Model

Traded
composite
D

Q0

A
E
Q'0
Q1
P0
B F

P1

O C Nontraded good

Source: Author’s model described in the text.

In figure 1 above, the PPF is depicted as ABC, where the linear stretch AB
(except at points A and B) corresponds to S. The consumption possibility
frontier (CPF), given a transfer AD (in units of the traded composite good)
from the rest of the world, is depicted by DEFC. It is simply a vertical
shift of the PPF by the distance AD. Assume that preferences are represented
by a homothetic, quasi-concave utility function. Maximizing utility subject
to the CPF leads to the initial equilibrium consumption at Q0 and production
at P0 vertically below it. At Q0 an indifference curve touches the PPF so that
the common slope of the two is the slope of DE, which equals the relative
price of the traded composite good, that is, the real exchange rate.
Suppose now that the transfer is withdrawn so that the CPF coincides with
the PPF. If the real exchange rate does not change, by virtue of homothetic
preferences, consumption shifts to Q′0 on AB, where the ray OQ0 from the
origin intersects AB. If production remains at P0, there will be an excess
supply of the nontraded good. However, since the marginal rate of trans-
formation between traded and nontraded goods is constant along AB, the
Maurice Obstfeld and Kenneth S. Rogoff 139

production point shifts to Q′0 and the excess supply is eliminated. Thus the
economy adjusts to the elimination of the trade deficit by a pure quantity
adjustment with no change in the real exchange rate.
One could depict an endowment economy by reinterpreting the PPF as
just a single point P0. With income transfer P0Q0, the economy consumes
at Q0. Let DE be the tangent to the indifference curve at Q0, so that it is
the real exchange rate from a consumption perspective. The withdrawal
of the transfers requires that the consumption point move to P0. But if
the consumption real exchange rate does change, it will move to Q′0,
once again creating an excess demand for traded goods. To eliminate this
excess demand, the real exchange rate for consumption has to change to the
slope of the indifference curve through P0. This means (under the standard
assumption of convexity of preferences and both goods being normal)
there has to be a rise in the relative price of the traded good in terms of the
nontraded good, or a real depreciation, since at P0 and Q0 the consumption
of the nontraded good is the same, whereas that of the traded good com-
posite is lower at P0.
Returning to the production economy, what if the initial consumption
point is on EF, such as Q1? By construction, because each point on EF is
at the same vertical distance (of AD) above the point on the stretch BC ver-
tically below it, the slope of the CPF at Q1 is the same as the slope of the
PPF at P1, and their common slope is equal to the equilibrium real exchange
rate from both a consumption and a production perspective. Now, with the
withdrawal of the income transfer, the CPF coincides with the PPF, and
production will remain at P1. At unchanged real exchange rates, consump-
tion will move to a point (not shown) to the left of P1 on the straight line that
is tangent to the PPF at P1, thus creating excess demand for traded goods.
To eliminate this excess demand, a real depreciation has to occur, with the
new equilibrium point lying to the left of P1 on the PPF, where an indif-
ference curve touches the PPF (not shown).
Short-run inflexibility and long-run flexibility in shifting resources start-
ing from the production point P1 can be easily illustrated. The short-run
PPF touches the long-run PPF at the initial production point P1 but is
below it otherwise, with the vertical distance between the two increasing as
the production of the nontraded good increasingly deviates from its level
at P1 in either direction. Under these assumptions, which seem natural for
depicting short-run inflexibility, it is clear that the short-run equilibrium
point where an indifference curve touches the short-run PPF will imply a
 140 Brookings Papers on Economic Activity, 1:2005

larger depreciation in terms of the real exchange rate than its long-run
value at the point where an indifference curve touches the long-run PPF.
In other words there is overshooting of the real exchange rate in the short
run. If we add on a dynamic adjustment of the short-run PPF to the long-run
PPF over time, it follows that, after overshooting, the real exchange rate
will converge to its long-run value.
The essential features of the adjustment will remain in its extension to a
multicountry general equilibrium setup. However, its diagrammatic expo-
sition will not. The reason is that the convenient device of a Hicksian com-
posite traded good depends on the relative price of traded goods not changing.
This cannot hold in general in the general equilibrium setup, because the
relative prices are endogenous. Although the Obstfeld-Rogoff model is a
three-country general equilibrium model, it replicates the essential quali-
tative conclusion of adjustment in a small open economy, namely, that a
real depreciation is generally (though not necessarily always) needed to
eliminate global imbalances.
For their purpose, which is to arrive at a quantitative estimate of the extent
of the real depreciation needed to eliminate global imbalances, Obstfeld and
Rogoff have had to calibrate their general equilibrium model. Let me there-
fore conclude with a couple of comments on the calibration.19 Obstfeld and
Rogoff relate their choice of values for the two crucial parameters, θ (the
elasticity of substitution in consumption between the traded aggregate and
the nontraded good) and η (the so-called Armington elasticity of substitu-
tion between the domestic and foreign traded goods in the traded goods
aggregate), to econometric estimates in the literature. There are several
problems with this procedure. First, although θ is arguably a “deep” param-
eter in the Lucas sense, since it relates to preferences, η is not. As such,
any estimate of η will depend on the trade policy regime and therefore
cannot be stably estimated econometrically. Second, setting aside the policy
dependence of parameter values, since the Obstfeld-Rogoff model involves
aggregates, alternative schemes of aggregation will influence parameter
values, and whether the estimates in the literature are all comparable and
correspond to the implied aggregation of the Obstfeld-Rogoff model is not
obvious. To be fair, the authors are certainly aware of these issues, and
their simulations cover a range of values for the two parameters. Perhaps
they should cover an even broader range of values, particularly for η.

19. For a more detailed discussion, see Dawkins, Srinivasan, and Whalley (2001).
Maurice Obstfeld and Kenneth S. Rogoff 141

General discussion: William Nordhaus agreed with the authors that pro-
ductivity’s effect on the trade deficit depends critically on whether the
change in productivity occurs in the traded goods or the nontraded goods
sector. He reported his own recent findings showing that, whereas the pro-
ductivity slowdown in the United States during the 1970s had occurred in
both sectors, the acceleration of the 1990s was mainly in traded goods.
Although some estimates attribute almost all the acceleration to comput-
ers and associated industries, Nordhaus estimated that only somewhere
between a half and two-thirds came from that source. The most recent
data also suggest that productivity has accelerated, but he cautioned that
this may in part reflect mismeasurement of productivity in the retail sec-
tor. Noting that Jack Triplett had found the United States to be on the
frontier of improved measurement techniques that have tended to raise
estimates of productivity growth, Nordhaus speculated that Europe’s pro-
ductivity performance might look more like the United States’ if the same
measurement techniques were used for both.
Gian Maria Milesi-Ferretti noted that the April 2005 issue of the Inter-
national Monetary Fund’s World Economic Outlook (WEO) contains a
paper using a four-region global economic model that is similar to the
authors’ three-region model but allows for changes in production. As one
would expect, the WEO model finds that allowing for a production response
leads to a smaller, but still quite substantial, real depreciation of the dollar.
Sebastian Edwards observed that, in the authors’ Bretton Woods II sce-
nario, Asia’s surplus increases from 15 percent of U.S. traded-goods GDP
to 25 percent. The authors’ real model is unable to consider the monetary
consequences of this increase, but Edwards suggested that in the real world
it would create enormous pressure to expand the money supply in China
and the other Asian countries, requiring an extraordinary amount of steril-
ization to avoid inflation. Indeed, the latest data already show an increase
in inflation in China. Edwards also observed that the results of the authors’
global rebalancing scenario do not differ significantly from those of an
earlier two-region model of theirs; the introduction of the third region does
not appear to make a significant difference to the results.
Edmund Phelps reminded the panel of a paper he had co-written in 1986,
which argued that the expansionary U.S. fiscal policy of that era would
result in a boom in the United States while causing world real interest
rates to increase, leading to a recession in Europe. This analysis suggests
that if the United States were now to adopt fiscal austerity, world real interest
 142 Brookings Papers on Economic Activity, 1:2005

rates would decrease. This in turn would likely lead to an increase in asset
values in Europe and Asia, and thus an increase, rather than a decrease as
Richard Cooper had predicted in his comment, in output in those countries.
In the United States the shadow prices of business assets would fall, causing
a decline in production and investment.
Peter Garber argued that although Japan’s and Germany’s populations
are aging, and their populations growing slowly, there is significant under-
employment in their economies, especially in the nontradable services
industries. Underemployment is the reason that Japan has dramatically
increased its monetary base in order to maintain a high yen-dollar exchange
rate. Garber suggested that Germany is facing the same problem but is
incapable of making a similar intervention, and its difficulties are exerting
pressure on the European Central Bank to lower interest rates.
Maurice Obstfeld and Kenneth S. Rogoff 143

References
Broda, Christian, and David E. Weinstein. 2004. “Globalization and the Gains
from Variety.” Working Paper 10314. Cambridge, Mass.: National Bureau of
Economic Research (February).
Bryant, Ralph C. 1990. “Comment on ‘Eight Myths About the Dollar.’ ” In The
Economics of the Dollar Cycle, edited by Stefan Gerlach and Peter A. Petri.
MIT Press.
Calvo, Guillermo A. 1998. “Capital Flows and Capital-Market Crises: The Simple
Economics of Sudden Stops.” Journal of Applied Economics 1: 35–54.
Campa, José Manuel, and Linda S. Goldberg. 2002. “Exchange Rate Pass-Through
into Import Prices: A Macro or a Micro Phenomenon?” Working Paper 8934.
Cambridge, Mass.: National Bureau of Economic Research (May).
Cline, William R. Forthcoming. The United States as a Debtor Nation: Risks and
Policy Reform. Washington: Institute for International Economics.
Cooper, Richard N. 2001. “Is the U.S. Current Account Sustainable? Will It Be
Sustained?” BPEA, no. 1: 217–26.
Croke, Hilary, Steven B. Kamin, and Sylvain Leduc. 2005. “Financial Market
Developments and Economic Activity during Current Account Adjustments in
Industrial Economies.” International Finance Discussion Papers 827. Washington:
Board of Governors of the Federal Reserve System (February).
Dawkins, Christina, T. N. Srinivasan, and John Whalley. 2001. “Calibration.” In
Handbook of Econometrics, vol. 5, edited by James J. Heckman and Edward
Leamer. Amsterdam: Elsevier.
Dooley, Michael P., David Folkerts-Landau, and Peter M. Garber. 2004a. “The
Revived Bretton Woods System: The Effects of Periphery Intervention and
Reserve Management on Interest Rates and Exchange Rates in Center Coun-
tries.” Working Paper 10332. Cambridge, Mass.: National Bureau of Economic
Research (March).
________. 2004b. “Direct Investment, Rising Real Wages, and the Absorption of
Excess Labor in the Periphery.” Working Paper 10626. Cambridge, Mass.:
National Bureau of Economic Research (September).
Dornbusch, Rudiger. 1976. “Expectations and Exchange Rate Dynamics.” Jour-
nal of Political Economy 84, no. 6: 1161–76.
________. 1990. Comment on Tobin, “Eight Myths about the Dollar.” In The
Economics of the Dollar Cycle, edited by Stefan Gerlach and Peter A. Petri.
MIT Press.
Edwards, Sebastian. 2004. “Thirty Years of Current Account Imbalances, Current
Account Reversals, and Sudden Stops.” International Monetary Fund Staff Papers
51(special issue): 1–49.
Feenstra, Robert C. 1994. “New Product Varieties and the Measurement of Inter-
national Prices.” American Economic Review 84, no. 1: 157–77.
 144 Brookings Papers on Economic Activity, 1:2005

Frankel, Jeffrey A., and Andrew K. Rose. 1995. “Empirical Research on Nominal
Exchange Rates.” In Handbook of International Economics, vol. 3, edited by
Gene M. Grossman and Kenneth Rogoff. Amsterdam: Elsevier.
Freund, Caroline, and Frank Warnock. 2005. “Current Account Deficits in Indus-
trial Countries: The Bigger They Are, The Harder They Fall?” World Bank and
Darden Graduate School of Business, University of Virginia (May).
Gagnon, Joseph E. 2003. “Productive Capacity, Product Varieties, and the Elastic-
ities Approach to the Trade Balance.” International Finance Discussion Paper 781.
Washington: Board of Governors of the Federal Reserve System (October).
Gerlach, Stefan, and Peter A. Petri. 1990. The Economics of the Dollar Cycle. MIT
Press.
Goldberg, Pinelopi Koujianou, and Michael M. Knetter. 1997. “Goods Prices and
Exchange Rates: What Have We Learned?” Journal of Economic Literature 35,
no. 3: 1243–72.
Gourinchas, Pierre-Olivier, and Hélène Rey. 2005a. “International Financial
Adjustment.” Working Paper 11155. Cambridge, Mass.: National Bureau of
Economic Research (February).
________. 2005b. “U.S. External Adjustment: The Exorbitant Privilege.” Univer-
sity of California, Berkeley, and Princeton University (April).
Greenspan, Alan. 2004. “The Evolving U.S. Payments Imbalance and Its Impact on
Europe and the Rest of the World.” Cato Journal 24 (Spring-Summer): 1–11.
Grubert, Harry, Timothy Goodspeed, and Deborah Swenson. 1993. “Explaining
the Low Taxable Income of Foreign-Controlled Companies in the United
States.” In Studies in International Taxation, edited by Alberto Giovannini, R.
Glenn Hubbard, and Joel Slemrod. University of Chicago Press.
Harris, David, Randall Morck, Joel Slemrod, and Bernard Yeung. 1993. “Income
Shifting in U.S. Multinational Corporations.” In Studies in International Taxa-
tion, edited by Alberto Giovannini, R. Glenn Hubbard, and Joel Slemrod. Uni-
versity of Chicago Press.
Hooper, Peter, Karen Johnson, and Jaime Marquez. 2000. Trade Elasticities for
the G-7 Countries. Princeton Studies in International Economics 87. Princeton
University (August).
Hooper, Peter, and John Morton. 1982. “Fluctuations in the Dollar: A Model of
Nominal and Real Exchange Rate Determination.” Journal of International
Money and Finance 1, no. 1: 39–56.
Krugman, Paul R. 1985. “Is the Strong Dollar Sustainable?” In The U.S. Dollar—
Recent Developments, Outlook, and Policy Options. Kansas City, Mo.: Federal
Reserve Bank of Kansas City.
________. 1989. “Differences in Income Elasticities and Trends in Real Exchange
Rates.” European Economic Review 33, no. 5: 1031–54.
________. 1991. Has the Adjustment Process Worked? Policy Analyses in Inter-
national Economics 34. Washington: Institute for International Economics.
Maurice Obstfeld and Kenneth S. Rogoff 145

Lane, Philip R., and Gian Maria Milesi-Ferretti. 2004. “The Transfer Problem
Revisited: Net Foreign Assets and Real Exchange Rates.” Review of Econom-
ics and Statistics 86, no. 4: 841–57.
________. 2005a. “Financial Globalization and Exchange Rates.” Working Paper
05/3 (January). Washington: International Monetary Fund.
________. 2005b. “A Global Perspective on External Positions.” Trinity College,
Dublin, and International Monetary Fund (May).
________. Forthcoming. “The External Wealth of Nations Mark II: Revised and
Extended Estimates of Foreign Assets and Liabilities, 1970–2003.” Trinity
College, Dublin, and International Monetary Fund.
Mann, Catherine L. 2005. “Breaking Up Is Hard to Do: Global Co-Dependency,
Collective Action, and the Challenges of Global Adjustment.” CESifo Forum
6, no. 1: 16–23.
Marazzi, Mario, and others. 2005. “Exchange Rate Pass-through to U.S. Import
Prices: Some New Evidence.” International Finance Discussion Paper 2005-833.
Federal Reserve Board (April).
McCulloch, Rachel. 1990. “Macroeconomic Policy, Trade and the Dollar: Remark-
able Events of the 1980s.” In The Economics of the Dollar Cycle, edited by Ste-
fan Gerlach and Peter A. Petri. MIT Press.
Meese, Richard, and Kenneth Rogoff. 1983. “Empirical Exchange Rate Models
of the Seventies: Do They Fit Out of Sample?” Journal of International Eco-
nomics 14 (February): 3–24.
Mendoza, Enrique G. 1991. “Real Business Cycles in a Small Open Economy.”
American Economic Review 81, no. 4: 797–818.
Mussa, Michael. 2005. “Sustaining Global Growth while Reducing External
Imbalances.” In The United States and the World Economy: Foreign and Eco-
nomic Policy for the Next Decade, edited by C. Fred Bergsten. Washington:
Institute for International Economics.
Obstfeld, Maurice. 2004. “External Adjustment.” Review of World Economics 140,
no. 4: 541–68.
Obstfeld, Maurice, and Kenneth Rogoff. 1996. Foundations of International
Macroeconomics. MIT Press.
________. 2000a. “Perspectives on OECD Economic Integration: Implications
for U.S. Current Account Adjustment.” In Global Economic Integration:
Opportunities and Challenges. Kansas City, Mo.: Federal Reserve Bank of
Kansas City (August).
________. 2000b. “The Six Major Puzzles in International Macroeconomics: Is
There a Common Cause?” In NBER Macroeconomics Annual 2000, edited by
Ben S. Bernanke and Kenneth Rogoff. MIT Press.
________. 2004. “The Unsustainable US Current Account Position Revisited.”
Working Paper 10869. Cambridge, Mass.: National Bureau of Economic
Research (November).
 146 Brookings Papers on Economic Activity, 1:2005

Obstfeld, Maurice, and Alan M. Taylor. 1997. “Nonlinear Aspects of Goods-

Market Arbitrage and Adjustment: Heckscher’s Commodity Points Revisited.”
Journal of the Japanese and International Economies 11, no. 4: 441–79.
________. Forthcoming. “Sources of America’s ‘Exorbitant Privilege.’ ” Univer-
sity of California.
Orcutt, Guy H. 1950. “Measurement of Price Elasticities in International Trade.”
Review of Economics and Statistics 32, no. 2: 117–32.
Ostry, Jonathan D., and Carmen M. Reinhart. 1992. “Private Saving and Terms of
Trade Shocks: Evidence from Developing Countries.” International Monetary
Fund Staff Papers 39, no. 3: 495–517.
Porter, Richard D., and Ruth A. Judson. 1996. “The Location of U.S. Currency:
How Much Is Abroad?” Federal Reserve Bulletin 82 (October): 883–903.
Reifschneider, David, Robert Tetlow, and John Williams. 1999. “Aggregate Dis-
turbances, Monetary Policy, and the Macroeconomy: The FRB/US Perspec-
tive.” Federal Reserve Bulletin 85, no. 1: 1–19.
Roubini, Nouriel, and Brad Setser. 2004. “The U.S. as a Net Debtor: The Sustain-
ability of U.S. External Imbalances.” Stern School of Business, New York Uni-
versity, and University College, Oxford (August).
Ruhl, Kim. 2003. “Solving the Elasticity Puzzle in International Economics.”
Federal Reserve Bank of Minneapolis (November).
Salter, W. E. G. 1959. “Internal and External Balance—The Role of Price and
Expenditure Effects.” Economic Record 35: 226–38.
Stockman, Alan C., and Linda L. Tesar. 1995. “Tastes and Technology in a Two-
Country Model of the Business Cycle: Explaining International Comove-
ments.” American Economic Review 85, no. 1: 168–85.
Swan, T. 1963. “Longer Run Problems of the Balance of Payments.” In The Aus-
tralian Economy, edited by H. W. Arndt and W. M. Corden. Melbourne:
Cheshire.
Tille, Cédric. 2004. “Financial Integration and the Wealth Effect of Exchange
Rate Fluctuations.” Federal Reserve Bank of New York (August).
Tobin, James. 1990. “Eight Myths about the Dollar.” In The Economics of the
Dollar Cycle, edited by Stefan Gerlach and Peter A. Petri. MIT Press.
Ventura, Jaume. 2001. “A Portfolio View of the U.S. Current Account Deficit.”
BPEA, no. 1: 241–53.
Warnock, Francis E. 2003. “Exchange Rate Dynamics and the Welfare Effects of
Monetary Policy in a Two-Country Model with Home-Product Bias.” Journal
of International Money and Finance 22, no. 3 (June): 343–63.
MICHAEL DOOLEY
University of California, Santa Cruz
PETER GARBER
Deutsche Bank

Is It 1958 or 1968? Three Notes on

the Longevity of the Revived
Bretton Woods System

IT IS NOW widely accepted that the broad outlines of the current interna-
tional monetary system are as we described them almost two years ago
and labeled “the Revived Bretton Woods system.” This system’s main
features are
—the emergence of a macroeconomically important group of economies
that manage their currencies vis-à-vis the dollar to support export-driven
growth
—the United States as center and reserve currency country, providing
financial intermediation services for foreign, and particularly Asian, sav-
ing through its national balance sheet, and willing to accept large current
account imbalances
—a group of poorer economies implementing export-led development
policies and exporting large amounts of capital to richer economies, mostly
the United States
—unusually low and even falling short- and long-term real interest rates
as a result of this glut of mobile global savings, and
—a group of industrial and emerging economies with floating exchange
rates, whose currencies are under incessant pressure to appreciate.
Not agreed and under vigorous discussion is how long this system can
last. Will it be a meteoric flash with a spectacular end soon to come? Or

We are grateful to Daniel Riera-Crichton for his able and diligent research assistance in
producing the empirical results of this paper.

147
 148 Brookings Papers on Economic Activity, 1:2005

will it last for the reasonably foreseeable future? We package these questions
here in an analogous question: Is the Revived Bretton Woods system at
the point in its development where the original Bretton Woods was in 1958
or 1968 or 1971?
In a series of publications we have provided a fundamental underpin-
ning for why we believe the system will last—that the situation today is
more like 1958.1 We argue that the gains to the players from continuing
their actions outweigh the costs that many have argued will arise in an
endgame asset price shift or in unexploited benefits of portfolio diversifi-
cation. Rather than characterize the situation with geopolitically charged
rhetoric like “balance of financial terror,”2 we think it more valid to think
in the familiar economic terms of “mutually beneficial gains from trade,”
such as might exist between any borrower and lender or between any pur-
veyor of goods and its customer.
Here we further develop our argument in the form of three notes address-
ing particular issues that have cropped up in critiques of the Revived Bretton
Woods view. These notes both respond to the critiques and continue to
expand our ideas. The first note explains how we think about what is driving
capital flows to the United States and keeping interest rates low. We view
the fact of unusually low long-term real interest rates for this stage of the
business cycle as a direct challenge to those who, exaggerating the impor-
tance of rumors about central bank reserve management practices, claim
that the end is near.
The second note seeks to provide some information about the experi-
ence of those emerging economies with chronic current account surpluses
since the breakdown of the first Bretton Woods system. A very large empir-
ical literature evaluates the experience of emerging economies that have
run chronic deficits, and the costs and frequency of associated financial
crises. But we are not aware of any similar evaluations of the durability and
stability of those foreign exchange regimes that have resulted in unusual
sequences of current account surpluses and accumulations of international
reserves. The widespread view that the surplus regimes at the core of the

1. Eichengreen (2004), in contrast, seems to favor 1968, that is, to allow the system a
few years more to run, whereas Frankel (2005a) favors 1971. Roubini and Setser (2004)
call for something even more immediate and apocalyptic, yet they acknowledge that the
day of reckoning may be as long as two years off.
2. Summers (2004a, 2004b).
Michael Dooley and Peter Garber 149

Revived Bretton Woods system will come to a quick and costly end has
likely been inferred in part from the recent experience of debtor emerging
economies. Our interpretation of the experience of these surplus regimes
is that they have been and may well remain durable and immune from
financial crises.
The third note addresses an issue that has been raised frequently in crit-
icisms of our comparing the current system to the Bretton Woods system,
namely, that the United States is running large current account deficits
now, but it was not then. Of course, many aspects of the current system are
different from what they were in the heyday of Bretton Woods: Konrad
Adenauer is no longer chancellor of Germany, Charles de Gaulle is dead,
the United States no longer guarantees gold convertibility, and there is
now a serious pretender to reserve currency status. Our first reaction was
that this difference was as superficial as these others and not at the heart
of the comparison we wanted to draw.
But the United States did have a major balance of payments deficit during
the Bretton Woods era, which was the proximate driver of the deterioration
of the system. So we relate the U.S. balance of payments deficits under
Bretton Woods to the U.S. current account deficits under the Revived
Bretton Woods to show that there is a close analogy. This is something more
than an exercise in the history of economic ideas, because it plays into our
view that collateral is the key to opening sizable gross cross-border trade
in assets in a system that is short on trust.

Real Interest Rates Say It Is 1958

Why is the real interest rate in the United States so low and falling
today, in the growth phase of the U.S. and global business cycles, even as
the U.S. current account deficit reaches record levels? At the end of June
2002, about when the euro began its recent appreciation, realized ten-year
real annual interest rates on U.S. Treasury securities were 3.70 percent
on nominal notes and 3.07 percent on inflation-protected securities. The
corresponding numbers at the end of December 2003 were 2.35 percent
and 1.95 percent, respectively. One year later the respective numbers were
1.17 percent and 1.63 percent, and as we write in mid-May 2005, they are
1.02 percent and 1.65 percent. This fall in rates has come in a period when
the media swirls daily with stories about foreigners losing confidence,
 150 Brookings Papers on Economic Activity, 1:2005

foreign exchange reserve managers diversifying portfolios, and imminent

collapse as everyone tries to be the first out the door. If all this is true, the
bond and credit markets have not noticed.
Three broad realities underpin our view of this global phenomenon, all
of which we expect to continue into the foreseeable future:

Reality 1: About fifteen years ago, hundreds of millions of under-

employed workers joined the world’s market economies. They had no
capital to speak of, but they had a desire to work in industry and to get
rich. One might expect that such an increase in the global supply of labor
would drive real interest rates up, but these workers came with an enor-
mously high saving rate and lived under the yoke of a dead financial sys-
tem, which had served them in the past as a capital destroyer, as it does to
this day. They lacked modern technology and management. Theirs was a
communist society that was and is problematic geopolitically, which might,
in turn, make their access to cross-border credit problematic. This created
a profound global disequilibrium for the industrial world, equal in magni-
tude to the global unemployment problem of the Great Depression although
much more concentrated geographically. The industrial world’s economic
system has to resolve this fundamental imbalance over the course of time
by absorbing these workers. To focus today on trade imbalances when in
fact there is an enormous labor market imbalance is to make the same
mistake that economists and policymakers made in the 1930s.
Reality 2: The emerging economies that have developed most success-
fully are those that export capital on net. Joshua Aizenman, Brian Pinto,
and Artur Radziwill demonstrate, in a sample of forty-seven developing
countries from 1981 to 2001, that the net exporters of domestic savings
among them had significantly higher growth rates.3 They conclude that “a
rise in the self-financing ratio [the stock of tangible capital supported only
by past national saving, divided by the actual stock of capital] from 1 to
1.1 is associated with an increase in the growth rate from 2.8% to 4.4%.
Further, reducing the self-financing ratio from 1 to 0.9 is associated with a
drop in the growth rate from 2.8% to 2.2%.”4 These estimates control for
differences in institutional quality as well as in trade and financial openness.
They clearly contradict the usual assumption that developing countries

3. Aizenman, Pinto, and Radziwill (2004).

4. Aizenman, Pinto, and Radziwill (2004, p. 9).
Michael Dooley and Peter Garber 151
Figure 1. Market Equilibrium for Loanable Funds

Real interest rate

U.S. demand

Private sector
supply

Quantity of funds

Source: Authors’ model described in the text.

have been successful in using net foreign saving to augment capital for-
mation and economic growth. We argue below that their results are con-
sistent with the idea that net exports of saving from poor countries support
two-way trade in private financial assets that improves the quality and
productivity of domestic capital formation.
Reality 3: The United States has a large and growing current account
deficit, funded at this moment by the foreign private and official sectors at
low and falling real interest rates. In recent years the official sector has
taken up a large share of this deficit.

Let us focus on reality 3 for a moment. We like to think about the United
States’ external deficit problem in a simple loanable funds flow framework.
After netting U.S. public and private investment demand from U.S. saving,
the United States has a demand for saving from the rest of the world that is
downward sloping when plotted against the real interest rate, as in figure 1:
the lower the real interest rate that it faces, the less the United States wants
to save and the more it wants to invest. Given a real interest rate, then, we
 152 Brookings Papers on Economic Activity, 1:2005
Figure 2. Effect of Expansionary Fiscal Policy

Real interest rate

U.S. demand

Private sector
supply

Quantity of funds

Source: Authors’ model described in the text.

can read off the U.S. current account deficit. Meanwhile there is an upward-
sloping supply of saving coming from the foreign private sector (we intro-
duce foreign official flows below), perhaps from asset managers looking
only at Sharp ratios and benchmarks, or perhaps from foreign industrial
corporations interested in return on capital. The higher the real interest rate
available in the United States, the more of this private foreign saving flows
in. The intersection of these two curves determines the global real interest
rate, the U.S. current account deficit, and the rest of the world’s current
account surplus.
A looser fiscal policy might shift the demand for foreign saving upward
as in figure 2. This would bring in more foreign saving or, equivalently,
increase the current account deficit. And it would cause the real interest
rate to rise, as in the Reagan-era deficits of the early 1980s. Some of this
may be going on today, but it is clearly not dominant. Since 2002, mar-
Michael Dooley and Peter Garber 153
Table 1. Marketable U.S. Government Debt, Interest Rates, and Fiscal Deficits,
2002–05
Item 2002 2003 2004 2005
Marketable debt (billions of dollars)a 3,020 3,317 3,721 4,085
Average interest rate paid (percent a year)a 4.99 4.11 3.61 3.94
Fiscal deficit (percent of GDP)b 1.5 3.4 3.6 3.0
Source: Bloomberg data.
a. As of March 31.
b. In preceding fiscal year.

ketable U.S. debt has increased by more than one-third while nominal and
real interest rates have declined (table 1). Moreover, relative to those of other
industrial countries, the U.S. budget deficit is not unusually large (figure 3);
it is hard to see why this factor alone would increase the U.S. current account
deficit relative to the deficits or surpluses of other industrial countries, espe-
cially given the much more rapid GDP growth rate (real and nominal) in
the United States.
Instead consider figure 4, which adds a vertical official sector supply
curve, reflecting the fact that policymakers have objectives other than a
narrow risk-return calculus localized to this small portion of national saving.
Their development goals require the export of domestic saving, and they will
accept whatever interest rate the market determines. Adding, horizontally,
this new public sector supply of foreign saving to the private sector supply
shifts total supply rightward and brings down the interest rate that clears
this global market for savings.
So, if a falling U.S. real interest rate is observed alongside a rising U.S.
current account deficit, it can only mean that official capital is being pushed
into the United States and private capital is being pushed out, but by a smaller
amount than the official capital coming in. On net, capital is not being pulled
in by U.S. demand shifts. This is the combination of facts that shows us
that the United States is passive and that the foreign official sector is the
active player in global imbalances.
This means that the typical denunciation of U.S. “profligacy” is worse than
useless for understanding the situation: it is actually misleading. Usually, this
rhetoric includes a reference to the role of the U.S. fiscal deficit in reducing
net U.S. saving, but a larger fiscal deficit should increase the interest rate.
Whatever the size of this effect, it has clearly been more than overcome by
the effects of foreign official capital pushing in.
 154 Brookings Papers on Economic Activity, 1:2005

Figure 3. Fiscal Deficits, Growth, and Inflation, Selected Countries, 2004a

Percent of GDP or percent a year

Germany
7 U.S.
France
6 Italy
Euroland b
5 Japan
4
3
2
1
0
–1

Deficit–GDP ratio Nominal GDP growth Inflation GDP growth

Source:–Bloomberg.
a. Budget deficit data are for fiscal 2004; all other data are for calendar 2004.
b. Countries that have adopted the euro.

Figure 4. Effect of Adding an Inelastic Official Sector Supply

Real interest rate

Pre-intervention
equilibrium Private sector
supply

Official sector
supply

U.S. demand

Official + private
sector supply

Quantity of funds

Source: Authors’ model described in the text.

Michael Dooley and Peter Garber 155

One often hears that private saving flows to the United States are falling
because of increased risks, stemming perhaps from the worsening U.S. inter-
national investment position. This would show up as a shift of the private
supply curve in figures 1, 2, and 4 upward and to the left, and it would put
further upward pressure on real interest rates. But this is exactly the oppo-
site of what we observe. Rather, the evidence is far more consistent with a
downward slide along a given private supply curve after the public sector
supply is added. To be sure, foreign private saving is financing far less of
the U.S. current account deficit than it did, say, five years ago. But the rea-
son is that private investors are being driven out by official sector flows
willing to replace them at much lower interest rates.
One should beware of making too much of a rising or falling fraction
of official sector finance in any given quarter or year. A steadily growing
flow from the foreign official sector to the United States year in and year
out is not necessary to maintain the system. Official flows are necessary only
when the foreign central bank must intervene to keep its currency under-
valued. As in a target zone exchange rate regime, when the private sector
is confident that the regime is durable and will be sustained by future inter-
ventions as the need arises, private inflows are sufficient to provide the
deficit financing.
So far, so good. But it means that today’s low real interest rate is a
momentary flow effect that will evaporate should official sector lending to
the United States dry up permanently. If this were to happen, the picture
would snap back from figure 3 to figure 1, and interest rates would jump.
If this is what the market expects, we should today see low short-term real
interest rates and much higher long-term rates. But we do not see this. Long-
term real rates are low—hence the conundrum that we are studying now.
Implicit in the real yield curve is that the equilibrium of figure 3 should
last a long time.
It follows that even a hint that Asian governments might reduce their flow
demand for dollar assets will generate an immediate jump in the ten-year
rate in the United States. Indeed, many observers doubt that foreign official
interests in funding the U.S. current account deficit are sustainable.

Back to Reality 1
The stakes are high indeed. Why should Asian authorities remain willing
to increase their claims against the United States? To answer this, we have
 156 Brookings Papers on Economic Activity, 1:2005

to focus on the strategy that Asia (notably China) has chosen to solve the
development and unemployment problems that are part of reality 1.
The problem for China is to mobilize its existing enormous domestic
saving to create a growing, internationally competitive capital stock that
can rapidly employ hundreds of millions of workers in productive activity.
A serious constraint is the lack of a domestic financial system capable of
channeling this saving into productive capital, technology, and management
skills.
The solution, perhaps stumbled upon inadvertently, has been to engage
in export-led growth, thereby providing an immediate global quality check
on the goods produced. This avoids falling off the cliff of another Great
Leap Forward. To get export markets open, part of the policy has been to
offer a large incentive to potential industrial exporters, both domestic and
foreign-based, in the form of low dollar wages and the expectation that
wages will rise only slowly toward world levels. Slowly rising dollar
wages could be associated with a gradual nominal revaluation of the ren-
minbi or a slightly higher rate of inflation than in China’s trading partners.
For example, a 3 to 5 percent revaluation of the renmimbi later this year and
the adoption of a carefully controlled float of the exchange rate would not
signal the end or even a material change in the development strategy we
have described.5
The typical problem in emerging economies is how not to offer too high
an industrial wage relative to wages elsewhere in the economy: too-rapid
industrialization could drive industrial wages sharply above agricultural
wages, deterring investment in industry and triggering a flood of migration
to the cities. By keeping wages low and relatively uniform, an initially
low but rising currency helps both to induce resource transfers to industry
and to restrain migration to a rate consistent with capital formation in the
industrial sector.6
Foreign direct investors have been encouraged because they bring the
discipline of international financial intermediation. Additional benefits
include technology transfer and the proven political clout to keep export

5. In Dooley, Folkerts-Landau, and Garber (2004b), we treat the initial stock of labor
as an exhaustible resource. In that context it is optimal for the government to absorb labor
more rapidly at the beginning of the regime. It follows that dollar wages are initially set at
a low level but rise over time to the world wage when the last worker is absorbed. See
Salant (1976) for a more general discussion.
6. We thank Vincent Reinhart for this insight.
Michael Dooley and Peter Garber 157

markets open. The importance of direct investment in keeping U.S. export

markets open has been questioned on a variety of grounds. Eswar Prasad
and Shang-Jin Wei argue that most foreign direct investment into China
comes from outside the United States and is not likely a significant factor
in keeping U.S. markets open to Chinese exports.7 Leaving aside the inher-
ent difficulty in determining the nationality of direct investors, we would
point out that Asian direct investors in China also have an enviable record
in penetrating U.S. markets and in dealing with the threat of U.S. protec-
tion. It seems likely to us that the building political pressure in the United
States to do something about the bilateral trade deficit with China would be
more effective if U.S. and other multinational corporations were not active
in direct investment in China.
We are more than willing, however, to base our forecast of the durability
of the system on the lasting inability of domestic credit markets in emerg-
ing economies to efficiently intermediate domestic saving. The difficulty
in reforming financial markets in these economies, and the frequency and
costs associated with crises in economies that have not been successfully
developed, are in our view the primary lesson provided by the failures of
development in Latin America.
But why does the need for international financial intermediation (two-
way trade in financial assets) create saving-investment imbalances and a
flood of net capital exports in the first place? After all, an export-based
development policy need not imply a net export of capital. All that is
needed is export growth, and this can just as well be balanced by import
growth as not.
In general, the successful emerging economies have not needed net
foreign saving; such inflows are generally small and unreliable relative to
domestic saving (reality 2). Nevertheless, other things equal, even a small
addition of net foreign saving should contribute to investment and growth
in poor countries. A positive argument in favor of net exports of saving
requires that some other important ingredient to growth not be available
in equal measure.
Our hypothesis is that net exports of domestic saving are necessary to
earn the collateral required for efficient international intermediation of
domestic saving. Asian emerging economies do not need net foreign saving,
but they do need efficient financial intermediation. We have emphasized

7. Prasad and Wei (2005).

 158 Brookings Papers on Economic Activity, 1:2005

foreign direct investment and other types of international financial inter-

mediation because we are not optimistic about the rapid development of
domestic credit markets. That is, residents of these countries can avoid
these markets by placing some of their assets off shore. These will return if
international investors are protected from political risk—especially impor-
tant when private capital is flowing to and from a geopolitically problematic
country in large amounts. The government can relax this credit constraint
by keeping its balance sheet very strong versus the rest of the world, that
is, by building net reserves.
The government’s net reserves then provide protection for private inter-
national financial intermediation against various geopolitical risks. In effect,
the emerging economy’s government promises to stay on the sidelines by
becoming a net creditor to the rest of the world. Note that a government can-
not borrow this credibility; it has to earn it by placing goods and services in
the rest of the world on net. And placing more goods and services in the
rest of the world than one is taking in means a current account surplus.
Imagine that foreign direct investment flows are matched by official sec-
tor reserve growth in the balance of payments accounts and that the current
account is balanced. Then the capital account is balanced in terms of both net
and gross flows. But the country sending the foreign direct investment is tak-
ing an unbalanced risk position, effectively buying equity and borrowing in
fixed-interest securities. Usually, in private markets, this requires some col-
lateral from the lesser credit. The way an emerging economy delivers collat-
eral is by running a current account surplus. The faster the gross positions in
the capital account grow, the faster must the current account imbalance grow
to support the unbalanced risk positions. We believe that this view of current
account surpluses as collateral provides a first explanation of the connection
between net and gross capital flows, currently a noteworthy lacuna in models
of international finance, which ignore gross trade in assets.

Other Asia and Japan

A reasonable objection to our argument is that it does not fit the more
developed countries in Asia, especially Japan, that have been the most eager
buyers of U.S. assets. In fact, it is useful to consider China and Japan as
spanning the problems facing Asia. Both Japan and China have an employ-
ment problem, but in Japan it is the result of a very long cyclical downturn,
whereas in China it is a long-term development problem. Both governments
Michael Dooley and Peter Garber 159

look to export growth as a solution to this problem, and both have a long
history of managing the exchange rate.
For quite different reasons, both countries have been able to sterilize
very large reserve accumulations. In deflationary Asian countries, notably
Japan, it is difficult to understand why there might be some limit on the
ability or motivation of the authorities to create yen in stemming an attack
on the currency. With interest rates at zero, it is costless to create as much
yen cash as is demanded, whereas dollar reserves produce a positive yield.
Normally, a limit on foreign exchange acquisition is reached when the
resulting monetary expansion causes excessive overheating and inflation.
But such an expansion is still not in sight for Japan and would not, in any
case, be the appropriate monetary policy.
The lessons of attacks on weak currencies and fixed exchange rate re-
gimes seem to be the ones being applied by the global private financial sec-
tor here. The authorities in such regimes face a limit on reserves or credit or
the amount of pain they are willing to put the economy through, and so each
attack on the currency is simultaneously a ratcheting up of the probability
that the currency will indeed collapse. Some observers seem to be holding a
case study of a typical speculative attack against a mirror and thinking that
private capital inflows likewise ratchet up the pain in Japan. Yet quite the
opposite is true in deflationary Japan. Japan has ceased its massive interven-
tion since the first quarter of 2004, and the yen has actually depreciated
somewhat against the dollar. Our expectation is that the authorities will re-
turn to the market if private flows to the United States again decline and the
yen again appreciates, especially if it is tested in another attack.
In China financial repression has allowed the authorities to place domestic
assets generated by sterilization without much increasing domestic interest
rates, and it has been very successful in containing inflation. The People’s
Bank of China currently places three-year domestic currency debt in the
banks at an annual interest rate of about 3 percent and is experiencing pos-
itive carry on its foreign exchange.8 Other emerging economies in Asia
with relatively open capital markets have followed a middle course of try-
ing to stay competitive with China but allowing some appreciation of their
currencies against the dollar, although still with heavy currency management
and accumulation of reserves. The success and durability of these efforts

8. This in contradiction to the continual alarmist statements from Goldstein and Lardy
(2004), among others.
 160 Brookings Papers on Economic Activity, 1:2005

are a matter of intense debate, but we doubt there is much to be gained from
continuing the debate at the theoretical level. We turn to the empirical
debate in the next section.

How Do Episodes of Reserve Accumulation End?

In a series of papers,9 we have argued the case for a meaningful distinc-

tion between countries that allow private international investment decisions
to determine important macroeconomic variables such as the real exchange
rate and the current account balance, and countries for which government
investment decisions determine these magnitudes. We have referred to these
as “capital account countries” and “trade account countries,” respectively.
Trade account countries repress private financial flows and overwhelm
with official flows those that slip through the repression. Capital account
countries, in contrast, do not block cross-border flows or significantly
intervene in foreign exchange markets. It is often assumed that the con-
ventional analytical framework developed to understand the behavior of
capital account countries applies also to trade account countries, because
capital and foreign exchange controls are mostly ineffective. In our view
this is entirely an unresolved empirical issue.
The opinion that the U.S. current account deficit is unsustainable flows
from a conviction that private international investors will be unwilling to
continue to accumulate net claims on the United States. In this view, more-
over, either the official capital flows that have partly financed the U.S.
current account deficit will be overwhelmed by private sector flows, or
governments will come to their senses in time to avoid a crisis. The usual
dark warning is that the longer it takes the official sector to realize the
inevitable truth, the harsher will be the consequences. The phase diagrams
of the speculative attack models dance in our collective heads.
We fully agree with half of this prediction. Two years ago we predicted
that private investors would become more reluctant to finance the U.S. cur-
rent account deficit as official sector capital flowed in.10 We also predicted
the very large appreciation of the euro and other currencies whose trade in
foreign exchange markets is dominated by private capital flows. This was

9. Dooley, Folkerts-Landau, and Garber (2003, 2004a, 2004b, 2004c).

10. Dooley, Folkerts-Landau, and Garber (2003).
Michael Dooley and Peter Garber 161

not unusual in itself. But we also argued that governments of a group of

what we called “trade account countries” (countries where repression of
private financial flows determine the real exchange rate and the current
account balance) had good reasons to continue to invest in the United States
for an extended period and that this would keep U.S. interest rates low, con-
trary to then-prevailing opinion.11 The length of this period is derived from
an optimal rate of absorption of those countries’ unemployed labor. In our
view of the real forces behind this system, this suggests a decade at least.
So it is important to understand why nonresidents are supplying net
saving to the United States at very low expected yields and why this may
or may not continue. To us, it is irrelevant to the overall picture whether the
net foreign investment is in Treasury securities, agency securities, private
fixed-income securities, equity, or something else. It is irrelevant whether
private or official foreigners take larger or smaller shares of the foreign
investment in the United States in any given year. It is mostly irrelevant how
the spreads across different classes of financial instruments in the United
States might be affected. This is not a discussion of investment strategy or
asset allocation; it is entirely directional.

Historical Evidence
One way to begin to evaluate the durability of the Revived Bretton
Woods system and the likely consequences of its demise is to study the
experience of economies that have had unusually long sequences of current
account surpluses and accumulations of official reserves. Doubts about the
durability of the system have generally centered on the ability and willing-
ness of surplus economies to maintain an undervalued currency for an
extended period. Does historical experience suggest that periods of reserve
accumulation are followed by speculative attacks that generate a real
appreciation (through either inflation or a nominal appreciation), losses on
dollar reserves, and painful recessions as resources are transferred from
traded goods industries?
The experience of emerging economies with chronic current account sur-
pluses since the breakdown of the original Bretton Woods system in 1971

11. The logic was that an increasingly indebted United States, with an interest rate
effectively underwritten by the trade account countries, would attract less private funding
from other (capital account) countries. Smaller net capital flows meant smaller net current
account flows, which would be accomplished through appreciation against the dollar.
 162 Brookings Papers on Economic Activity, 1:2005

has not attracted much attention, perhaps because until recently they have
been quantitatively unimportant. An alternative possibility is that observers
assume that such regimes cannot last for long and will end badly, because
of the evidence provided by emerging economies with chronic deficits.
Assumptions are necessary because past empirical work on crises and cur-
rent account reversals has considered only episodes identified by large
depreciations or swings in current accounts from deficit to surplus.12
The theoretical symmetry between speculative attacks on undervalued
currencies and those on overvalued currencies is well known.13 In an attack
on a strong currency, anticipated capital gains generate private capital inflows
when speculators believe the regime can be overwhelmed. Intervention to
limit nominal appreciation takes either of two forms, both with unfavorable
side effects: an increase in the monetary base, which raises the domestic
price level, or sterilization, which increases reserve assets and the govern-
ment’s domestic currency liabilities. The regime can appear to be stable for
a time, but the government’s tolerance for inflation or reserve accumulation
is limited, and a speculative attack will bring the regime to an end.

Data Methods: Identifying Precedents

To identify historical precedents for today’s surplus economies, we first
identify sequences of reserve accumulation that might provide a typical
pattern for emerging economies that accumulate net reserves for an extended
period. We then examine the behavior of other variables in the years dur-
ing and after the accumulation sequence.
For a sample of 115 developing and industrial economies, we examine
yearly data from 1970 to 2004. We first identify sequences of consecutive
years in which the economy experienced current account surpluses on
average and the government increased its net foreign asset position. For
surplus economies the change in the government’s net foreign asset position
is usually dominated by changes in international reserve assets, but our
measure of net reserve accumulation also includes changes in government
debt and other official sector capital flows. We are interested in the con-
solidated government contribution to financing the change in national net
foreign assets or its mirror image the current account balance. To further
restrict attention to episodes in which the government was an important

12. Frankel and Rose (1996); Razin and Milesi-Ferretti (1998).

13. Grilli (1986).
Michael Dooley and Peter Garber 163
Figure 5. Real Exchange Rate and Current Account Balances during and after
an Episodea

Real exchange rate

Indexb

160
140
120
100 100.94 102.23 104.09 102.89 101.14 97.54
80
60
40
20

Current account–GDP ratio

Percent

10
5
0.73 0.47 0.18 0.05
0
–2.06 –0.08
–5
–10
–15

–3 –2 –1 0 1 2
Year

Source: International Monetary Fund, International Financial Statistics; International Institute of Finance; World Bank, World
Development Indicators; Organization for Economic Cooperation and Development.
a. Year zero refers to the first year after the end of an episode. An episode is defined as a sequence of years (three or more) in
which the official sector increases its stock of international assets, runs a current account surplus (on average), and generates more
than 25 percent of the change in national net foreign assets. Each observation denotes the real exchange rate or current account ratio
of an economy at the specified point in its episode. Numbers in italic are means.
b. The index equals 100 three years before the end of an episode.

participant in international financial markets, we exclude episodes during

which the government generated less than 25 percent of the change in
national net foreign assets.
The typical experience of surplus economies during the three years
before the end of a sequence of net reserve accumulations and the three
years that followed is summarized in figure 5. Definitions and data sources
 164 Brookings Papers on Economic Activity, 1:2005

are reported in appendix A, and detailed data for all accumulation episodes
in appendix C.
During periods of reserve accumulation, several regularities stand out.
First, with very few exceptions, the current account begins in surplus, and the
surplus increases during the period of net reserve accumulation. The average
increase in the surplus was 0.73 percent of GDP in year t − 3, 0.47 percent in
t − 2, and 0.18 percent in t − 1, the final year of net reserve accumulation.
In t − 0, the first year following the sequence of reserve accumulation,
the current account surplus declines by an average of 2.06 percent of GDP.
This certainly suggests that some important shock has occurred. Moreover,
the shock typically persists, with little further change in the current account
balance in the two years that follow.
In the final three years of net reserve accumulation, the currency typi-
cally appreciates in real terms each year; the average cumulative increase
(we define the real exchange rate such that an increase represents an appre-
ciation) is about 4 percent. The behavior of the exchange rate before the
official sector leaves the market is not surprising and is fully consistent with
the conjecture that the growing current account surplus, appreciation of the
currency, and reserve accumulation reflect a growing fundamental disequi-
librium in the real exchange rate and the current account. This sequence is
supposed to end with a jump (appreciation) in the real exchange rate and a
gradual decline in the current account balance. Instead, in the average case,
the government retreats from the market, the currency depreciates in real
terms in the following year by 1.2 percent,14 and the depreciation continues
for two more years. The real exchange rate ends more than 3 percent below
its level five years earlier, and the government enjoys a substantial capital
gain on its reserve accumulation.
The behavior of the macroeconomic variables when the government
stops accumulating net reserves clearly does not fit with the standard model
of a speculative attack on a strong currency. With the government out of
the market, there is a “sudden start” of private capital inflows, as the model
predicts. But these inflows are associated with a persistent real deprecia-
tion of the currency, not an appreciation. Economic growth is above trend

14. At the end of the original Bretton Woods system, Germany experienced a three-
year episode of reserve accumulation and currency appreciation followed by depreciation
in 1974. In this case, however, the cumulative appreciation was larger than the depreciation
in the following and subsequent two years. The usual assumption that reserve accumula-
tion ends with large capital losses on reserves is probably influenced by this episode.
Michael Dooley and Peter Garber 165

during the reserve accumulation episode and generally moderates there-

after. In most cases growth remains positive and recovers in a year or two.
A variety of shocks to the world economy could account for these empir-
ical regularities. For example, a decline in foreign demand for the country’s
exports could explain the swing in the current account and the decline in
growth. The deterioration in the national net foreign asset position is con-
sistent with a decline in the real exchange rate. Indeed, such a decline may
create the expected yield differentials necessary for the sudden start in
capital inflows. The deterioration of the national net foreign asset position
would be consistent with a real depreciation.
Another possibility is that intended or unintended financial liberalization
allows residents to diversify away from domestic assets toward interna-
tional assets. For example, if China suddenly opened its capital markets,
residents’ desire to diversify into foreign currency assets would suggest a
depreciation of the renmimbi rather than the appreciation currently expected.
In this context it would be fully rational for the authorities to build a stock
of dollars now in anticipation of private demand when financial markets
are liberalized. Moreover, it makes no sense to allow the renmimbi to
appreciate now only to depreciate sharply later.

China, Japan, and Korea

Seven economies accounted for two-thirds of worldwide international
reserve holdings at the end of 2004 and for three-quarters of the $600 bil-
lion growth in international reserve assets in that year. Three economies—
Japan, China (excluding Hong Kong), and Korea—held 45 percent of the
global total and acquired 60 percent of the 2004 increase. The general
sequence of current account imbalances, reserve gains, growth, and real
exchange rate changes described above holds even more clearly for this
group of economies. Several are in the midst of a stretch of reserve accu-
mulation today and have experienced two or three similar episodes in the
past. All seven have in the past experienced unusually long episodes of
net reserve accumulation relative to our complete sample: the champion
to date is Singapore with its twenty-seven-year run from 1974 through 2000.
It seems particularly relevant, in evaluating how the current episode of
reserve accumulation might end, to look at the previous experiences of
these seven economies. In this section we review the recent experience
of the “big three.” For each we present charts comparing, for each reserve
 166 Brookings Papers on Economic Activity, 1:2005

accumulation episode, current account deficits and net reserve accumula-

tion on the left-hand side, and economic growth and the real effective
exchange rate on the right-hand side. Appendix B provides similar figures
for the other four economies: Hong Kong, India, Singapore, and Russia.
Japan has experienced three extended episodes of net reserve accumu-
lation in the post–Bretton Woods era: a three-year sequence from 1986 to
1988, a five-year sequence from 1992 to 1996, and a six-year sequence
that began in 1999 and continued through 2004 (figure 6). What might the
first two episodes suggest for the current episode in Japan? We look first
at the current account. In each episode the current account surplus (in bil-
lions of dollars) is growing before the net reserve accumulation begins. The
surplus moderates during the accumulation, then declines in the year before
and the year in which the accumulation ends. In the first episode reserve
accumulation accounted for about 28 percent of the current account surplus
during the period of reserve growth, and in the second about 26 percent.
In the present episode the current account surplus has continued to grow.
Reserve accumulation absorbed about half of the surplus through 2004.
Considering just these data, one might expect a moderation in the rate of
reserve accumulation going forward, but not an end to the sequence of net
reserve gains.
In the two previous episodes, the real effective exchange rate and the
growth rate behaved as described above for the typical experience. The
exchange rate rose before and during the reserve accumulation but then
fell sharply for two or three years. In the first episode the real exchange rate
fell in the year following the end of the accumulation and in 1986, the year
the authorities withdrew from the foreign exchange market. As is typical
for the larger sample, in both episodes the GDP growth rate rose during
the accumulation sequence and then turned down, for three years in the first
episode and two years in the second.
During the recent episode both the real exchange rate and the growth rate
have departed from the norm. Growth increased in the first year of the recent
episode, and the real effective exchange rate rose as would be expected, but
growth collapsed in 2001 and 2002, and the real exchange rate fell. Since then
there has been a recovery in output and a small rise in the real exchange rate.
If history is a reliable guide, reserve accumulation in this episode will
moderate relative to the current account surplus but will continue until there
is a significant decline in the surplus. Meanwhile the real exchange rate
will continue to rise, but at a moderate rate, and output growth will
Michael Dooley and Peter Garber 167

Figure 6. Japan: Current Account Balance, Net Official Assets, Exchange Rates, and
GDP Growth in Three Episodes

1986–88
Billions of dollars Percent Index

8 Exchange rate 110

150
(right scale) 90
6
100 Current account 70
4
50 50
2 GDP growth
(left scale) 30
0 0
Net official assets 10

1983 1985 1987 1989 1983 1985 1987 1989

1992–96

Exchange rate
8 (right scale) 110
150 Current account
90
6
100 70
4
50 50
2 30
0 Net official assets 0 GDP growth 10
(left scale)
1991 1993 1995 1997 1991 1993 1995 1997

1999–2004

8 Exchange rate 110

150 (right scale)
Current account 90
6
100 70
4 GDP growth
50 Net official 50
(left scale)
assets 2
30
0 0 10

1998 2000 2002 2004 1998 2000 2002 2004

Source: International Monetary Fund, International Financial Statistics; International Institute of Finance; World Bank, World
Development Indicators; Organization for Economic Cooperation and Development.
a. Year in bold refers to the first year after the end of an episode. See footnote a of figure 5 for the definition of an episode.
 168 Brookings Papers on Economic Activity, 1:2005
Figure 7. China: Current Account Balance, Net Official Assets, Exchange Rates, and
GDP Growth in One Episode

Billions of dollars Percent Index

12
200 Net official assets GDP growth 130
10
(left scale)
150 8 120
6
100 110
4
50 2 100
Exchange rate
Current account 0 (right scale)
0 90
–2

1998 2000 2002 2004 1998 2000 2002 2004

Source: International Monetary Fund, International Financial Statistics; International Institute of Finance; World Bank, World
Development Indicators; Organization for Economic Cooperation and Development.
a. See footnote a of figure 5 for the definition of an episode.

continue to improve slowly. When the current account deteriorates,

reserve accumulation will end and the real value of the yen will fall.
China has the longest continuing sequence of net reserve accumulation
in our sample. From 1990 to 2001 small current account surpluses were
roughly matched by reserve accumulation, with little participation by the
domestic private sector in international financial markets (figure 7). Since
then the current account surplus has grown rapidly, and reserve accumu-
lation has consistently been about double the surplus, as large net inflows
of direct investment have been matched by reserve accumulation. Clearly,
the reserve buildup since 2002 is unusual by historical measures. We have
not seen a sequence of private capital inflows financing reserve accumula-
tion on anything like this scale before.
Nor, as the right-hand panel shows, have we yet seen any of the predicted
precursors of a successful speculative attack. The real exchange rate fell
until 1999, when the nominal rate was fixed. Since then the rate has moved
with the dollar, falling by about 4 percent from 1999 through 2003. Recall
that this is a real effective rate, so that the standard model would predict a
gradual erosion of the authorities’ ability to control inflation. We have seen
no evidence of this to date.
Finally, high growth rates during the sequence of current account sur-
pluses are clearly a feature of this history. As in the case of Japan, our
Michael Dooley and Peter Garber 169
Figure 8. Korea: Current Account Balance, Net Official Assets, Exchange Rates, and
GDP Growth in Two Episodes

1986–89
Billions of dollars Percent Index

40 150
10
30 140
8
20 Current account 130
10 6
GDP growth 120
0 4 (left scale)
Net official 110
–10
assets 2 Exchange rate 100
–20
0 (right scale)
–30 90

1982 1984 1986 1988 1990 1982 1984 1986 1988 1990

1998–2004

40 10 150
30 GDP growth 140
8 (left scale)
20
130
10 6
Current account 120
0 4
110
–10
2 100
–20 Net official
0 Exchange rate
–30 assets 90
(right scale)
1998 2000 2002 2004 1998 2000 2002 2004

Source: International Monetary Fund, International Financial Statistics; International Institute of Finance; World Bank, World
Development Indicators; Organization for Economic Cooperation and Development.
a. Year in bold refers to the first year after the end of an episode. See footnote a of figure 5 for the definition of an episode.

reading of history is that the reserve accumulation will continue until the
current account surplus turns around for other reasons. In China’s case an
interruption of direct investment inflows or liberalization of capital outflows
might generate a real depreciation and an end to the sequence of reserve
accumulations.
Korea experienced one sequence of net reserve accumulations that meets
our criteria from 1986 to 1989, and a second episode started in 1998 and
continues today (figure 8). During the earlier episode, reserve accumulations
 170 Brookings Papers on Economic Activity, 1:2005

roughly matched an increasing current account surplus and came to an end

when the surplus declined. The real exchange rate rose in the final two
years of the episode and declined the year following and for the next two
years. In the familiar pattern, growth slowed in the final year of accumu-
lation but then rebounded for the next two years.
The episode that started in 1998 is not unusual. Reserve accumulation
has approximately matched a U-shaped sequence of current account sur-
pluses, and the real exchange rate has risen.

Summary of Findings
To conclude, we have looked at a large body of data to evaluate the rel-
evance of the standard model for understanding developments in emerg-
ing economies with chronic current account surpluses since 1970. We find
almost no support for the standard model, which predicts an eventual spec-
ulative attack on a strong currency. Episodes of net reserve accumulation
coincide with growing current account surpluses. Reserve accumulations
end when the current account surplus declines or (as often happens) swings
all the way into deficit. Most important, the real exchange rate weakens at
the end of accumulation episodes, and there is generally a small downturn
in economic activity. Such a sequence is consistent with a variety of real
and financial shocks to the surplus economy. But a real depreciation fol-
lowing the authorities’ decision to stop accumulating reserves is not con-
sistent with a speculative capital inflow or a successful speculative attack.
Recall that, in the standard model, the regime ends with a burst of inflation
or a forced nominal appreciation of the currency, either of which would be
associated with a real appreciation. We do observe “sudden starts” of private
capital inflows to finance a current account deficit, but these are associated
with a falling real value for the currency, presumably to generate increases
in expected yields that draw private capital into the economy.
Let us reemphasize what we did not find in the data. We did not find
sequences of reserve accumulation followed by revaluations that generated
capital losses for the government.15 We did not find sequences of reserve
accumulation followed by recessions generated by a real appreciation of

15. Calculations of such book losses have become a central arithmetical exercise
among those issuing dire warnings and calling for an end to the system. See, for example,
Roubini and Setser (2004) and Eichengreen (forthcoming).
Michael Dooley and Peter Garber 171

the currency. This history suggests to us that the contemporary pattern of

current account surpluses can continue in these economies until there is a
major negative shock to demand for their exports. A cyclical downturn in
the United States might be a likely candidate.

Nothing Lasts Forever

The historical record we have presented suggests that most current

account surplus regimes have not been terminated by speculative attacks.
If this interpretation is correct, there is no obvious constraint on the abil-
ity of existing surplus regimes to continue to finance a current account
deficit in the center country. A common theme in international finance is
that repressed systems do not last forever. We agree they last for no more
than twenty years and probably less, but the important point is that they
are effective for substantial periods.
Of course, what countries can do tells us nothing about what they will
want to accomplish. They could listen to the eminent advice and join the
Washington consensus and the international finance textbooks by importing
capital and developing internally. Our Revived Bretton Woods argument
suggests that they will want to do just the opposite. That is, the governments
of trade account countries will want to lend to the rest of the world and,
in particular, to the center country. And they will counter efforts by the
domestic private sector to export capital, through controls and sterilized inter-
vention. An important part of our story is that the real exchange rate distor-
tion will decline over time and vanish at the end of the adjustment period.
So the big speculative incentive is front-loaded, and the beginning of a
reserve accumulation episode is precisely the time in an emerging econ-
omy’s history when financial repression is most likely to be effective. An
important constraint on capital inflows into China is the underdeveloped
and bankrupt domestic financial market. As the industrial sector grows and
that sector lobbies for a better domestic financial system, the whole fabric
of financial repression will unravel. But this takes time.
In our framework the Chinese government is not accumulating reserves
because of a mindless infatuation with a fixed nominal exchange rate. It is
instead using a real undervaluation of its currency to limit urban migration
and to subsidize rapid industrialization and absorption of unemployed
labor. So, at the end of the process, the government anticipates holding a
 172 Brookings Papers on Economic Activity, 1:2005

stock of dollar reserves that may or may not generate a capital loss. Clearly,
if, as is typical, the renminbi depreciates in real terms, there is no capital
loss in the endgame. In any case the government anticipates having by
that time a physical capital stock that is larger and more productive than
today’s and a labor force that is employed and paying taxes. The one is the
prerequisite for the other. The government’s portfolio of interests includes
the domestic capital stock as well as foreign exchange reserves; the value
of that portfolio should not be maximized locally over its individual sub-
components.

That Old-Time Religion, It’s Good Enough for Me: It Is 1958

The financial press and several widely quoted experts have argued that
our comparison of the current international monetary system to the Bretton
Woods system is problematic. In particular, they point out that the United
States did not finance a large and persistent current account deficit under
Bretton Woods, and indeed the mere forecast of such a deficit in the late
1960s was enough to bring the system to a painful end. In addition, unlike
in the original Bretton Woods system, there is now a viable alternative
reserve currency, the euro, and there are no formal arrangements to pre-
vent reserve diversification. We argue below that this is a misreading of
the nature of the system then and now and of the forces that brought the
Bretton Woods system to an end.

The Old-Time Religion: Balance of Payments Deficits Are Not

Current Account Deficits
During the Bretton Woods years, the United States did not run large
current account deficits, the measure of external imbalance that most draws
our attention today. But, in the reckoning of the day, it did run large and
persistent balance of payments deficits. The definition of an external deficit
that was natural to economists and policymakers at the time seems today
to have been forgotten or to be treated as a curious and outmoded accounting
convention. Almost all the old-timers focused on a liquidity definition of
the balance of payments, which Ragnar Nurkse explained as follows:
A country with a deficit in its balance of payments can cover the deficit either
by an outflow of gold or an inflow of foreign short-term funds. . . . These funds
are equivalent to a loan by foreigners and should be regarded as a draft on the
Michael Dooley and Peter Garber 173

recipient countries stock of international reserves. . . . The foreign short term

funds are a liability, can be withdrawn at any moment, and must be treated as a
negative gold reserve.16
Notice that this definition implicitly adds elements of the capital account,
namely, the balance of trade in longer-term assets, to the current account
in order to define a payments imbalance. It emphasizes strictly net flows
of gold and short-term claims, that is, liquidity, in defining the balance of
payments. Two generations of students of international economics have
been kept in the dark about this concept or, at most, trained to think of it as
an odd creation of the old-timers, mentally straitjacketed by the completely
controlled economies of their day. Yet there it is in the literature: they
harped continually about the growing U.S. balance of payments deficits.
For instance, in his valedictory on the old international monetary system,
French president Charles de Gaulle said
. . . . But in addition, the fact that a large number of countries accept, out of
principle, dollars in the same way as gold to compensate, when appropriate, any
deficits that arise to their advantage from the American balance of payments,
leads the United States to become voluntarily indebted to foreign countries. . . .
instead of paying them totally in gold, the value of which is real, that you can
only possess if you have earned it and that you cannot transfer to others without
risk and without sacrifice. . . .
The United States, for want of having necessarily to pay in gold, at least
totally, for their negative balances of payment in accordance with the old rules,
that required countries to take the required steps, sometimes rigorously, to rem-
edy their imbalance, is suffering year after year from a deficit balance. No less
because the total of their commercial exchanges is to their disadvantage. Quite
the opposite! Their material exports always exceed their imports. But that is
also the case for dollars, exports of which are always in excess of imports. In
other words, capital sums are being built up in America, by means of what
should really be called inflation, which, in the form of dollar loans granted to
countries or to private individuals, are being exported. As, in the United States
itself, the increase in currency circulation that results from this makes invest-
ments within the country less remunerative, there is an increasing trend there to
invest abroad. This leads, for certain countries, to a sort of expropriation of
some of their companies. . . .
But circumstances are such today that we can even wonder how far the problem
would go if the countries that hold dollars wanted, sooner or later, to change
them into gold? Although such a general movement would never take place, it
is still the fact that there is an imbalance that is, to a certain extent, fundamental.17

16. Nurkse (1945, p. 3).

17. Press conference at the Palais de l’Elysée, February 4, 1965.
 174 Brookings Papers on Economic Activity, 1:2005

The old fundamentalists said there was a balance of payments problem.

The modern secularists say there was not, because the current account was
in surplus. So what brought about this change? A change in definition.

The Modern Secular View: Yes, They Are

The intertemporal maximization model of the international monetary
system found in most modern textbooks assumes that the system is based
entirely on trust and freely flowing capital. Private international capital
transactions dominate and therefore undo official interference. Such trans-
actions are based on the assumption that debtors willingly repay creditors,
and those who suffer capital losses willingly repay those who enjoy capital
gains without the imposition of infrastructure to secure this result. Observed
net and gross capital transfers are interpreted as private intertemporal trade
in goods and services. Boiled down to this dimension, goods and services
should flow, on net, from high-income, slow-growing economies to low-
income, fast-growing economies so that consumption can be smoothed over
time. This flow imbalance can be sustained for a long time and reach high
levels because it can be repaid later with surpluses that come from rapid
growth. Trust is all that is needed.
That this theory generates more puzzles than insights is problematic but
has not hindered its dominance. For example, an inconvenient parallel lit-
erature on sovereign debt has difficulty concluding that anyone should
repay international debt, yet we somehow reconcile ourselves to this con-
tradiction in two basic traditions in international finance.

Which Is More Realistic, Collateral or Trust?

A unifying conceptual basis for both the original and our Revived Bretton
Woods system is the idea that the international monetary system was and is
based on collateral, not on trust. Nurkse and his contemporaries believed the
international monetary system depended on countries’ willingness and abil-
ity to deliver gold on demand. A country’s ability to deliver gold could be
instantly reduced by calling its short-term credits. It follows that the liquid-
ity balance was the natural measure of the change in the position of gov-
ernments, including the government of the center country.
It is our contention that the current system also runs on collateral, not on
trust. International net saving transfers are too small (except to the United
States) because no one trusts a net debtor (the Feldstein-Horioka puzzle).
Michael Dooley and Peter Garber 175

Gross two-way trade in assets is too small because no one trusts a poten-
tial loser (the home bias puzzle).
In the original Bretton Woods system, the United States was able to
provide intermediation services to the world because it posted a stock of
collateral in the only form that was acceptable at that time, that is, gold.18
Nurkse was right that the ability of the United States and other countries
to participate in international markets was limited by the stock and distri-
bution of gold. A similar implication of our view of collateral in the Revived
Bretton Woods system is that a country that wants to participate in private
international intermediation has to post collateral. In 1949 the United States
had, as it were, the only triple-A credit rating in the system, and so it could
hold its own collateral. As de Gaulle pointed out, however, this was no longer
the case in 1965, when liquid claims on its collateral were substantial.
The key idea in our analysis of the current system is that “earned” U.S.
dollar reserve assets have replaced gold as the ultimate reserve asset. The
only collateral “asset” that everyone trusts are goods already delivered to
the United States by other countries. These goods come to the United
States via U.S. current account deficits. Everyone trusts the United States
to keep these goods or, what is the same thing, to “default” on U.S. official
liabilities to selected foreign governments if those governments steal the
private assets of U.S. residents or others, especially in the context of a
geopolitical bump.
In this sort of default, the Treasury does not cease paying on its own
obligations owned by the problematic foreign government. In practice, it
has in the past frozen assets, converting them from liquid to completely
illiquid claims, placed service payments into blocked accounts, forced long-
term rollovers at Treasury bill rates, and redefined the ultimate claimants
and recipients of these payments in legal cases, which may emanate from
ex post legislation.
Moreover, as in Nurkse’s explanation above, a country cannot usefully
borrow reserves. It would then have nothing to lose, since it could simply
default on its liability.19 In our view reserves and other official or even pri-
vate foreign-held assets are collateral only if they have been earned by net

18. The United States did sit on its own gold reserve, and so there was some trust even
in this arrangement. It also sat on a large chunk of everyone else’s gold.
19. That is why Argentina’s reserves are not collateral, but rather loans from the Inter-
national Monetary Fund. To seize them would be to seize the IMF’s capital.
 176 Brookings Papers on Economic Activity, 1:2005

exports of goods and services. If they are not so earned, they are by defin-
ition borrowed. If borrowed, there are no already-delivered goods for the
United States to keep, and there is hence no collateral.
Critics of the collateral approach argue that the U.S. Treasury would
never damage its reputation by defaulting on an official reserve liability.
We have two reactions. First, the Treasury has frequently done so in the
past. Several such actions are described below, along with a more detailed
history of a recent case. Second, we argue that transferring collateral to
the rightful owner in circumstances envisioned in the collateral relation-
ship preserves the reputation of the U.S. government both as a debtor and
as an impartial and reliable enforcer of collateral arrangements. In deliv-
ering its liability to the injured party, the United States is not defaulting on
its obligations. It is honoring both its promise to pay and its promise to pay
the rightful owner of its obligation. The identity of the rightful owner is
conditioned by the terms of the collateral arrangement. Both reputations
contribute to the demand for U.S. international reserves. But an important
implication of our approach is that the second of the dual roles, that of
enforcer of collateral arrangements, is the only unique function of an inter-
national reserve currency.
In a private collateral arrangement, the rights and obligations of the
participants are clear and explicit. The rights and obligations of govern-
ments in the collateral arrangement we have described are implicit and
necessarily less clear. For example, the event that would trigger transfer
of ownership of U.S. official liabilities is not defined, as it would be in a
private collateral arrangement. But historical precedents exist. The United
States has transferred ownership following major geopolitical incidents such
as wars, invasions, revolutions, hostage takings, and nationalizations of
foreign investment. That there is uncertainty about what set of events would
trigger transfer of collateral does not mean that there are no such events or
that private investors do not value the protection offered by collateral in
those circumstances.
There is also uncertainty concerning what set of creditors to a country
would actually benefit from collateral arrangements. But even a random
distribution among creditors would be a significant disincentive for a sov-
ereign on the international periphery considering whether to seize assets,
provided it had enough collateral at risk. Uncertainty about what events
will trigger transfer of collateral and uncertainty about the distribution of
Michael Dooley and Peter Garber 177

the transfer make governments’ collateral less powerful than private col-
lateral. Our conclusion is that more of it is needed to support a given scale
of financial intermediation.
Ricardo Caballero and A. Krishnamurthy have similarly argued that
international collateral is necessary to support private financial intermedi-
ation within advanced and emerging economies.20 They also emphasize
that an important market failure in emerging economies is the inability to
produce assets that can be used as collateral, making it necessary to import
such assets. Caballero elsewhere relates this to the private financing of the
U.S. current account deficit as follows: “There is an enormous demand for
saving instruments in the world, and the US is the most efficient producer
of such instruments. No other place combines the volume from new oppor-
tunities and ability to generate trustworthy saving instruments from each
unit of physical investment put on the ground.”21 An important aspect of
their analysis is that financial crises can reduce the supply of collateral
assets in emerging economies, and that this might constitute the real costs
of such crises. Moreover, even developed financial markets can lose their
ability to produce safe assets following a severe financial crisis like that
which has plagued Japan in recent years.
We are just beginning to explore the economic significance of private and
official holdings of international collateral and how the two might interact. Is
private collateral a substitute for official holdings of safe assets? Is official
collateral necessary for the credibility of cross-border private collateral
arrangements? Our framework is based on the idea that official collateral is
required because, when trouble comes, private international credit arrange-
ments are enforced, if at all, by governments. There is, of course, ample
room for clarification and improvement of our understanding of these mech-
anisms. But two things seem to us clear. The United States is a source of safe
assets that cannot be produced locally in most of the rest of the world. And,
since borrowed collateral is an oxymoron, most of the rest of the world has
to earn these assets by delivering goods to the United States.
Could Europe, offering the euro as an alternative reserve currency, re-
place the United States as the preferred custodian of collateral? Clearly this
is possible. As many observers have recently pointed out, the European

20. Caballero and Krishnamurthy (2001, 2003).

21. Caballero (2004).
 178 Brookings Papers on Economic Activity, 1:2005

Union already provides euro-denominated government debt that is a credi-

ble promise to pay. Moreover, some diversification from dollars to euros
might make sense in terms of a narrow risk-return calculus. But our con-
jecture is that the dollar will remain the dominant reserve currency as long
as the European Union is less willing or less able than the United States to
enforce collateral arrangements. Since the European Union has no track
record in this regard, it seems unlikely that the euro will soon challenge the
position of the dollar in the international monetary system. For this to change
would require markets to come to expect that some European governments
would be willing to accept large current account deficits and to block the
movement of euro reserves as a way of punishing a country (possibly an
aggressive one) that expropriates foreign assets. In our view both expecta-
tions are unlikely. At this point in history, substituting euros for dollars
places collateral out of the reach of creditors and therefore considerably
reduces its usefulness.
Our approach is based on the view that there is little trust between key
countries in the international monetary system. In such a system, everyone
sees tremendous benefits from international financial intermediation, but
no one can afford the risk of letting another country owe them substantial
amounts of goods. The best risk is the central reserve country. Put another
way, without trust, the stock of net financial indebtedness must always be
less than the stock of collateral that can be seized. In domestic financial
markets the stock of real capital that can be pledged as collateral is large
relative to credit balances. Although collateral is a universal feature of
domestic credit relationships, it is seldom a binding constraint, at least in
the aggregate. In international finance just the opposite is the case. Huge
stocks of national wealth exist but are useless in creating incentives for
repayment, because mass default is often generated by government via
the domestic legal system.

Some Evidence on the Durability of Reserve Currency Status

The International Emergency Economic Powers Act of 1977 (IEEPA,
which supplanted the Trading with the Enemy Act of 1917) empowers the
president of the United States to freeze foreign-owned assets under U.S.
control. The IEEPA authorizes the use of sanctions when the president
sees an “unusual and extraordinary threat” to the “national security, for-
eign policy, or economy” of the United States and declares a national
Michael Dooley and Peter Garber 179

emergency.22 The word “emergency” allows the window to be slammed

shut if, for example, a foreign country threatens to launch a financial
attack by withdrawing funds or to pull out a substantial amount of funds
out in order to prevent their seizure. As described by the U.S. Information
Agency, a freeze on foreign-owned assets
can be applied selectively to a particular country, or to a group of countries, in
time of war or in response to a national emergency. . . .
The procedure can be used to serve three purposes:
—to deny authorities in blocked countries access to assets that might be used
against the US
—to protect the true owners of the assets from illegal attempts to seize their
property
—to create a pool of assets for possible use in settling US claims against blocked
countries, or for use as a bargaining chip in negotiating an eventual return to
normal relations.23
During World War II, assets owned by Germany, Japan, and Italy were
blocked and eventually used in settling war claims against them. Simi-
larly, assets of Hungary, Romania, Latvia, Lithuania, Estonia, Bulgaria,
and Czechoslovakia were blocked after these countries fell under Soviet
domination. Asset blockings were subsequently imposed against North
Korea and China in 1950, Cuba in 1963, North Vietnam in 1964, Rhode-
sia (now Zimbabwe) in 1965, Kampuchea (Cambodia) in 1975, Iran in
1979, Libya in 1986, Panama in 1988, the Federal Republic of Yugoslavia
(Serbia and Montenegro) in 1992, and Afghanistan in 1999. In 1990 the
United States blocked $30 billion in assets belonging to Iraq and Kuwait.
In 1979 it blocked $12 billion of Iran’s assets, including $5 billion in off-
shore branches of U.S. banks; part of this was used to pay off syndicated
loans by U.S. banks to Iran, and $1.4 billion was sent to the Bank of Eng-
land to cover claims in the United Kingdom. Another $1 billion was held
against awards from the Iran-U.S. claims tribunal.24
These asset freezes have occurred under a variety of circumstances. Some
of the asset blockings were aimed at adversaries in a declared or undeclared
war (Germany, Japan, Italy, China, North Korea, and Iraq). Some were aimed

22. See the International Emergency Economic Powers Act (IEEPA), United States Code
(www.treas.gov/offices/enforcement/ofac/legal/statutes/ieepa.pdf). Assets frozen under the
IEEPA are administered by the Treasury’s Office of Foreign Asset Control (OFAC).
23. U.S. Information Agency, “Freeze of Iraq, Kuwait Assets Has Many Precedents,”
August 28, 1990 (www.fas.org/news/iraq/1990/900828-152460.htm).
24. U.S. Information Agency, “Freeze of Iraq, Kuwait Assets Has Many Precedents.”
 180 Brookings Papers on Economic Activity, 1:2005

at friendly countries that had been occupied, with the aim of preserving the
assets pending the restoration of a government recognized by the United
States (Latvia, Lithuania, Estonia, and Kuwait). Some countries saw their
assets blocked when they opposed the United States geopolitically or be-
came hostile without war breaking out (Cuba, Iran). Some freezes were im-
plemented as part of a global imposition of sanctions (F.R. Yugoslavia,
Rhodesia). The differences in circumstances notwithstanding, this history
shows that the center country can repeatedly “default” on official liabilities
and still remain the only important provider of reserves.

Conclusion

The international monetary system must create collateral in order to

support international capital transactions. In the industrial countries, the
lack of such collateral might account for the relatively small net and gross
capital flows among them. Collateral is expensive, and the benefits of
trade in financial assets among similar countries are probably not great,
even though the legal and expropriation risks are relatively small. For
emerging economies, in contrast, the benefits of trade in financial assets
are very large. The irony here is that, to accumulate collateral (or “net
reserves” to the more traditional among our readers), the emerging econ-
omy must export national saving. This is bad from the modern secularist
perspective, but it is orthodoxy in the old-time religion. The benefits of
two-way trade in financial assets are potentially enormous for countries that
have high saving rates but waste the resources thus generated when they
are channeled through inept domestic financial systems. These countries
need to run the modern version of a liquidity surplus.
Some observers have taken a too-legalistic interpretation of our defini-
tion of international collateral. We do not argue that any set of private
investors in an emerging economy would benefit or would expect to ben-
efit from the collection of collateral by the United States, and in the case
of China we have in mind much more the sort of expropriation that might
result from a geopolitical clash. Nevertheless, both U.S. and non-U.S. pri-
vate (portfolio and direct) investors know that an emerging economy that
is an international creditor has something to lose from confiscation of its
investments abroad. It seems clear to us that European direct investors in
Argentina, for example, would have fared much better in recent years if
Michael Dooley and Peter Garber 181

the government of Argentina had owned net assets in the United States. In a
general sense our argument is that the government of an emerging economy
needs a strong incentive to stay out of the way of private international finan-
cial intermediation. Building a positive net international asset position seems
to us the obvious way for it to create that incentive. The real potential
for globalization of international finance lies in governments of emerging
economies posting collateral in the United States to support private two-way
trade in financial assets. The current general move in emerging economies,
in both Asia and Latin America, toward reducing sovereign debt and build-
ing international reserves may be based on an implicit understanding of
how the system really works.

APPENDIX A

Data Sources and Methodological Notes

Episodes of Official Asset Accumulation

We define an episode as a period of three or more years where

—the official sector increases its stock of international assets
—on average the official sector entirely or partly finances the current
account, and
—the official sector generates more than 25 percent of the change in
national net foreign assets.
The second part of the definition is equivalent to the country running
current account surpluses during the episode. As described above, the sec-
ond requirement binds only on average; it is possible to find one or more
observations where the country runs current account deficits, although in
the data this appears very rarely.

Net Official Assets

Net official assets are defined as the sum of the following items in the
“general government” and “monetary authority” accounts in the balance of
payments (all on a net basis): capital transfers, portfolio investment assets
(equity and debt, the latter including bonds, notes, and money market
 182 Brookings Papers on Economic Activity, 1:2005

instruments), financial derivatives, other investment (trade credits, loans,

currency and deposits), and reserve assets.
For the following countries, partial quarterly estimations have been
calculated for 2004: Japan, Pakistan, Russia, and Ukraine (two-quarter
estimations); Denmark, Indonesia, and Korea (three-quarter estimations).

Current Account

Current account data were obtained from the International Financial

Statistics (IFS) of the International Monetary Fund and from the Interna-
tional Institute of Finance (IIF) dataset. All forecasts for 2005 and 2006
come from the IIF dataset. The current account data are expressed as a
flow variable in millions of dollars from the end of one fiscal period to the
beginning of the next.

GDP Growth

GDP growth data were obtained from the World Bank’s World Devel-
opment Indicators and the IIF dataset. All forecasts for 2005 and 2006 come
from the IIF dataset.

Real Effective Exchange Rates

The commonly used definition of the real effective exchange rate is

REER = Π i [( e ei ) ( P Pi )] ,
t wi

where e is the exchange rate of the subject currency against the dollar (in
dollars per subject currency unit), in index form; ei is the exchange rate of
currency i against the dollar (in dollars per currency unit), in index form;
wi is the weight attached to currency i; P is the consumer price index (CPI)
of the subject country; and Pi is the consumer price index of country i.
REER data were retrieved from the IFS, IIF, and Organization for Economic
Cooperation and Development datasets. Data are reported as index values,
where an increase indicates an appreciation of the local currency.

Coverage

The list of countries in the sample is available from the authors. The initial
sample of 164 countries was reduced to 115 because of unavailability of data.
APPENDIX B

Reserve Accumulation Episodes in Hong Kong, India,

Russia, and Singapore
Hong Kong, 1999–2001
Billions of dollars Percent

40 10
30 8
20 Current account 6 GDP growth
10 4
0 2
0
–10 Net official assets –2
–20 –4
1996 1998 2000 2002 1996 1998 2000 2002

India, 1976–79

8 10 GDP growth
8
6 6
4 Current account 4
2 2
0 0
–2
–2 Net official assets –4
1974 1976 1978 1980 1974 1976 1978 1980

Russia, 1999–2004
Index

40 Current 10 GDP growth 150

account 8 (left scale) 140
30
20 6 130
10 4 120
2 Exchange rate 110
0 0
–10 (right scale) 100
Net official assets –2
–20 –4 90
1998 2000 2002 2004 1998 2000 2002 2004

Singapore, 1974–2000

40 10 110
30 8 90
Current account 6
20 70
10 4 Exchange rate
2 50
0 Net official assets 0 (right scale)
–10 30
–2 GDP growth
–20 –4 (left scale) 10
1991 1993 1995 1997 1999 2001 1991 1993 1995 1997 1999 2001

Source: International Monetary Fund, International Financial Statistics; International Institute of Finance; World Bank, World
Development Indicators; Organization for Economic Cooperation and Development.
a. Year in bold refers to the first year after the end of an episode. See footnote a of figure 5 for the definition of an episode.
 184 Brookings Papers on Economic Activity, 1:2005
APPENDIX C

Real Exchange Rate Changes, Reserves, Current

Account Balances, and GDP Growth in Reserve
Accumulation Episodes

Average real Cumulative change

depreciation in reserves
(percent a year) (millions of dollars)

First year First year

Duration During after During after
Episode (years) episode episode episode episode
Argentina, 1976–79 4 6.7 −23.9 8,810 −2,598
Bulgaria, 1982–85 4 2 −26.9 1,053 −885
Canada, 1996–2001 6 −0.2 3.7 19,926 −185
China, 1990–present 15+ −4.5 ... 410,674 ...
Colombia, 1976–80 5 n.a. 4.6 4,287 −21
Colombia, 1989–94 6 2 6.3 4,728 −4
Croatia, 1993–95 3 14.3 0.3 1,652 533
Denmark, 2001–present 4+ 3.3 ... 13,537 ...
Egypt, 1990–94 5 −2.9 4 15,615 409
Finland, 1997–2002 6 −1.2 4.2 3,261 −508
France, 1977–80 4 2.3 −3.7 11,796 −3,588
France, 1983–86 4 1.4 −1.1 10,417 −2,602
Germany, 1971–73 3 4.2 −1.7 18,312 −523
Germany, 1976–78 3 1.2 −4.3 21,011 −3,590
Hong Kong, 1999–2001 3 n.a. −4.4 24,755 −2,377
Hungary, 1990–92 3 7.9 1.1 2,905 2,575
Hungary, 1995–97 3 1.2 0.5 2,983 791
India, 1976–79 4 n.a. n.a. 6,578 −624
Indonesia, 2002–present 3+ 6.4 ... 8,245 ...
Ireland, 1993–95 3 −2.2 −1 4,823 −52
Italy, 1987–89 3 1.4 1.3 25,245 11,623
Japan, 1986–88 3 12.1 −4.4 70,747 −13,058
Japan, 1992–96 5 1.8 1.6 147,110 6,567
Japan, 1999–2004 6+ 0.4 ... 398,985 ...
Jordan, 1999–2003 5 −0.3 −0.2 3,411 n.a.
Korea, 1976–78 3 −13.5 −16.9 3,385 749
Korea, 1986–89 4 2.9 −3 13,036 −1,208
Korea, 1998–present 7+ −0.6 ... 122,894 ...
Malaysia, 1976–80 5 −3 3.7 2,652 −235
Malaysia, 1985–93 9 −2.2 0.4 25,398 −3,160
Malaysia, 2001–03 3 −3 −3 14,838 n.a.
Morocco, 1996–2002 7 0.3 0.3 6,782 1,649
Michael Dooley and Peter Garber 185

Cumulative change in
current account balancea
(millions of dollars)
Real GDP growth (percent a year)
First year
During after Years of deficit During Average for three years
episode episode after episode episode after episode
3,120 −4,774 3+ 10 −1.5
612 −951 2 3 5.1
24,902 14,447 0 4.2 2.6
315,476 ... 0 9.3 ...
1,029 −1,961 3+ 5.4 6
−2,185 −4,527 3+ 3.8 4.4
−413 −1,049 3+ 1.6 5.1
22,033 ... 0 1 ...
−2,937 −1,000 3+ 3.6 5.1
49,845 9,295 0 3.4 1.9
7,592 −4,811 3+ 2.9 1.8
−3,646 −4,446 3+ 1.7 3.8
4,627 9,085 0 3 2.3
16,929 −5,387 3+ 3.7 1.8
27,309 12,596 0 4.7 2.6
1,134 −4,262 3+ −6.2 1.3
−3,451 −2,228 3+ 2.5 4.7
4,441 −1,785 3+ 2.4 5.6
18,676 ... 0 4.1 ...
5,063 2,049 0 6.1 9.3
−22,628 −16,479 3+ 3.3 1.4
249,000 63,000 0 4.5 4.6
551,000 97,000 0 1.5 0.3
743,000 ... 0 1.4 ...
1984 208 0 3.9 5.2
−1383 −4,151 3+ 10 3.8
34,634 −2,003 3+ 9.8 8.1
130,483 ... 0 4.2 ...
130 941 0 8.6 6.4
−6,326 −4,521 3+ 8.1 9.7
27,885 14,400 0 3.3 5.8
2,253 1,593 0 4 3.7 (continued )
 186 Brookings Papers on Economic Activity, 1:2005

Average real Cumulative change

depreciation in reserves
(percent a year) (millions of dollars)

First year First year

Duration During after During after
Episode (years) episode episode episode episode
Netherlands, 1970–77 8 1.5 −1.3 3,902 −770
Netherlands, 1992–95 4 2.2 −1.8 11,349 −5,695
Norway, 1980–85 6 1 1 9,442 −3,211
Norway, 1993–97 5 −0.5 −1.7 14,353 −6,384
Norway, 1999–2003 5 1.7 −4.8 13,393 n.a.
Oman, 1989–92 4 −1.4 −6.6 1,302 −1,058
Oman, 1999–2001 3 −1.7 −4.9 3,483 309
Pakistan, 2001–present 4+ −0.3 ... 10,244 ...
Portugal, 1985–92 8 2.2 −0.4 16,235 −2,848
Romania, 1986–89 4 −2.7 −31.1 1,614 −1,494
Russia, 1999–2004 6 −0.1 7.2 63,731 n.a.
Saudi Arabia, 1979–82 4 1.6 −3.3 11,432 −1,509
Singapore, 1974–2000 27 −0.3 −2.1 81,135 −861
South Africa, 1989–91 3 2.9 −1 2,030 503
Sweden, 1970–73 4 −0.4 0.2 1,499 −753
Sweden, 1998–2000 3 −2.7 −0.1 5,306 −1,048
Switzerland, 1982–87 6 2.1 −0.3 10,189 −2,382
Ukraine, 1999–present 6+ −4.7 ... 5,383 3,092
Venezuela, 1979–81 3 11.3 −5.9 7,839 −8,160
Sources: International Monetary Fund, International Financial Statistics; International Institute of Finance; World Bank,
World Development Indicators; Organization for Economic Cooperation and Development.
a. A positive number indicates a change in the direction of surplus.
n.a., data not available; . . . , not applicable.
Michael Dooley and Peter Garber 187

Cumulative change in
current account balancea
(millions of dollars)
Real GDP growth (percent a year)
First year
During after Years of deficit During Average for three years
episode episode after episode episode after episode
16,476 −904 1 3.6 2
63,117 21,502 0 2 3.7
11,736 −4,551 3+ 3.4 1.9
33,520 6 0 4.6 2.5
113,812 n.a. 0 1.9 n.a.
563 −1,190 3+ 6.7 4.6
5 2 0 4.3 3.2
9,277 ... 0 5.3 ...
−14 233 0 4.6 1.1
9,874 −3,254 3+ −0.8 −9.1
199,007 40,800 0 6.8 5.2
98,912 −16,852 3+ 2.5 −5.2
116,349 16,137 0 7.5 0.7
5,734 1,967 0 0.4 0.8
2,149 −529 3+ 3.4 2.3
29,648 8,531 0 4.2 1.5
26,869 8,846 0 1.9 3.7
10,606 ... 0 8.3 ...
9,078 −4,246 1 −1.3 −1.5
Comments and
Discussion

Barry Eichengreen: The first rule of forecasting is, “Give them a fore-
cast or give them a date, but never give them both.”1 Michael Dooley and
Peter Garber have given us a forecast, namely, that the dollar will fall
and U.S. Treasury yields will rise. Bravely, they have also given us a date.
Unfortunately for those of us interested in the future, that date is 1971.
Like Dooley and Garber, I agree that what cannot go on forever gener-
ally will not. But unlike them I do not believe that recent events in finan-
cial markets can help us pin down the timing. The middle of March, just
before the Brookings Panel meeting, saw an increase in noise about the pos-
sibility that foreign central banks might diversify out of dollars. The gov-
ernor of the Bank of Korea made some widely reported comments about
the need for more-active reserve management. Prime Minister Junichiro
Koizumi of Japan told a parliamentary committee that reserve diversifi-
cation was “necessary.”2 Y. V. Reddy, governor of the Reserve Bank of
India, said that the diversification of reserves was under active discussion.3
Ukrainian economy minister Sergiy Teriokhin argued publicly that the
country should diversify its reserves out of dollars and into euros.4 This
upsurge in noise was associated with an eight-month high in Treasury yields,
reinforcing the belief that reserve diversification could eventually force
the dollar down and Treasury yields up.
At the same time, that eight-month high in Treasury yields was not all that
high. I would acknowledge that this is a troubling point. I am not alone, of
1. Attributed to Edgar R. Fiedler by Dickson (1978).
2. Steve Johnson, “Dollar Wobbles on Japan Diversification Talk,” Financial Times,
March 10, 2005.
3. Reuters, “Asian Foreign Exchange Reserves: A $2.46 Trillion Question,” March 11,
2005, 11:36 AM.
4. Reuters, March 16, 2005, 12:50PM.

188
Michael Dooley and Peter Garber 189

course: Federal Reserve Chairman Alan Greenspan has commented on this

issue extensively, to the point where it is now known as the Greenspan
conundrum. Factors invoked to help explain it include the relatively short
supply of new long-term Treasury debt coming onto the market as the debt
managers at the U.S. Treasury shorten maturities, and the inelastic demands
of various institutional investors for government securities. In Dooley and
Garber’s view, the proper interpretation is that financial market partici-
pants are attaching a positive probability to Asian central banks continuing
to support the dollar by making massive purchases of Treasury bonds. This
is the substance of the first of the authors’ three notes.
Who am I to second-guess the markets, much less to second-guess our
esteemed authors? Well, I’m an economic historian who can recall a sub-
stantial number of previous episodes where major imbalances leading to
sharp changes in exchange rates were not obviously factored into financial
markets until immediately before the event. For example, the January-
March 1933 run on the dollar, a suggestive precedent, had virtually no dis-
cernible impact on interest rate differentials or forward exchange rates
until almost immediately before it occurred, despite the fact that the possi-
bility had been actively discussed for the better part of a year.5 The 1992
attacks on the pound sterling, a currency that commentators regularly cited
as ready for a fall, were similarly not preceded by the emergence of notice-
able interest rate differentials or a forward discount in the foreign
exchange market until a couple of weeks before the denouement.6 Particu-
larly interesting, given the context, is that in 1968–71, in the run-up to the
collapse of the Bretton Woods system, the forward discount on the dollar
was very modest, as was the interest rate differential between the United
States and Germany.7 Then, in the summer of 1971, the forward discount
jumped upward. Although one can always ascribe such behavior to the
arrival of new information, it is not as if people failed to see the collapse of
the Bretton Woods system and a substantial devaluation of the dollar com-
ing. To the contrary, there was an immense contemporary literature warn-

5. See Hsieh and Romer (2001).

6. See Eichengreen and Wyplosz (1993).
7. See Obstfeld (1993). Imperfect capital mobility complicates the drawing of infer-
ences from these data, although, as Obstfeld notes, capital mobility was rising strongly
toward the end of the Bretton Woods period. Imperfect capital mobility also complicates
drawing inferences from interest rates today, as Dooley and Garber themselves note.
 190 Brookings Papers on Economic Activity, 1:2005

ing that the system would dissolve and that the dollar would have to fall
substantially. Yet there seemed to be a striking reluctance to take a posi-
tion on this basis until one minute before the clock struck midnight. This
behavior may be hard to reconcile with perfect foresight, but, if it exists,
asset prices will not be telling us when it is 11:58.8
Dooley and Garber’s second note is an analysis of the end of several
episodes of large current account surpluses accompanied by reserve accu-
mulations. The authors’ intriguing finding is that many such episodes come
to an unhappy end with a sharp real and nominal depreciation, which is
not how most observers of China expect the current situation to play out.
I would simply observe that how ancillary variables like the real exchange
rate will react when China’s accumulation of reserves ends will depend
on why it ends. If it ends because the People’s Bank of China and the gov-
ernment, observing robust economic growth and mounting inflationary pres-
sure, choose to tighten monetary policy in order to cool fears of overheating,
the currency will appreciate. But if they wait until domestic financial
excesses further infect the banking system, creating a crisis of confidence,
there may instead be a scramble to get out, causing the currency to crash.
The wisdom of moving away from the peg while confidence is strong,
capital is still flowing in, and reserves are still being accumulated is the
central lesson of the literature on exit strategies.9 It provides the strongest
argument for why China should abandon its peg to the dollar now.
Dooley and Garber’s third note extends an earlier paper of theirs that
characterizes the role of the United States in the current international
system as providing financial intermediation services to the rest of the
world.10 The United States borrows short, indeed increasingly short given
the shorter and shorter tenor of Treasury debt, and lends long in the form

8. It is worth recalling that, in the original Krugman-Flood-Garber model, interest rates

remain perfectly stable until the moment the exchange rate collapses, despite the fact that
everyone has perfect foresight and can see the end coming. I understand, of course, that the
interest rate in their model is the instantaneous rate, not the long-term rate, and that the result
depends on some restrictive assumptions. But add to this the well-known fact that the mar-
kets appear to attach a heavier weight to short-term rates than the term-structure hypothesis
would lead one to predict, and it is possible to reconcile the conviction that the end is near
with the observed low level of long-term rates.
9. See, for example, Eichengreen and Masson and others (1998) and, for an application
to China, Eichengreen (forthcoming).
10. Dooley, Folkerts-Landau, and Garber (2004b).
Michael Dooley and Peter Garber 191

of foreign direct investment (FDI). It thus provides liquidity transformation

services like those of a bank. It enhances the efficiency of resource alloca-
tion, also like a bank. Again, there is a suggestive analogy with the Bretton
Woods system, there having been an argument in that period that the United
States was acting as banker to the world, borrowing short and lending long.
It was similarly argued that the U.S. deficit on liquidity account was not a
problem because foreigners were simply the happy recipients of these
intermediation services.11
But there is a difference between the Bretton Woods era and today,
namely, that the present situation occurs against the backdrop of large,
ongoing current account deficits for the country that is banker to the world.
In principle, there is no reason why the country with the most efficient finan-
cial system, which is therefore providing intermediation services to the
rest of the world, cannot run a balanced current account or, for that matter,
a surplus. There is no reason why importing short-term capital and export-
ing long-term capital should also require it to run a current account deficit,
as the United States is doing. The United States ran current account sur-
pluses following World War II, even after contemporary economists stopped
referring to the dollar gap. Britain ran persistent current account surpluses
before World War I, when it was similarly acting as banker to the world.
Being an international financial center and providing maturity transfor-
mation services to the rest of the world does not doom a country to current
account deficits.
So China must be buying something else through its bilateral surpluses
and heavy investment in U.S. Treasury bonds. According to Dooley and
Garber, it is buying custodial services. The United States, in their view, is
now custodian to the world (this is not a comment on the postindustrial
economy). In other words, the United States holds the collateral against
which countries like China are able to borrow. China can then attract FDI
from abroad, because if it ever decided to nationalize U.S. or other private
assets, the United States would then default on its official liabilities held
by the Chinese government.
This is a provocative hypothesis whose validity is highly questionable.
For one thing, the story is specific to China, whereas the accumulation of

11. See Despres, Kindleberger, and Salant (1966). Others objected to this view on the
grounds that the capital inflow was really the temporary balancing item that offset a U.S.
current account surplus that was too small to fully finance U.S. FDI abroad.
 192 Brookings Papers on Economic Activity, 1:2005

reserves and chronic surpluses in trade with the United States is a pan-Asian
phenomenon. No one worries that Japan, Korea, or Taiwan will expropriate
U.S. investments, yet they, too, hold massive claims on the United States.
I am not aware of any U.S. corporate executives pointing to China’s large
dollar reserves as a form of collateral justifying their decision to invest
there. Nor am I aware of statements by Chinese officials explaining that
they are accumulating Treasuries as a way of posting collateral for FDI
inflows. Dooley and Garber rightly emphasize the importance of looking
at the current international financial system through the eyes and statements
of Asian officials. Here is an instance where their point works against them.
Moreover, the timing is wrong: U.S. FDI in China began to rise around
1992, yet the massive reserve accumulation started nearly a decade later.
Then there is the fact that the United States accounts for only a small frac-
tion of FDI in China: Morris Goldstein and Nicholas Lardy find that it
accounts for less than 10 percent.12 Thus one must assume that the United
States would be willing to go to bat not just on behalf of U.S. private foreign
investors but also on behalf of investors from other countries. In addition,
the way foreign investments in China have been expropriated historically
is through the surreptitious stripping of assets by Chinese managers and
joint-venture partners. It is hard to imagine that the U.S. government would
risk tarnishing its public credit in response to such behavior. Rather, one
has to assume a major geopolitical blow-up between the United States and
China, a decision by Beijing to freeze all U.S. investments there, and the
U.S. government retaliating by freezing all U.S. Treasury bonds held by
China in custody in the United States. Such events are not beyond the
realm of possibility, but they do not exactly strike me as an obvious way
of explaining the current pattern of global imbalances.
This suggests testing the hypothesis with a systematic analysis of the
impact of inward FDI and property rights protection on the demand for
reserves. Specifically, I am imagining a regression of the level of reserves
on FDI, where the coefficient on the latter is expected to be larger in
economies where property rights are less secure, the government is com-
munist, and the country is not politically allied with the United States. Of
course, one would have to control for the other standard determinants of
the demand for reserves and correct for the endogeneity of FDI. But this

12. Goldstein and Lardy (2005).

Michael Dooley and Peter Garber 193

would be a much more direct way of marshalling evidence for Dooley and
Garber’s hypothesis.
If one accepts that the collateral story is implausible or at best unsubstan-
tiated, how then to explain the Chinese authorities’ insistence on keeping
their currency down against the dollar? We are forced to fall back on the
traditional rationale for export-led growth. The export sector, in this account,
is the locus of knowledge spillovers and productivity growth in a devel-
oping economy. Distortions affecting the economy justify the imposition
of another distortion in the form of an undervalued currency, which pushes
more resources into the export sector than would occur under the unfettered
operation of market forces. The original distortion might be the fact that
the productivity effects associated with producing for export are external
to the firm, providing an inadequate incentive for private investors to shift
resources into the sector absent other interventions. Or it might be an inef-
ficient financial system that prevents saving in the developing economy
from underwriting adequate investment in the export sector. Or it might
be a shortage of organizational knowledge that is strongly complementary
with exports and can only be augmented by export-linked FDI that imports
this organizational knowledge from abroad.
Which of these distortions provides the primary motivation for pursu-
ing the export-led growth strategy matters importantly for how quickly one
should expect the government to move away from current arrangements. I
have the sense that Chinese managers and entrepreneurs are rather quickly
gaining the organizational knowledge necessary to run a modern, export-
oriented manufacturing firm. I also have the sense that the productivity
effects from learning by exporting are internal as well as external to the
firm, much as they are in other countries. In other words, the time may
not be very long in coming when these justifications for keeping the exchange
rate artificially low are no stronger in China than in a variety of other
middle-income countries.
The strongest argument in favor of the indefinite maintenance of the
status quo is that the export sector, where productivity is higher than in
the rest of the economy, is being starved of funds by an inefficient Chinese
banking system and that an undervalued renminbi is needed to offset this
distortion, perhaps by artificially boosting the prices of traded goods rela-
tive to nontraded goods and thus enhancing the profitability of investing
in the traded goods sector. But this would boost the prices of traded goods
across the board, whether manufactures or agricultural products, and whether
 194 Brookings Papers on Economic Activity, 1:2005

produced by state enterprises or by private firms. This hardly looks like a

sensible strategy for enhancing efficiency.13 Alternatively, an undervalued
renminbi could lead to China accumulating Treasuries and thus encouraging
efficiently allocated FDI in China. This is the reasoning that forces Dooley
and Garber into the logical corner in which they now find themselves. It
also brings us back to my earlier objections to the collateral-and-selective-
default story.
If one rejects, as I do, their collateral argument, one must hang one’s
hat on the first two distortions: learning spillovers external to the firm, and
shortages of organizational knowledge. It is then hard to resist the conclu-
sion that these justifications for an undervalued currency will grow less
compelling rather quickly, for the reasons I have just enumerated. As the
authors point out, capital losses on dollar-denominated reserves should be
counterbalanced against the returns, and these returns are likely to be
strongly declining. In addition, the liability side of the equation should
include also the other costs of undervaluation, such as the limited extent
of monetary control, the incentive under current conditions for large
amounts of credit to flow into the property market (heightening financial
fragility), and the additional difficulty that this poses for efforts to raise
lending standards and otherwise strengthen the banking system. These
feature not at all in Dooley and Garber’s analysis.
These are all reasons, then, why Asian governments are likely to move
away from the strategy of export-led growth through undervaluation
before long. (But recall the first rule of forecasting.) Once countries see
their neighbors doing so, and thereby lending less support to the dollar,
there will be an obvious incentive not to be late in jumping off the band-
wagon.14 At that point the familiar models of speculative attacks, which
the authors helped to pioneer, will not just be “dancing in our heads.”

13. There is also the question of whether undervaluing relative to the dollar is an appro-
priate strategy for boosting exports in general. The United States takes about a third of
China’s exports (including exports that go via Hong Kong). Even if one aggregates the
other dollar peggers, the share of the “dollar area” is still only 40 percent. Europe mean-
while takes 25 percent of Chinese exports. Over the last ten years the dollar has risen as
well as fallen against the euro. Between 1994 and 2002 it rose very substantially. Was Chi-
nese policymakers’ preference then for increasing overvaluation on an effective basis?
14. As emphasized in Eichengreen (2004).
Michael Dooley and Peter Garber 195

Jeffrey A. Frankel: In a recent series of papers that have deservedly received

a lot of attention, Michael Dooley and Peter Garber (usually with David
Folkerts-Landau) have put forward the view that the essential elements of
the current international monetary situation bear a strong resemblance to
the old Bretton Woods system, with the Asian central banks today playing
the role of dollar accumulators that Europe played in the 1960s. There is a
lot of insight in their views, which are all the more useful for their uncon-
ventional perspective and language. This is not just another recitation of
familiar positions in a fixed-versus-floating debate. For example, I think
the authors are right that “The problem for China is to mobilize its existing
enormous domestic saving to create a growing, internationally competi-
tive capital stock that can rapidly employ hundreds of millions of workers
in productive activity. A serious constraint is the lack of a domestic finan-
cial system capable of channeling this saving into productive capital, tech-
nology, and management skills.” To link these issues to the U.S. balance of
payments situation was an original, imaginative, and thought-provoking
leap.
It is fair to claim, as the authors do, that much of their analysis has now
been generally accepted. This is not to say that one has to credit them for
the insight that Asian central banks are now financing much of the U.S.
current account deficit; that much is obvious. And, at the other end of the
argument, I think the authors have been forced to clarify that they are not
claiming that the current system can continue indefinitely—that the dollar
will never have to depreciate—which many of us thought we heard them
saying at first.1 But I do credit them with the useful insight that the willing-
ness of Asia, and especially China, to pile up unheard-of quantities of dollars
is not simple myopia or mercantilism, but rather part of a deliberate strategy
of export-led growth, which may not be foolish given the structural limi-
tations of their financial and corporate governance systems.
I will begin by restating the authors’ case. It is not always easy to
understand everything they say, perhaps because of the novelty of much of
it. I will try to be helpful in defending them against some of the critiques
that have demanded their response. Indeed, I may carry the parallel with

1. See, for example, Eichengreen (2004). Dooley, Folkerts-Landau, and Garber (2003,
abstract) wrote, “there is a line of countries waiting [to follow the development strategy of
keeping their currencies undervalued against the dollar] sufficient to keep the system intact
for the foreseeable future.”
 196 Brookings Papers on Economic Activity, 1:2005

the Bretton Woods period further than they themselves intended or would
like. I will then, in some respects, try to suggest some different (and perhaps
less unconventional) language to make what seems to me the same argu-
ment. On the key question posed by the title of their paper, however, which
amounts to asking whether we are at the beginning of the revived Bretton
Woods system or the end, I come down on the latter side, or, more precisely,
that we are closer to the end than to the beginning. One could argue that
we are actually at the end, 1971, since the dollar depreciated substantially
against most major currencies during 2002–04, as so many of us predicted
it would when the United States shifted macroeconomic gears four years
ago. But the system may still have a little ways to go in other respects,
such as the much-anticipated end of the peg of the renminbi to the dollar. I
shall argue that in one respect we may be at 1967.
IS THE BRETTON WOODS ANALOGY APPROPRIATE? The five features of
the current system that the authors enumerate at the beginning of their paper
in fact bear little resemblance to the Bretton Woods system. Certainly, with
no mention of pegged exchange rates or gold, their list bears little relation
to the system that was agreed upon at Bretton Woods, New Hampshire, in
1944. But I think readers are meant to take “Bretton Woods” to refer to
the de facto system under which all other currencies were pegged to the
dollar, and all other central banks held dollars as the reserve asset because
the dollar was convertible into gold.2 Specifically, the “Bretton Woods”
period in this sense refers to the years from 1958, when the European
countries restored convertibility as envisaged in the International Mone-
tary Fund Articles of Agreement, to 1971, when the dollar was devalued,
or perhaps just to 1968, when the United States ceased to allow foreign
citizens to convert dollars to gold.
The ten to thirteen years that this de facto system lasted is not very long
in the broad scheme of things. One might even question whether it was a
“system,” given that it was already breaking down throughout much of that
period, under strain from the U.S. balance of payments deficit. But almost
everyone has long understood these points. That the dollar is still the main
international reserve currency today, that some Asian central banks see it
in their interest to pile up large quantities of dollars, and that their doing
so finances a large U.S. balance of payments deficit like those of the

2. See, for example, McKinnon (1996, especially p. 41).

Michael Dooley and Peter Garber 197

1960s, together probably suffice to justify the analogy with Bretton Woods.
The picture is indeed that of a “system,” in the sense that it shows how the
two halves, the United States and the periphery, fit together into a whole—
a mode of analysis that has largely died out from international macro-
economics since 1971, as Barry Eichengreen points out in his comment.
True, only China and a couple of other Asian economies (Hong Kong and
Malaysia) have currencies that are fixed to the dollar today, even de facto.
But dollar purchases by foreign central banks, especially in Asia, seek-
ing to prevent their currencies from appreciating against the dollar have
nonetheless been very large recently. They have financed an increasing
share of the U.S. current account deficit, which itself has been widening
sharply, over the last four years.
In the third of the three “notes” that constitute the Dooley-Garber paper,
the authors take pains to defend themselves against the criticism that the
Bretton Woods analogy is invalid because the United States was not run-
ning a current account deficit in the 1960s as it is today. I agree with them
on this. In the first place, the long-term trend in the U.S. current account
balance during the Bretton Woods period was negative, even if the bal-
ance was usually greater than zero. The United States had run substantial
surpluses in the late 1940s, because it had emerged from World War II
with its productive capacity intact. (This was called the period of “dol-
lar shortage.”) The surpluses diminished in the 1950s. True, they then
recovered a bit, peaking in 1964. But from then on the general trend was
downward until the system fell apart in 1971.
In the second place, more-comprehensive measures of the balance of pay-
ments, starting with the basic balance (which includes foreign direct invest-
ment and other long-term capital flows) and the liquidity balance (which adds
short-term, nonliquid flows and errors and omissions), did turn negative. The
overall U.S. balance of payments went into deficit in 1958, which is presum-
ably why the authors single out that year for their Bretton Woods analogy.
These deficits defined the 1960s as a period of excess supply of dollars.3
Both in Europe in the 1960s and in China today, rising foreign direct
investment (FDI)—bilateral FDI as well as overall outward FDI from the
3. The United States was losing reserves throughout 1958–67, and in large amounts
during 1970–71, which forced the devaluation and the closing of the gold window. Foreign
central banks were also piling up dollars, and in ever-larger magnitudes, through most of
the 1960s, and especially starting in 1970. The trend in 1968–69 was actually in the other
direction, perhaps because the United States had begun to make an effort to tighten fiscal
policy with a tax surcharge.
 198 Brookings Papers on Economic Activity, 1:2005

United States and overall inward FDI to the partners in question—was an

important part of the story, indeed just as important as the trade balance.
One difference is that in Europe the voices decrying “the American chal-
lenge” were louder than those favoring the entry of foreign multinational
companies as part of an intelligent growth strategy, whereas China seems
more welcoming to FDI. When the French complained about the United
States’ “exorbitant privilege”—the ability to trade pieces of paper for
items of more tangible value—they had in mind Americans acquiring
French factories as much as, or more than, French goods. So I think the
authors are on firm ground in focusing on a broader measure of the balance
of payments, one that includes direct investment, such as the basic balance or
the liquidity balance.
TERMINOLOGY. It may be worth spending a moment on language choices,
so that we are all sure we are talking about the same things. I have some
translating to propose.
The authors use the phrase “trade account country” to refer to a coun-
try that offsets current account imbalances with large changes in official
reserves. They particularly have in mind China, which has been running
a trade account surplus and offsetting it with increases in reserves. They
use the phrase “capital account country” to refer to one that offsets its
current account deficits with capital inflows. (Then there is the third set of
countries—Europe and a few others—that are floaters.) In standard models
the first group would be characterized by low capital mobility and a high
propensity for the authorities to intervene in foreign exchange markets to
stabilize the exchange rate, and the second group by high capital mobility
and less of a tendency to intervene. I am afraid that I don’t care for the terms
“trade account country” and “capital account country.” Each time I read
them, I have to stop and remind myself which one is which in the authors’
scheme. I think it would be much more intuitive to call Asia the “exporting
countries” and the United States the “consuming country.”
AMERICA AS THE WORLD’S BANKER. I learned from Charles Kindle-
berger thirty years ago that one way to think of the chronic willingness of
other countries to absorb dollars is that the United States acts as the
world’s banker, taking short-term deposits and investing in longer-term,
higher-risk, higher-return assets such as FDI.4 Thus I accept the idea that

4. Kindleberger (1965).
Michael Dooley and Peter Garber 199

both parties stand to gain from an exchange of gross capital flows, whereby
U.S. companies undertake direct investment in China and China acquires
dollar bonds. Notwithstanding that U.S. Treasury bills tend to pay a much
lower rate of return on average than the United States earns on its invest-
ments in other countries, I accept the argument that this can be part of a
useful development strategy. (I leave it to others to reconcile this with the
fact that the recent surges of capital into China have been portfolio invest-
ment, not FDI, and that even the FDI has for the most part not come from
the United States.5) I am still, however, mulling over the notion that FDI is,
as the authors put it, the “lesser credit.” This must mean that the danger of
expropriation by China’s government is greater than the danger that the
United States will depreciate away the value of the bonds, and that there-
fore China has to offer collateral to the United States, not vice versa, and
that supplying exports up front constitutes this collateral. It is an interest-
ing notion.
But while I am thinking about this, I have a question. Where do private
short-term liquid portfolio flows belong in the story? Are we sure it is the
liquidity balance that matters, as the authors say, and not the official settle-
ments balance (or official reserve transactions balance), which is after all
the most comprehensive measure of the balance of payments? Why not
also include short-term portfolio capital flows, even those that are liquid?
In other words, why not draw the line so that all private transactions
appear above the line, and nothing besides official reserve transactions
appears below it? Isn’t the main point that the United States can run
deficits on the overall balance of payments, and that other countries are
forced to run corresponding surpluses in order to earn foreign exchange
reserves? Wasn’t that the point of the two-country version of the monetary
approach to the balance of payments in the 1960s?6 Don’t the arguments
about how developing countries are forced to post “collateral” against cap-
ital inflows apply just as much—actually, more—to their inflows of short-
term portfolio capital as to FDI?
If it is just the U.S. monetary authorities who are playing the role of
world banker, the answer is that what matters is the official settlements
balance. If it is the entire U.S. commercial banking sector, what matters is
the liquidity balance. The language of private versus public domination of

5. See, for example, Prasad and Wei (2005).

6. See, for example, Mundell (1971) and Dornbusch (1973).
 200 Brookings Papers on Economic Activity, 1:2005

the capital account makes it sound as if the official settlements balance were
key, that one should draw the line below all private transactions, even liq-
uid short-term banking flows. Consistent with this, the biggest increase in
China’s balance of payments over the last few years has been neither in
the current account nor in FDI, but rather in unmeasured inflows that are
generally considered to be speculative and short-term: it is Chinese citi-
zens bringing back onshore liquid investments that they had previously
managed to accumulate offshore. Nor is the circumstance unique to China
of short-term capital flowing from the United States to the emerging mar-
ket, that is, in the same direction as the FDI. This has been a feature of
other emerging markets, both during the most recent boom phase of the
international debt cycle (2002–05) and during the preceding one (1990–96).
Hence all the talk about how capital inflows are more likely to lead to
crises if their composition is tilted in the direction of short-term loans,
unless they are fully offset by reserves.
One reason to skip directly to the official settlements balance is that the
liquidity balance and the others are no longer computed. The United States
stopped trying to calculate these statistics long ago, in part because the cap-
ital account statistics are not reliable, in part because it is difficult even in
principle to define what is a short-term and what is a long-term flow, and
in part because the entire exercise of distinguishing between autonomous
and accommodating transactions became obsolete with the end of the Bretton
Woods system. (The “basic balance,” however, is still reported for some
countries, such as Japan.)
LONG-TERM INTEREST RATES. The authors cite low real interest rates
as evidence that the world is experiencing a glut of saving—that the rela-
tionship is being pushed by the Asian desire to save, rather than pulled by
the U.S. desire to consume. It is true, as they say, that “the fact of unusually
low long-term real interest rates for this stage of the business cycle [is] a
direct challenge to those who . . . claim that the end is near.” My view is
that one of the ways economists can most usefully contribute to real-time
analysis of the economy is by pointing out when some market price seems
to be out of line with historical relationships, the implication being that
it is likely to correct itself within a couple of years. (Examples include
the undervalued U.S. stock market in 1980, the overvalued dollar in 1985,
the overvalued yen in 1995, the overvalued stock market in 2000, and the
undervalued euro in 2002.) Perhaps those observers are right who say that
the recent coexistence of low U.S. interest rates with large U.S. budget
Michael Dooley and Peter Garber 201

deficits shows that there has been a major fundamental shift in the willing-
ness of global investors to hold U.S. debt. But I am willing to bet against
it. Here is one testable disagreement.
My view is that the bond market was buoyed over 2001–04 by three
factors, each of which is already starting to come to an end:
—easy monetary policy in the United States, that is, the purchases of
U.S. Treasury securities by the Federal Reserve (which began to reverse
in the summer of 2004)
—easy monetary policy in Asia, that is, the purchases of U.S. securities
by Asian central banks (who get center stage in the Dooley-Garber story),
and
—the fact that investors appear still to be putting some weight on official
government projections of declining future U.S. budget deficits, despite
all the reasons to disbelieve them.
As these three factors come to an end, nominal long-term interest rates
should rise from their current levels, near 4 percent at the time of this
writing, to above 6 percent. I calculate this as approximately 21⁄2 percent
inflation plus 2 percent for a normal real short-term rate plus 1 percent for
a normal term premium plus 1 percent as an extra term premium for an
expected path of a rising debt-GDP ratio. This does not even count the
proposed dumping of huge quantities of new U.S. Treasury bonds on the
market to fund a transition to privatized Social Security accounts. Nor
does it count possible unforeseen factors such as further instability com-
ing from the Middle East or new oil price increases. It seems to me that a
crash is more likely to come in the bond market than anywhere else. But
time will tell.
I would also guess that the dollar will resume its depreciation long before
all unemployed or badly employed labor in China is reallocated to world-
class production (which the authors say will take at least ten years). It is
now in China’s own interest to move away from the peg. It has more
reserves than it can use.7 But the Chinese authorities do not want to be
pushed into revaluing the renminbi against the dollar.8 It is safer to bet

7. Frankel (2005b).
8. Li Ruogu, deputy governor of the People’s Bank of China, may have best captured
the Chinese perspective when he said in May 2004, “I think those who call for a fixed
exchange rate are right in the short run. And those who call for a floating exchange rate are
right in the long run. How long is the short run, you ask? You must understand. China is
8,000 years old. So when I say, short run, it could be 100 years” (author’s paraphrase).
 202 Brookings Papers on Economic Activity, 1:2005

against the U.S. bond market: I think it will probably fall before the dollar
does, or at worst at the same time.
Perhaps my answer to the question of whether it is now 1958 or 1968 is
that it is 1967. In that year the seriousness of the U.S. balance of payments
imbalance had become clear, yet large increases in government spending
were still coming out of Washington—much of that spending, then as now,
on a foreign military adventure—with no evidence of any willingness to
pay for it by raising taxes.9 Perhaps if it had not been for the Vietnam War
and the Great Society programs—the determination to have both the guns
and the butter—and the associated fiscal and monetary profligacy, the
Triffin dilemma would have taken decades to work itself out fully.10 But
the excessively expansionary U.S. macroeconomic policy accelerated the
process, so that the end came in 1971. To my way of thinking, a funda-
mental systemic structure that produces rising U.S. balance of payments
deficits, combined with excessively expansionary macroeconomic poli-
cies that accelerate the trend, sounds like the situation today. It is in this
respect that I think the current period resembles the 1960s even more, per-
haps, than Dooley and Garber think. But it is also for this reason that I
think the situation is unlikely to be sustainable for ten or twenty years.
Incidentally, easy monetary policy kept real interest rates fairly low in
1967 as well: the federal funds rate was 4.2 percent and the ten-year yield on
government bonds was 5.1 percent, compared with consumer price infla-
tion at 3.0 percent. Thus the Johnson deficits seem to be a better precedent
for the Bush deficits than are the Reagan deficits, which were not initially
accommodated by monetary policy.
THE EURO AS A RIVAL FOR THE DOLLAR. I both agree and disagree
with the authors’ rejection of recent media reports that central banks are
jumping on a bandwagon of diversification of their reserves out of dollars.
Where I agree is that there is an element of hysteria to such reports, which
come in highly predictable waves every time the dollar has been depreci-
ating for a couple of years in a row. The most recent previous cycle occurred
in 1994–95, when articles suggested that the dollar might lose its status as
the unrivaled international reserve currency, that it was in danger of “going

9. See, for example, Solomon (1977, pp. 102–04). In 1968 Congress did pass an income
tax surcharge to help pay for the spending. It was too little, or too late, or both, to head off
rising fiscal deficits and inflation.
10. Triffin (1960).
Michael Dooley and Peter Garber 203

the way of sterling, the guilder, the ducat and the bezant.”11 The recent cycle
of alarmist articles in the financial media was not hard to predict.
The earlier cycle of alarms was wrong. In the first place, shares of
reserve holdings in central banks and other measures of international cur-
rency use change only slowly over time. In the second place, these mea-
sures in the 1990s actually showed that the dollar’s share had reversed the
downward trend of the 1970s and 1980s.12 But this time may be different,
and this is where I am less confident than the authors. The share of the
dollar in international reserves has again resumed its downward trend.
And there are two or three available explanations. One is that there is now
an obvious rival, the euro, which is more credible as an alternative than
the deutsche mark or the yen ever were. Another is that the United States
now has a net international debt that is large and rapidly rising.
Some argue that, because central banks in Asia hold so many dollars
already, they will be reluctant to diversify into other assets for fear of pre-
cipitating a sharp depreciation of the dollar, in which the value of their
holdings will be the hardest hit. No doubt individual central banks have
this concern. But, as often, I agree with Eichengreen: when each individ-
ual participant decides that it stands to lose more by holding pat than by
joining the run, it will act in its own self-interest.13 If narrow economic
self-interest is not sufficient to stop the slide, will enlightened geopolitical
calculation do it? In the meantime, the United States has lost popular sym-
pathy and political support in much of the rest of the world. Our past deficits
due to imperial overstretch were manageable when others paid the bills
for our troops overseas: Germany and Japan during the Cold War, Kuwait
and Saudi Arabia in 1991. Now the hegemon has lost its claim to legiti-
macy in the eyes of many. The next time the United States asks other cen-
tral banks to bail out the dollar, will they be as willing to do so as Europe
was in the 1960s, or as Japan was in the late 1980s after the Louvre
Accord? It seems unlikely.
Menzie Chinn and I have documented econometrically some of the com-
monly hypothesized determinants of reserve currency shares: size of the
home economy, rates of return, stability of the currency, historical inertia,

11. Kindleberger (1995, p. 6).

12. Frankel (1995).
13. Eichengreen (2004).
 204 Brookings Papers on Economic Activity, 1:2005

and a tipping phenomenon.14 We have found each of these to be statisti-

cally significant during the period 1973–98. We found surprisingly little
support for another hypothesized variable, the net international invest-
ment position of the home country. We use the admittedly brief period of
European Economic and Monetary Union (EMU), 1999–2003, as a crude
test of the out-of-sample performance of the estimated equation. The equa-
tion correctly predicts a small increase in the euro share at the expense of
the dollar during this period. We then use the parameter estimates to pro-
ject into the future. The projections are naturally very sensitive to what one
assumes about the explanatory variables, particularly the size of Euroland.
Our preliminary finding is that, if some Central and Eastern European
countries join EMU by 2010, and if Sweden, Denmark, and (most impor-
tant) the United Kingdom join by 2015, so that Euroland becomes larger
economically than the United States, the tipping phenomenon could set in
soon thereafter. The euro could definitively pass the dollar, perhaps in the
subsequent decade. Even if some of the EU countries stay out of Euroland,
as is likely, the tipping could result in response to a future trend deprecia-
tion of the dollar that continues at the same pace as in the past.
It is still true that measures of international currency use such as reserve
shares change slowly. So when one sees newspaper headlines warning of
the euro overtaking the dollar, it is important to realize that this is unlikely
to happen for a long time. But if it does happen, the consequences are
likely to be large. It would mean the end of America’s “exorbitant privi-
lege.” It would be not just a ten-year “system” coming to an end, but the
end of a century of U.S. economic hegemony.

General discussion: Paul Krugman saw the crucial question as not

whether China will stop accumulating foreign exchange reserves, but
whether it will diversify its reserve portfolio, switching to the euro or pos-
sibly the yen. He also noted that loss of reserve currency leadership could
arise not only from diversification of existing reserves, but also from diver-
sification at the margin. He speculated that this would have huge exchange
rate implications. Kenneth Rogoff was skeptical of the authors’ claim that
demand for U.S. assets by Asia’s official sector can explain today’s low

14. See, for example, Bergsten (1975), Dooley, Lizondo, and Mathieson (1989), and
Eichengreen and Mathieson (2000). Chinn and Frankel (2005) provide a comprehensive
review.
Michael Dooley and Peter Garber 205

dollar interest rates, and he drew a parallel with recent work by Glenn
Hubbard and Eric Engen, who concluded that the recent huge U.S. budget
deficits do not matter for interest rates. Rogoff also doubted that the huge
pool of surplus Chinese labor entering the global market is driving interest
rates down, noting that standard models would predict the opposite and
that both the capital share of income in the advanced industrial economies
and corporate profits are surging. Michael Dooley replied that the pressure
on interest rates came not directly from labor supply but from China’s very
high saving rate: those savings, effectively, are exported to the rest of the
world.
Rogoff emphasized the role of Japan, whose current account surplus in
recent years is several times China’s and whose reserves are larger. He
noted that arguments based on labor surplus and the collateral hypothesis
do not apply to Japan. He added that other Asian economies also run larger
surpluses than China and together accumulate more reserves than China,
but they are much more open and their capital markets are more integrated
than Europe’s were in the 1960s. Gian Maria Milesi-Ferretti was also skep-
tical about the authors’ hypothesis that emerging economies accumulate
U.S. reserves as a form of collateral. He reminded the panel that many of
the countries that have grown rapidly, including Korea, Malaysia, Thailand,
and Indonesia, have typically been borrowers on international markets. The
main exception has been Taiwan, which is a special case for various rea-
sons. Singapore, which is now a large creditor, borrowed heavily in its
early stages of development. Milesi-Ferretti also noted that a country’s
external balance can depend importantly on the terms of trade. Asia’s emerg-
ing economies are mostly commodity importers and are today running large
current account surpluses despite unfavorable terms of trade. If there is a
temporary component in the current high prices of oil and other commodi-
ties that these countries import, the structural surpluses may be even larger
than they appear.
Richard Cooper questioned the presumption of many observers that the
Chinese renminbi is undervalued. He noted that China has substantially effec-
tive controls on resident capital outflows, an investment rate of 40 percent
of GDP but an even higher saving rate, and a modest current account sur-
plus relative to the size of the economy. Given these facts and the fact that
exchange rates in the rest of the world are generally floating, Cooper saw
little evidence that the renminbi is undervalued. Furthermore, he expected
a reform of the Chinese banking system and, eventually, full currency con-
 206 Brookings Papers on Economic Activity, 1:2005

vertibility. In that event the renminbi might well depreciate, because cap-
ital outflows would increase as wealthy Chinese households begin to invest
abroad to diversify their portfolios. He added that one already observes
increasingly aggressive foreign direct investment by China, with Chinese
oil firms just one example.
Michael Dooley and Peter Garber 207

References
Aizenman, Joshua, Brian Pinto, and Artur Radziwill. 2004. “Sources for Financ-
ing Domestic Capital—Is Foreign Saving a Viable Option for Developing
Countries?” Working Paper 10624. Cambridge, Mass.: National Bureau of
Economic Research (July).
Bergsten, C. Fred. 1975. The Dilemmas of the Dollar. New York University Press.
Caballero, Ricardo J. 2004. “The Wrong Call: The Euro Is No Match for the Dol-
lar.” Massachusetts Institute of Technology (December).
Caballero, Ricardo J., and Arvind Krishnamurthy. 2001. “International and Domes-
tic Collateral Constraints in a Model of Emerging Market Crises.” Journal of
Monetary Economics 48: 513–48.
_________. 2003. “Excessive Dollar Debt: Financial Development and Under-
insurance.” Journal of Finance 58, no. 2: 867–93.
Chinn, Menzie, and Jeffrey Frankel. 2005. “Will the Euro Eventually Surpass the
Dollar as Leading International Reserve Currency?” Working Paper 11508.
Cambridge, Mass.: National Bureau of Economic Research.
Despres, Emile, Charles P. Kindleberger, and Walter S. Salant. 1966. “The Dollar
and World Liquidity: A Minority View.” The Economist (February 5).
Dickson, Paul. 1978. The Official Rules. New York: Delacorte Press.
Dooley, Michael P., David Folkerts-Landau, and Peter Garber. 2003. “An Essay
on the Revived Bretton Woods System.” Working Paper 9971. Cambridge,
Mass.: National Bureau of Economic Research (September).
_______. 2004a. “The Revived Bretton Woods System: The Effects of Periphery
Intervention and Reserve Management on Interest Rates and Exchange Rates
in Center Countries.” Working Paper 10332. Cambridge, Mass.: National
Bureau of Economic Research (March).
_______. 2004b. “Direct Investment, Rising Real Wages and the Absorption of
Excess Labor in the Periphery.” Working Paper 10626. Cambridge, Mass.:
National Bureau of Economic Research (July).
_______. 2004c. “The US Current Account Deficit and Economic Development:
Collateral for a Total Return Swap.” Working Paper 10727. Cambridge,
Mass.: National Bureau of Economic Research (September).
Dooley, Michael P., J. Saul Lizondo, and Donald J. Mathieson. 1989. “The Cur-
rency Composition of Foreign Exchange Reserves.” International Monetary
Fund Staff Papers 36, no. 2: 385–434.
Dornbusch, Rudiger. 1973. “Devaluation, Money and Nontraded Goods.” Ameri-
can Economic Review 63, no. 5: 871–80.
Eichengreen, Barry. 2004. “Global Imbalances and the Lessons of Bretton
Woods.” Working Paper 10497. Cambridge, Mass.: National Bureau of Eco-
nomic Research (May).
 208 Brookings Papers on Economic Activity, 1:2005

_______. Forthcoming. “Chinese Currency Controversies.” Asian Economic

Papers.
Eichengreen, Barry J., and Donald J. Mathieson. 2000. “The Currency Composi-
tion of Foreign Exchange Reserves: Retrospect and Prospect.” Working Paper
wp/00/131. Washington: International Monetary Fund.
Eichengreen, Barry, and Charles Wyplosz. 1993. “The Unstable EMS.” BPEA,
no. 1: 51–124.
Eichengreen, Barry, Paul Masson, and others. 1998. “Exit Strategies: Policy Options
for Countries Seeking Greater Exchange Rate Flexibility.” Occasional Paper 168.
Washington: International Monetary Fund (April).
Frankel, Jeffrey. 1995. “Still the Lingua Franca: The Exaggerated Death of the
Dollar.” Foreign Affairs 74, no. 4: 9–16.
_______. 2005a. “The Twin Deficits.” Notes presented at a meeting of the Bellagio
Group, Amsterdam, January 20–21.
_______. 2005b. “On the Renminbi: The Choice between Adjustment under a
Fixed Exchange Rate and Adjustment under a Flexible Rate.” Working Paper
11274. Cambridge, Mass.: National Bureau of Economic Research (April).
Frankel, Jeffrey A., and Andrew K. Rose. 1996. “Currency Crashes in Emerging
Markets: An Empirical Treatment.” Journal of International Economics 41,
no. 3–4: 351–66.
Goldstein, Morris. 2004. “Adjusting China’s Exchange Rate Policies.” Washington:
Institute for International Economics (May).
Goldstein, Morris, and Nicholas R. Lardy. 2004. “What Kind of Landing for the
Chinese Economy?” Policy Briefs in International Economics PB04-7. Wash-
ington: Institute for International Economics (November).
_______. 2005. “China’s Role in the Revived Bretton Woods System: A Case of
Mistaken Identity.” Working Paper 05-2. Washington: Institute for International
Economics.
Grilli, Vittorio U. 1986. “Buying and Selling Attacks on Fixed Exchange Rate
Systems.” Journal of International Economics 20: 143–56.
Hsieh, Chang-Tai, and Christina D. Romer. 2001. “Was the Federal Reserve Fet-
tered? Devaluation Expectations in the 1932 Monetary Expansion.” Working
Paper 8113. Cambridge, Mass.: National Bureau of Economic Research (Feb-
ruary).
Kindleberger, Charles P. 1965. “Balance of Payments Deficits and the Interna-
tional Market for Liquidity.” Essays in International Economics 46. Interna-
tional Economics Section, Department of Economics, Princeton University.
________. 1995. “Dollar Darkness.” The International Economy (May/June).
McKinnon, Ronald. 1996. The Rules of the Game: International Money and
Exchange Rates. MIT Press.
Mundell, Robert A. 1971. Monetary Theory. Pacific Palisades, Calif.: Goodyear.
Michael Dooley and Peter Garber 209

Nurkse, Ragnar. 1945. “Conditions of International Monetary Equilibrium.” Essays

in International Economics 4. International Economics Section, Department of
Economics, Princeton University.
Obstfeld, Maurice. 1993. “The Adjustment Mechanism.” In A Retrospective on the
Bretton Woods System, edited by Michael D. Bordo and Barry Eichengreen.
University of Chicago Press.
Prasad, Eswar, and Shang-Jin Wei. 2005. “The Chinese Approach to Capital
Inflows: Patterns and Possible Explanations.” Working Paper 11306. Cam-
bridge, Mass.: National Bureau of Economic Research (May).
Razin, Assaf, and Gian Maria Milesi-Ferretti. 1998. “Current Account Reversals
and Currency Crises: Empirical Regularities.” Working Paper 6620. Cam-
bridge, Mass.: National Bureau of Economic Research (June).
Roubini, Nouriel, and Brad Setser. 2004. “The US as a Net Debtor: The Sustain-
ability of the US External Imbalances.” Stern School of Business, New York
University, and University College, Oxford (November).
Salant, Stephen W. 1976. “Exhaustible Resources and Industrial Structure: A
Nash-Cournot Approach to the World Oil Market.” Journal of Political Econ-
omy 84, no. 5: 1079–93 .
Solomon, Robert. 1977. The International Monetary System: An Insider’s View,
1945-1976. Harper and Row.
Summers, Lawrence H. 2004a. “The United States and the Global Adjustment
Process.” Speech at the Institute for International Economics, Washington,
March.
_________. 2004b. “The U.S. Current Account Deficit and the Global Economy.”
Per Jacobsson Lecture, Washington, October 3.
Triffin, Robert. 1960. Gold and the Dollar Crisis. Yale University Press.
SEBASTIAN EDWARDS
University of California, Los Angeles

Is the U.S. Current Account

Deficit Sustainable? If Not, How
Costly Is Adjustment Likely to Be?

MANY ANALYSTS IN academia, the private sector, and applied research

institutions express increasing concern about the growing U.S. current
account deficit. There is a general sense that current global imbalances
are unsustainable and that adjustment must come sooner rather than later.
The unprecedented magnitude of the U.S. current account deficit and
the United States’ growing net foreign indebtedness have fueled these
worries, with many analysts arguing that, unless something is done, the
world will move toward a major financial crisis.1 Some have gone as far
as to suggest an imminent collapse of the dollar and a global financial
meltdown.2 Underlying this view is the fact that, if the deficit continues
at its current level, U.S. net international liabilities will eventually reach
100 percent of GDP, a figure widely considered to be excessively large.3
The source of financing of the U.S. current account deficit has also
become a matter of concern. A number of authors have argued that, by
relying on foreign and particularly Asian central banks’ purchases of

1. Although most of these alarmist discussions have taken the form of newspaper opin-
ion pieces, there have also been a few policy papers on the subject. See, for example,
Roubini and Setser (2004).
2. See, for example, Roubini and Setser (2004). For an excellent set of papers on the
subject, see Bergsten and Williamson (2004).
3. For example, in a very clear discussion of this issue, Michael Mussa said, “[T]here
probably is a practical upper limit for US net external liabilities at something less than 100
percent of US GDP and, accordingly . . . current account deficits of 5 percent or more of US
GDP are not indefinitely sustainable” (Mussa, 2004, p. 114).

211
 212 Brookings Papers on Economic Activity, 1:2005

Treasury securities, the United States has become extremely vulnerable to

sudden changes in expectations and economic sentiments.4
Robert Skidelsky recently argued in the New York Times that the value
of the dollar is one of the most important sources of political tension
between the United States and Europe. Arguing that “[U]nilateralism is
not more acceptable in currency matters than in foreign policy,” Skidel-
sky points out that,
The United States is the only major country proclaiming itself indifferent to its
currency’s value. In countries running persistent current account deficits, gov-
ernments normally—indeed must—reduce domestic consumption. But so far,
the United States has relied on other countries to adjust their economies to prof-
ligate American spending. . . .5

There is, however, an alternative view. Some authors have argued that,
in an era of increasing financial globalization and rapid U.S. productivity
gains, it is possible—indeed, even logical and desirable—for the United
States to run very large current account deficits for a very long period
(say, a quarter of a century). In this view, growing international portfolio
diversification implies that the rest of the world will be willing to accu-
mulate large U.S. liabilities during the next few years, maybe even in
excess of 100 percent of U.S. GDP. From this perspective, since the U.S.
current account deficit poses no threat, there are no fundamental reasons
to justify a significant fall in the value of the dollar. 6
This paper analyzes the relationship between the dollar and the U.S.
current account, with particular attention to the issue of sustainability
and the mechanics of current account adjustment. I develop a portfolio
model of the current account and show that, even under a very positive
scenario where foreigners’ (net) demand for U.S. assets doubles from its
current level, the U.S. current account will have to go through a signifi-
cant adjustment in the not-too-distant future. Indeed, one cannot rule out
a scenario where the U.S. current account deficit shrinks abruptly by
3 to 6 percent of GDP. To get an idea of the possible consequences of
such an adjustment, I also analyze the international historical evidence

4. See, for example, Martin Wolf’s article in the Financial Times, “Funding America’s
Recovery is a Very Dangerous Game,” October 1, 2003, p. 15.
5. “Winning Back Europe’s Heart: Rogue Dollar,” New York Times, February 20,
2005, p. 9.
6. Dooley, Folkerts-Landau, and Garber (2004a, 2004b); see also Cooper (2004) and
Caballero, Farhi, and Hammour (2004).
Sebastian Edwards 213

on current account reversals. The results of this empirical investigation

indicate that significant current account reversals have tended to result
in large declines in GDP growth.

The U.S. Dollar and the Current Account:

A Thirty-Year Perspective

This section analyzes the behavior of the U.S. real exchange rate
(RER) and current account since the adoption of floating exchange rates
in the early 1970s.7 I begin by discussing the course of the U.S. RER and
current account during that period and the changing nature of the U.S.
trade-weighted RER index. I argue that the last thirty years of U.S. RER
behavior can be divided into six distinct phases. Second, I discuss the
most recent data on the U.S. current account, including its sources of
financing. And third, I provide some international evidence on current
account imbalances during the last three decades. This comparative
analysis helps in placing the recent U.S. experience in historical context.

Six Phases of Real Exchange Rate Behavior

Figure 1 presents quarterly data for the U.S. current account balance as
a percentage of GDP and for the Federal Reserve’s trade-weighted index
of the U.S. RER, both for the period 1973–2004; in this figure, as in the
rest of the paper, the RER index is defined such that an increase represents
a real appreciation. The figure shows that current account deficits have
become increasingly large since 1992. It also shows that, over the first
decade of floating exchange rates (1973–82), the United States ran small
current account surpluses or deficits, which averaged to a small surplus of
0.04 percent of GDP. In contrast, over 1983–2004 the current account bal-
ance was in deficit, on average, by 2.4 percent of GDP. Figure 1 also

7. Because of space considerations, I do not discuss in detail some important issues

such as the stationarity of the RER and its (changing) volatility through time. Most recent
analyses based on panel data have found that the RER is stationary and that its half-life
cycle is less than the three to five years traditionally considered to be the consensus view.
See Choi, Mark, and Sul (2004). An analysis of U.S. RER volatility indicates that for the
complete period 1975–2004 the U.S. RER index exhibited one of the highest volatilities in
the sample. Only the British pound, the Japanese yen, and the euro had greater volatility.
Within this period, the RER volatility of the U.S. dollar was at its highest in 1985–89,
which roughly corresponds to phase III (a period of rapid depreciation) in figure 2.
 214 Brookings Papers on Economic Activity, 1:2005
Figure 1. Real Exchange Rate and Current Account Balance, 1973–2004

Percent of GDP Index (March 1973 = 100)a

Real exchange rate

6 120
(right scale)
4
100
2
80
0

–2
Current account
–4 (left scale)
–6
I II III IV V VI

1975 1979 1983 1987 1991 1995 1999 2003

Source: Bureau of Economic Analysis, National Income and Product Accounts, International Transactions Accounts; Federal
Reserve data.
a. Price-adjusted Major Currencies index.

shows that during the whole period under consideration the RER index
fluctuated within a wide range: from 91.2 at its lowest point, it rose to
136.3 at its highest, with a mean for the period of 105.3. Finally, the fig-
ure demonstrates an apparent negative correlation between the trade-
weighted real value of the dollar and the current account balance: periods
of a strong dollar have tended to coincide with periods of larger current
account deficits. Although the relationship is not exact, the synchronicity
between the two variables is quite high: the contemporaneous coefficient
of correlation between the logarithm of the RER index and the current
account balance is −0.53; the highest value for the correlation coefficient
(−0.60) is obtained when the log of the RER is lagged three quarters.
Recent policy debates about the value of the dollar illustrate the massive
changes that have occurred in U.S. trade relations during the last three
decades. Discussions of the dollar in the early 1970s dealt almost exclu-
sively with bilateral exchange rates—both nominal and real—between
the dollar and the currencies of other industrial countries; more recently
the debate has increasingly focused on the dollar’s value in terms of the
currencies of emerging economies, including the Chinese renminbi, the
Korean won, and the Malaysian ringgit. During the last few years the
Mexican peso has also become an important determinant of the trade-
Sebastian Edwards 215

weighted value of the dollar; this was not the case when the Smithsonian
Agreement was abandoned and the regime of floating exchange rates
began in 1973. Between 1995 and 2005 the renminbi’s weight in the Fed-
eral Reserve trade-weighted RER index of the dollar rose from 5.7 per-
cent to 11.4 percent, and the Mexican peso’s from 7.0 percent to 10.0
percent. Meanwhile the yen’s weight declined from 16.5 percent to 10.6
percent. In fact, the trade-weighted RER of the dollar is dominated today
by the Asian nations, which (excluding India) have a combined weight in
the index of 38.8 percent. The currencies of commodity-exporting coun-
tries, as a group, are also very important, with a weight of 24.6 percent.
Finally, the launching of the euro in 1999 has marginalized the British
pound: although its weight of 5.2 percent is still quite respectable, the
pound is no longer among the top five currencies in the index. The situa-
tion was quite different in 1998, when the pound had a higher weight
than the currencies of all but one of the countries that would eventually
adopt the euro: in that year the deutsche mark had a weight of 6.4 percent,
and the pound a weight of 5.9 percent.8
One can distinguish in figure 1 six distinct phases in U.S. RER behav-
ior for the thirty-two-year period 1973–2004. A summary of these six
phases is also in large measure a summary of the history of the inter-
national financial system since the inception of floating:9
—Phase I, 1973:1–1978:4. The early years of floating were character-
ized by a depreciating trend for the dollar in real terms, with the decline
in value cumulating to 18.1 percent over the twenty-four quarters. Dur-
ing this period the standard deviation of monthly log differences of the
RER index was 0.0205. During the early part of this phase (1973–76),
the current account was in surplus, but this turned into a small deficit in
1977 and 1978.
—Phase II, 1979:1–1985:1. During the next twenty-five quarters the
dollar experienced a 49.3 percent appreciation in real terms. Meanwhile the
current account went into deficit, which reached 2.9 percent of U.S. GDP in
1984:4. The standard deviation of monthly log differences of the RER
index was 0.022, slightly higher than in phase I. In view of the dollar’s
strengthening and the related increase in the U.S. current account deficit, on

8. In 2005 the euro has a weight of 18.8 percent in the Federal Reserve index; in 1995
the currencies that the euro later replaced had a combined weight of 17.3 percent.
9. Figure 1 presents the Federal Reserve’s broad RER index. The same six phases are
also apparent when alternative indexes are used.
 216 Brookings Papers on Economic Activity, 1:2005

September 22, 1985, the members of the Group of Five major industrial
countries (the G-5: the United States, Japan, the United Kingdom, France,
and Germany) agreed to implement concerted and coordinated interven-
tions in the foreign exchange market. As part of this agreement, which
came to be known as the Plaza Agreement, the G-5 countries also commit-
ted themselves to put in place coordinated macroeconomic policies that
would reduce the costs of the global adjustment process.
—Phase III, 1985:2–1988:4. During the period following the Plaza
Agreement, the dollar experienced a rapidly depreciating trend, with a peak-
to-trough change in the index of −28.7 percent. RER volatility increased
substantially during these fifteen quarters: the standard deviation of
monthly log differences of the RER index was 0.0268. The current
account deficit continued to grow, however, until in mid-1987 it stabilized
at around 3.6 percent of U.S. GDP. From that point onward the current
account began to improve, and by 1988:4 the deficit had declined to
2.4 percent of GDP. On February 22, 1987, the ministers of finance and
central bank governors of the G-6 (the former G-5 plus Canada) released
a communiqué, which came to be known as the Louvre Accord, declaring
that significant progress had been made in achieving global adjustment,
and that “Further substantial exchange rate shifts among their currencies
could damage growth and adjustment prospects in [the G-6] countries.”
The communiqué went on to say that the G-6 “agreed to cooperate closely
to foster stability of exchange rates around current levels.”10
—Phase IV, 1989:1–1995:2. During the next phase the dollar contin-
ued to depreciate in real terms, but at a much lower rate than in the pre-
ceding phase: during these twenty-six quarters the total real depreciation
was 10 percent. The standard deviation of monthly log differences of the
RER index over this period was 0.0232, and the current account balance
continued to improve, until in 1991:1 the United States posted its first cur-
rent account surplus in many years. The current account balance averaged
−1.15 percent of U.S. GDP during this phase.
—Phase V, 1995:3–2002:1. This phase was characterized by a trough-
to-peak real dollar appreciation of 33.4 percent (although, as figure 1
shows, between 1998:1 and 1999:4 there was a short-lived period of real
depreciation). Interestingly, during this phase RER volatility declined sig-
nificantly: the standard deviation of monthly log differences of the RER
10. Quoted from the text of the Louvre communiqué, available at www.g8.utoronto.
ca/finance/fm870222.htm.
Sebastian Edwards 217

index was 0.0196. This phase was also characterized by an increasing cur-
rent account deficit: whereas in late 1995 and early 1996 the deficit was
on the order of 1.5 percent of U.S. GDP, by early 2002 it was hovering
just below 4 percent of GDP. In 1999, for the first time in many years, the
U.S. federal government posted a budget surplus.
—Phase VI, 2002:2–2004:4. In the most recent (and continuing)
phase, the real value of the dollar has accumulated a 14 percent real
depreciation. The current account deficit has continued to widen, exceed-
ing 5 percent of U.S. GDP toward the end of the sample. RER volatility
has increased slightly: the standard deviation of log differences of the
RER index was 0.0212. Other important macroeconomic developments
during this phase included a worsening of the U.S. fiscal position and stiff
increases in the prices of oil and other commodities.
Figure 2 breaks down the U.S. current account balance for 1973 through
2004 into its main components: the balance of trade in goods and services,
the balance of trade in nonfinancial services, the income account, and
transfers, all as percentages of GDP on a yearly basis. As the top left panel
shows, large and persistent trade surpluses preceded the era of large cur-
rent account deficits: already in the late 1970s the trade account was nega-
tive, and since mid-1976 it has had only one surplus quarter (1992:2).11
The top right panel shows that since 1996 the trade surplus in nonfinan-
cial services has declined steadily, so that by 2004 it was only 0.3 percent
of GDP. The income account remains positive, as the bottom left panel of
the figure shows. The surplus has declined sharply since 1980, but given
that for many years now the U.S. international investment position has
been negative—that is, the United States has been a net debtor—the fact
that the income account is still positive may seem surprising. The reason is
that the return on U.S. assets held by foreigners has been systematically
lower than the return on foreign assets in the hands of U.S. nationals.
Finally, the bottom right panel shows that the transfers account has been
negative in every year except one since 1946. Recently the transfers deficit
has been stable at approximately 0.7 percent of GDP.
Recent Current Account Imbalances
Table 1 presents data on the U.S. current account deficit and its financ-
ing for the period 1990–2004. The nature of external financing of the
11. Mann (2004) shows that most of the U.S. trade deficit is explained by deficits in
automobiles and consumer goods.
Figure 2. Components of the Current Account Balance, 1973–2004

Goods and services Services

Percent of GDP Percent of GDP

1 1.2
0 1.0
–1
0.8
–2
0.6
–3
–4 0.4
–5 0.2
–6 0.0

Income Transfers

1.4
1.0
1.2
1.0 0.5
0.8
0.0
0.6
0.4 –0.5
0.2
–1.0
0.0

1979 1985 1991 1997 2003 1979 1985 1991 1997 2003

Source: Bureau of Economic Analysis, International Transactions Accounts.

Table 1. Net Financial Flows and Current Account Balance, 1990–2004a
Billions of dollars
Item 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
Change in reserves 31.8 23.2 44.4 70.4 44.9 100.1 133.4 18.0 −26.7 52.3 42.5 23.1 110.3 250.1 358.1
Foreign private -2.5 18.8 37.1 24.4 34.3 91.5 147.0 130.4 28.6 -44.5 -70.0 -14.4 100.4 113.4 108.1
purchases of
U.S. Treasuries
Currency 18.8 15.4 13.4 18.9 23.4 12.3 17.4 24.8 16.6 22.4 5.3 23.8 21.5 16.6 14.8
Securitiesb −27.2 −10.5 −19.1 −66.2 −6.2 −45.1 −46.0 44.6 32.1 182.6 338.0 309.2 301.4 178.6 323.2
Debt n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. 267.7 300.3 269.8 241.8 360.1
Equity n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. 93.0 12.6 37.5 −63.2 −36.8
Foreign direct 11.3 −14.7 −28.4 −32.6 −34.0 −41.0 −5.4 0.8 36.4 64.5 162.1 24.7 −62.4 −133.9 −133.0
investment
Claims reported by 17.3 8.0 13.2 11.3 −35.0 14.4 −32.6 −5.2 −15.1 −21.5 31.9 57.6 32.6 55.1 −41.5
nonbanks
Claims reported by 8.6 3.4 37.4 55.7 100.1 −44.9 −75.1 7.9 4.2 −22.0 −31.7 −7.5 66.1 65.2 −15.6
banks
Net financing 58.0 43.5 97.9 81.8 127.4 87.3 138.7 221.3 76.2 233.8 478.0 416.6 569.9 542.7 614.0
Current account −79.0 3.7 −48.0 −82.0 −118.0 −109.5 −120.2 −136.0 −209.6 −296.8 −413.4 −385.7 −473.9 −530.7 −665.9
balance
Source: Bureau of Economic Analysis, U.S. International Transactions and International Investment Position.
a. Each financing item is reported net (inflows minus outflows).
b. Debt and equity do not add up to total securities because of rounding.
 220 Brookings Papers on Economic Activity, 1:2005

deficit has changed significantly in the last few years. Whereas from 1997
to 2000 inflows of foreign direct investment (FDI) contributed in an
increasingly important way to financing the deficit, net FDI inflows fell
sharply in 2001 and have been negative since then. Also, after four years
(1999–2002) in which net equity flows were positive, these became nega-
tive in 2003–04. Table 1 also shows that during 2003 and 2004 the U.S.
current account deficit was fully financed through net fixed-income flows
(the sum of the first, second, and fifth rows in the table). Official foreign
purchases of government securities—“Change in reserves” in table 1—
played a particularly important role in financing the 2003 and 2004 cur-
rent account deficits. A number of analysts have argued that reliance on
foreign central bank purchases of Treasury securities has made the United
States particularly vulnerable to sudden changes in expectations and eco-
nomic sentiments.12
Current account imbalances are reflected in changes in a country’s
net international investment position (NIIP): deficits result in a deterio-
ration of the NIIP, and surpluses in an improvement. Figure 3 shows
that the U.S. NIIP as a percentage of GDP has become increasingly neg-
ative since the mid-1980s: in 2004 U.S. net international liabilities
reached 29 percent of GDP. An important feature of the U.S. NIIP that
distinguishes it from those of most other countries is that gross inter-
national assets and gross international liabilities are held in different
currencies. Whereas more than 70 percent of gross foreign assets held
by U.S. nationals are denominated in foreign currency, approximately
95 percent of gross U.S. liabilities in the hands of foreigners are denom-
inated in dollars. This means that U.S. net liabilities as a percentage of
GDP are subject to “valuation effects” stemming from changes in the
value of the dollar: a dollar depreciation reduces the value of U.S. net
liabilities; as a result, the deterioration of the U.S. NIIP during 2002–04
was significantly smaller than the accumulated current account deficit
during those two years (table 2).
A key question in current account sustainability analyses—discussed
in detail below—is, What is the “reasonable” long-run equilibrium ratio
of U.S. net international liabilities to GDP? The higher this ratio,
the greater will be the sustainable current account deficit. According to
some, the current ratio of almost 30 percent of GDP is already excessive;

12. See, for example, the article by Martin Wolf cited above.
Sebastian Edwards 221

Figure 3. Net International Investment Position, 1976–2004a

Percent of GDP

10
5
0
–5
–10
–15
–20
–25
–30

1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002

Source: Bureau of Economic Analysis, National Income and Product Accounts, International Investment Position.
a. Data for 2004 are author's projection. Direct investment positions are valued at current cost.

others believe that a NIIP-to-GDP ratio of up to 50 percent would be

reasonable.13
One of the first things that undergraduate students of open-economy
macroeconomics learn is that a country’s current account is equal to the
difference between its national saving and its domestic investment.
Over the years a number of authors have argued that a worsening of a
current account balance that stems from an increase in investment is
very different from one that results from a decline in national saving.
Some have gone as far as to argue that very large deficits in the current
account “don’t matter,” as long as they are the result of increased (pri-
vate sector) investment.14 As figure 4 shows, the recent deterioration of
the U.S. current account has largely corresponded to a decline in
national saving, and in particular of public and household saving, rather
than a rise in investment. A simple implication of this trend—and one
that is emphasized by most authors—is that an improvement in the U.S.
current account will require not only an RER adjustment, but also an
increase in the national saving rate to GDP. Symmetrically, a correction

13. See Obstfeld and Rogoff (2004) and Mussa (2004).

14. Corden (1994).
 222 Brookings Papers on Economic Activity, 1:2005
Table 2. Net International Investment Position and Current Account Balance,
1998–2004
Billions of dollars
Item 1998 1999 2000 2001 2002 2003 2004
Net international 900.0 775.5 1,388.7 1,889.7 2,233.0 2,430.7 n.a.
liabilities
position
Change from 79.3 −124.5 613.3 500.9 343.3 197.7 n.a.
previous year
Current account 209.6 296.8 413.4 385.7 473.9 530.7 665.9
deficit
Valuation changesa −130.2 −421.3 199.8 115.2 −130.6 −333.0 n.a.
Source: Bureau of Economic Analysis, U.S. International Transactions and International Investment Position.
a. Current account deficit plus valuation changes equals change in net international investment position from previous year.

of current global imbalances will also require a decline in Europe’s and

Japan’s saving rates, or an increase in their investment rates, or some
combination of the two.

The U.S. Current Account Deficit in International Perspective

How large are the recent U.S. current account deficits from a compara-
tive point of view? And how large is the U.S. net international liabilities
position compared with those of other developed countries through his-
tory? The top panel of table 3 presents data on the distribution of current
account balances (as a percentage of GDP) in the world economy, as well
as in six country groups—the industrial countries, Latin America, Asia,
the Middle East, Africa, and Eastern Europe—for 1971 through 2001 and
1984 through 2001. (The data are unweighted averages for each country.)
At almost 6 percent of GDP, today’s U.S. current account deficit is very
large indeed from a historical and international perspective: it is in the top
decile of all the deficits recorded by all industrial countries in the first
thirty years of floating exchange rates.
Since 1971 the United States has been the only large industrial country
that has run current account deficits in excess of 5 percent of GDP. This
reflects the unique position of the United States in the international finan-
cial system: U.S. assets have been in high demand, allowing the United
States to run large and persistent deficits. On the other hand, this fact also
suggests that the United States is moving into uncharted waters. As Mau-
Sebastian Edwards 223

Figure 4. Investment and Saving, 1970–2003a

Percent of GDP

Household saving
Corporate saving
15 Public saving
Foreign saving
Investment
10

–5

1972 1975 1978 1981 1984 1987 1990 1993 1996 1999 2002

Source: Bureau of Economic Analysis, National Income and Product Accounts, International Investment Position.
a. All series are on a net basis.

rice Obstfeld and Kenneth Rogoff,15 among others, have shown, if the
U.S. deficit continues at its current level, in twenty-five years U.S. net
international liabilities will exceed those recorded by any other country,
as a percentage of GDP, in modern times.
During the last thirty years the only industrial countries to have had
current account deficits in excess of 5 percent of GDP have been small
ones: Australia, Austria, Denmark, Finland, Greece, Iceland, Ireland,
Malta, New Zealand, Norway, and Portugal. What is even more striking is
that very few countries, industrial or developing, have had large current
account deficits that lasted for more than five years. Table 4 lists those
countries that have had “persistently large” current account deficits at
some time during the period 1970–2001. For purposes of this table, I
define a country as having a “large deficit” if, in any year, its current

15. Obstfeld and Rogoff (2004).

 224 Brookings Papers on Economic Activity, 1:2005
Table 3. Distribution of Current Account Deficits by World Region, 1970–2001
Percent of GDP
No. of 1st 1st 3rd 9th
Region countries Mean Median decile quartile quartile decile
1970–2001
Industrial countries 22 0.6 0.7 −3.8 −1.6 3.0 4.8
Latin America and 34 5.4 4.1 −2.5 1.1 8.0 16.9
Caribbean
Asia 22 3.0 2.7 −7.1 −0.6 6.3 11.3
Africa 51 6.3 5.3 −3.4 1.2 9.9 16.9
Middle East 12 0.0 1.4 −18.8 −5.0 6.4 13.6
Eastern Europe 25 3.9 3.0 −2.4 0.3 6.1 10.7
All countries 169 3.9 3.3 −5.0 −0.1 7.1 13.1
1984–2001
Industrial countries 25 0.2 0.3 −4.7 −2.3 2.7 4.8
Latin America and 34 5.1 3.7 −2.5 1.1 7.0 17.0
Caribbean
Asia 22 2.2 2.4 −8.0 −1.3 5.9 10.2
Africa 51 5.9 4.6 −3.5 0.9 9.1 16.2
Middle East 12 2.3 1.5 −12.4 −4.0 6.3 14.9
Eastern Europe 25 4.0 3.1 −2.5 0.3 6.6 10.9
All countries 169 3.8 3.0 −4.8 −0.4 6.7 12.9
Source: Author’s calculations using data from World Bank, World Development Indicators.

account deficit exceeded its region’s ninth decile.16 I then define a “persis-
tently large deficit” country as one that has at some time had a large
deficit, as defined above, for at least five consecutive years.17 The result-
ing list in table 4 is extremely short, and none of the countries listed is
large. This illustrates the fact that, historically, periods of large current
account imbalances have tended to be short lived and have been followed
by periods of current account adjustment.
Table 5 presents data on net international liabilities, as a percentage of
GDP, for a group of industrial countries that have historically had a large
negative NIIP position.18 The picture that emerges is different from that in

16. Notice that the thresholds for defining “large” deficits are year- and region-specific,
with a different threshold for each region each year.
17. For an econometric analysis of the persistence of current account deficits see
Edwards (2004). See also Taylor (2002).
18. Data for the United States are from the Bureau of Economic Analysis. For the other
countries the data are from the Lane and Milesi-Ferretti data set up to 1997. I have updated
them using national current account balance data. The updated figures should be interpreted
with a grain of salt, because I have not corrected them for valuation effects.
Sebastian Edwards 225
Table 4. Countries with Persistently Large Current Account Deficits
by World Region, 1970–2001a
Region and country Period
Industrial countries
Ireland 1978–84
New Zealand 1984–88
Latin America and Caribbean
Guyana 1979–85
Nicaragua 1984–90, 1992–2000
Asia
Bhutan 1982–89
Africa
Guinea-Bissau 1982–93
Lesotho 1995–2000
Eastern Europe
Azerbaijan 1995–99
Source: World Bank, World Development Indicators, various years.
a. A persistently large deficit is defined as one that exceeded the ninth decile for the country’s region for
at least five consecutive years.

table 4: a number of industrial nations have had—and continue to have—

a significantly larger net international liabilities position relative to GDP
than does the United States. This suggests that, at least in principle, the
U.S. NIIP could continue to deteriorate for some time into the future. But,
even if this should happen, at some point the process must come to an
end, and the U.S. position as a percentage of GDP will have to stabilize. It
makes a big difference, however, at what level it does stabilize. For exam-
ple, if, in the steady state, foreigners are willing to hold at most the equiv-
alent of 35 percent of U.S. GDP in the form of net U.S. assets, the United
States could sustain a current account deficit of only 2.1 percent of U.S.
GDP.19 If, on the other hand, foreigners’ net demand for U.S. assets were
to grow to 60 percent of GDP—which, as table 5 shows, is approximately
the ratio of (net) foreign holdings of Australian assets to that country’s
GDP—the sustainable U.S. current account deficit would be 3.6 percent
of GDP. And if foreigners are willing to hold (net) U.S. assets equivalent
to 100 percent of U.S. GDP—the figure that Mussa considered implau-
sible in the statement footnoted above—the sustainable U.S. current

19. This calculation assumes a 6 percent rate of growth of nominal GDP going forward.
See below for an analytical discussion and the relevant equations.
 226 Brookings Papers on Economic Activity, 1:2005
Table 5. Net Stock of Liabilities in Selected Industrial Countries, Selected Years
Percent of GDP
Country 1980 1985 1990 1995 2000 2003
Australia n.a. n.a. 47.4 55.1 65.2 59.1
Canada 34.7 36.3 38.0 42.4 30.6 20.6
Denmark n.a. n.a. n.a. 26.5 21.5 13.0
Finland 14.6 19.0 29.2 42.3 58.2 35.9
Iceland n.a. n.a. 48.2 49.8 55.5 66.0
New Zealand n.a. n.a. 88.7 76.6 120.8 131.0
Sweden n.a. 20.9 26.6 41.9 36.7 26.5
United States −12.9 −1.3 4.2 6.2 14.1 22.1
Source: Bureau of Economic Analysis, U.S. International Transactions and International Investment Position, various years;
Lane and Milesi-Ferretti (2001).

account deficit could be as large as 6 percent of GDP, or approximately its

current level.20
Since there are no historical precedents for a large, advanced nation
running persistently large current account deficits, it is extremely difficult
to get a clear idea of how foreigners’ demand for U.S. assets will behave
in the future. Given this lack of historical precedent, a reasonable strategy
is to model the RER dynamics to be expected if, as posited by Michael
Dooley and his coauthors,21 among others, foreigners’ demand for U.S.
assets continued to increase. This is the approach followed in the next
section.

The Analytics of Current Account

and Real Exchange Rate Adjustment

The current account and the RER are endogenous variables jointly
determined in a general equilibrium context. The key question is how
these two variables will move in response to a given exogenous shock—a
decline in capital inflows, say—if the other main variables, including eco-
nomic growth and the rate of unemployment, do not deviate significantly
from their long-term equilibrium paths. A number of authors have recently
addressed this issue using a variety of simulation and econometric models.

20. Mussa (2004).

21. Dooley, Folkerts-Landau, and Garber (2004a).
Sebastian Edwards 227

Most have asked what amount of RER adjustment would be “required” to

achieve a certain current account balance. Among these, some have con-
sidered the case where the deficit is completely eliminated,22 whereas oth-
ers have investigated the reduction of the deficit to a smaller but still
positive level. Appendix table A-1, which summarizes a selection of these
studies, shows that they have used different methodologies and reached
different conclusions.23 All, however, find quite large required adjustments
in the trade-weighted value of the dollar.24 The estimates from the studies
summarized in table A-1 are much larger than those discussed in most
investment bank newsletters and in the media.25

A Portfolio Model of the Current Account

and the Real Exchange Rate
From an analytical perspective the process of current account adjust-
ment may be broken down into two components: the dynamics of changes
in net foreign assets, and the “transfer” associated with changes in a coun-
try’s net foreign assets position. Changes in international investors’ will-
ingness to hold U.S. assets will affect total absorption of saving and
relative prices, including the RER. An increase in foreigners’ rate of accu-
mulation of U.S. assets will allow the United States to increase its absorp-
tion, generating a real appreciation and a current account deficit. In a
similar way, a reduction in the rate at which foreigners accumulate the
country’s assets—or, worse, a reduction in their holdings of domestic
assets—will result in a slowdown or a drop in absorption and a decline
in the relative price of nontradables, that is, a real depreciation. These
changes in absorption and the concomitant adjustment in relative prices
are reminiscent of discussions of the “transfer problem,” which go back at
least to the debates between John Maynard Keynes and Bertil Ohlin dur-
ing the 1920s. In large countries such as the United States, however, the

22. For example, Obstfeld and Rogoff (2000, 2004); Blanchard, Giavazzi, and Sa (this
volume).
23. See also the studies by Mann (2003, 2004), which extend her pioneering 1999
model.
24. See, for example, Blanchard, Giavazzi, and Sa (this volume).
25. Although many financial market practitioners do believe that the dollar will
weaken, they tend to expect more moderate adjustments. See, for example, the forex publi-
cations of some major investment banks.
 228 Brookings Papers on Economic Activity, 1:2005

story is more complex. First, as mentioned above, changes in relative

prices have valuation effects on net foreign asset holdings, which will feed
back into the dynamics of net foreign asset accumulation or decumula-
tion.26 Second, in a large country, changes in aggregate expenditure are
likely to affect the international terms of trade, and thus the general equi-
librium outcome of the original shock.
the basic model. Consider the following barebones portfolio model
of the current account.27 Equation 1, which is the basic equation for the
external sector (expressed in domestic currency), states that the current
account deficit (CAD) is equal to the trade deficit (TD) plus the income
account (net income payments to the rest of the world, IA) plus net trans-
fers to the rest of the world (NT):28
(1) CAD t = TD t + IA t + NTt .
The income account, in turn, is equal to

IA t = iD tf − i *F dt ,

where i is the interest rate paid on (gross) domestic assets in the hands
of foreigners Dtf, and i* is the interest rate on (gross) foreign assets held
by domestic residents Ftd. Since equation 1 is expressed in domestic
currency,
F td = Et F td ∗ ,

where E is the nominal exchange rate, defined as units of domestic cur-

rency (in this case dollars) per unit of foreign currency, and F td * denotes
(gross) foreign assets held by domestic residents, expressed in foreign
currency. Equation 1 can then be rewritten as

26. This effect has been emphasized by Lane and Milesi-Ferretti (2001, 2004a, 2004b),
Tille (2003), and Gourinchas and Rey (2005), among others. For a discussion of valuation
effects in the context of current account sustainability in emerging economies, see Edwards
(2003).
27. To concentrate on the problem at hand and to keep the analysis tractable, I have
made a number of simplifications; I have made no attempt to construct a full general equi-
librium model. Recent papers that have constructed portfolio models of the current account
include Blanchard, Giavazzi, and Sa (this volume), Edwards (1999, 2002), Gourinchas and
Rey (2005), and Kraay and Ventura (2002).
28. I have defined current account balances such that a deficit is a positive number. In
equation 1, then, negative numbers refer to a surplus.
Sebastian Edwards 229

(1′) CAD t = TD t + iδ t + ( i − i*) Ftd + NTt ,

where δ is net domestic assets in the hands of foreigners, δt = D tf − F td. The

terms iδt and (i − i*)F td capture valuation effects on the current account,
recently emphasized by a number of authors.29
Equation 2 is a portfolio balance equation that summarizes net inter-
national demand for the domestic country’s assets δt. Domestic and for-
eign assets are assumed to be imperfect substitutes. The variable α is the
share of foreigners’ wealth that they are willing to hold in the form of
the domestic country’s assets; W is world wealth, and Wc is the domestic
country’s wealth. The variable αjj is the domestic country’s asset alloca-
tion on its own assets, that is, the share of their wealth that domestic resi-
dents choose to hold in domestic assets. I assume that there is home bias
in portfolio decisions; this is reflected in the fact that α and (1 − αjj) are
below international market shares of domestic and foreign wealth. There
is no need, however, to assume that foreign and domestic investors have
the same degree of home bias. Hence the portfolio balance equation can
be written as
(2) δ t = α (Wt − W ct ) − (1 − α jj ) W ct .
An important question is how the asset allocation shares α and αjj are
determined. Under standard portfolio theory, α and αjj will depend on
expected real returns (i, i*), perceived risk (µ, µ*), and the degree of seg-
mentation of international financial markets. Here, however, I make the
simplifying assumption that α and αjj are exogenously determined. This
assumption allows me to focus on the effects of exogenous changes in
portfolio allocations—that is, exogenous changes in α and αjj—on net
asset dynamics and the current account. More specifically, I consider the
case where changes in α and αjj reduce the initial extent of home bias.
Later I discuss how the results are altered if some degree of substitutabil-
ity between domestic and foreign assets is allowed.
World wealth in foreign currency W* and in domestic currency W are
related by W *t = Wt /Et. Domestic and foreign interest rates are related by
i = i* + (dEe/E) + (µ − µ*) + k, where (dEe/E) is the expected rate of
depreciation of the domestic currency, and k is a term that captures the
effect of capital controls; in a world of perfect capital mobility, k = 0.
29. Including Lane and Milesi-Ferretti (2004a, 2004b) and Gourinchas and Rey (2005),
among others.
 230 Brookings Papers on Economic Activity, 1:2005

Note that in this model “investors” refers both to private and public
investors, including foreign central banks. Indeed, as already pointed
out, recent discussions have emphasized the key role played by foreign
(and especially Asian) central banks in helping finance the U.S. current
account deficit.
The counterpart of a current account deficit is the change in the coun-
try’s (net) assets in the hands of foreigners:

(3) CADt = ∆δ t .

Equation 4 defines the trade deficit:

(4) TDt = ∑p m
i
mi − ∑ pix xi ,

where pim and pix are prices of importables and exportables in domestic
currency; mi is demand for importables, which is assumed to depend on
the real exchange rate e, the international price of importable goods, the
country’s real income y, and other factors, including the degree of trade
protection v. Demand for exports xi, on the other hand, depends on the
RER, the international price of exportables, rest-of-world real income y*,
and other factors u:

(5) mi = mi ( e, y, v ); xi = xi ( e, y*, u ) .

Variables mi and xi, in turn, may be interpreted as excess demand for

importables and supply of exportables, respectively, in the domestic
country. The basic version of the model assumes that the law of one price
holds for both importables and exportables: pim = Epim*; pix = Epix*. How-
ever, in the simulation exercises below, alternative assumptions can be
made, including that exporters and importers price to market. Equation 6
is the equilibrium condition for the nontradable goods market in the
domestic country, where S tN is the supply of nontradables in period t,
assumed to depend on the RER and other factors z, and D tN is demand for
nontradables:

(6) S tN ( et , zt ) = DtN ( et , yt ).

The domestic price level P is assumed to be a geometric average of the

nominal prices of tradable goods (importables and exportables) and non-
tradable goods:
Sebastian Edwards 231

Pt = ( ptm ) ( p tx ) ( p Nt )
a b (1− a − b )
.

Equation 7 defines the real exchange rate:

(7) et = Pt Et P*,
t

where P*t is the foreign price level. As before, an increase in e represents

a real appreciation.
The interpretation of this model is simple. The domestic country can
run a current account deficit only to the extent that foreign investors are
willing to increase their net holdings of domestic assets—that is, to the
extent that ∆δt > 0. Once ∆δt is known, and for given values of other key
variables, it is possible to derive the real exchange rate e consistent with
the prevailing current account deficit or surplus. A particularly interesting
exercise, given the current U.S. situation, is to analyze how exogenous
changes in portfolio preferences—that is, changes in α, αjj, or both—will
affect the current account and the RER.
Closing the model would require specifying a number of market clear-
ing conditions, including the saving and investment equations for the
world economy; and the world market clearing conditions for each
importable and exportable good. These equilibrium conditions determine
endogenously both interest rates and all relevant tradable goods prices.
Doing this, however, would make the model significantly more complex
than is required for dealing with the problem at hand. For this reason I
work with a partial equilibrium version of the model, under alternative
assumptions regarding these variables’ behavior.30
It is important to emphasize that current account adjustment not only
implies changes in the RER, but also requires changes in saving and
investment in the domestic country (here the United States) and the rest of
the world. From a policy perspective the adjustment in domestic saving
would be greatly facilitated by an increase in public sector saving.
PORTFOLIO EQUILIBRIUM, DYNAMICS, AND CURRENT ACCOUNT
SUSTAINABILITY. External sustainability requires that a country’s net
external liabilities stabilize at a level compatible with foreigners’ net
demand for these claims, as specified by equation 2. On the assumption
that the domestic country’s wealth is a multiple λ of its (potential or full-

30. Most recent models on global imbalances and the U.S. current account have used a
partial equilibrium framework in the simulation phase.
 232 Brookings Papers on Economic Activity, 1:2005

employment) GDP, and that its wealth is a fraction β of world wealth W,

it is possible to rewrite (international) net demand for the country’s assets
as δ = [αθ − (1 − αjj)]λY, where Y is potential GDP, and θ = (1 − β)/β =
EW f*/Wc, where W f* is rest-of-world wealth expressed in foreign currency.
Denoting γ* = [αθ − (1 − αjj)]λ, then δ = γ*Y. This means that in
long-run equilibrium net international demand for the domestic country’s
assets can be expressed as a proportion γ* of its potential or sustainable
GDP. The determinants of this factor of proportionality γ* depend on rel-
ative returns and the perceived risk of the domestic country and of the rest
of the world, as well as on the degree of integration of international finan-
cial markets.
If g is the country’s sustainable rate of growth and π the country’s
long-term rate of inflation, the “sustainable” ratio of the current account
deficit to GDP is given by
(8) CAD/Y = ( g + π ) αθ − (1 − α jj )  λ = γ * ( g + π ) .
Notice that, if [αθ − (1 − αjj)] < 0, domestic residents’ demand for foreign
assets exceeds foreigners’ demand for domestic country assets. Under
these circumstances the country will have to run a current account surplus
in order to maintain a stable net external assets-to-GDP ratio. Most stud-
ies of the sustainability of the U.S. current account have used equations of
this type. Mussa,31 for example, argues that in long-term equilibrium γ * is
likely to be around 0.50.32 In long-run equilibrium the sustainable trade
balance will be given by TD/Y = (g − r)γ*, where r is the real interest rate.
In this model, as in earlier models developed by myself and by Aart
Kraay and Jaume Ventura,33 additional saving will be allocated in a way
that maintains domestic and foreign assets in the same proportion as in
the original portfolio. Kraay and Ventura have shown that models that
combine this assumption with the assumption of transactions costs in
investment go a long way toward explaining international current account
behavior in a large number of countries.
If the degree of perceived riskiness of the domestic country (exoge-
nously) declines, α, and thus γ *, will increase. As a result, the sustainable

31. Mussa (2004).

32. See also Edwards (1995), Ades and Kaune (1997), and O’Neill and Hatzius (2004)
for current account sustainability analyses of this type.
33. Edwards (1999); Kraay and Ventura (2002).
Sebastian Edwards 233

current account balance will deteriorate (see equation 9). Equally impor-
tant, changes in portfolio allocation, generated by changes in α or αjj, will
generate a dynamic adjustment process, during which the current account
will differ from its long-run sustainable level. This transitional dynamic
can be incorporated into the model through the following equation:
(9) ( CAD/Y ) t
= ( g + π ) γ *t + ψ ( γ *t − γ t − 1 ) − κ ( CAD/Y )t − 1 − ( g + π ) γ *t  .

According to equation 9, short-term deviations of the current account

from its long-run level can result from two forces. The first is a traditional
stock adjustment term (γ *t − γt−1) that captures deviations between the
demanded and the actual stock of the country’s assets in the hands of for-
eign investors. The coefficient ψ is the speed of adjustment, which will
depend on a number of factors, including the degree of capital mobility in
the country in question. The second force affecting this dynamic process,
captured by −κ[(CAD/Y)t−1 − (g + π)γ *t ] in equation 9, is a self-correcting
term, included to ensure that some form of consumption smoothing is pres-
ent. The importance of this self-correcting term will depend on the value
of κ.34
Whether the dynamic representation in equation 9 is appropriate is, in
the final analysis, an empirical matter. As I show below, under certain
parameterizations this model does a very good job at tracking the behav-
ior of the U.S. current account during the last few years. The dynamic
behavior for the net stock of the domestic country’s assets in the hands
of foreigners, as a percentage of GDP, will be given by γt = [γt−1 +
(CAD/Y)t−1](1 + g + π)−1.
Consider the case where for some exogenous reason the home bias in
the rest of the world is reduced—that is, α in equation 2 increases. This
will result in an increase in the sustainable current account deficit (equa-
tion 8). It will also unleash a dynamic adjustment process, captured by
equation 9. During this transitional period the current account deficit will
exceed its new long-run (higher) sustainable equilibrium; that is, the cur-
rent account deficit will overshoot its new sustainable level. During the
transition the trade account will move according to the following equa-
tion: ∆(TD/Y)t = ∆(CAD/Y)t − ∆(iγ*t )− ∆[(i − i*)(Fd/Y)]t − ∆(NT/Y)t. From

34. If ψ = κ = 0, the current account will jump from one sustainable level to the next.
There are many reasons to assume that both ψ and κ are different from zero, including the
existence of adjustment costs in consumption.
 234 Brookings Papers on Economic Activity, 1:2005

equations 4 through 6, and after making some assumptions regarding the

behavior of other key variables, such as the international terms of trade
and interest rates, the following equation for the current account may be
derived (to simplify the notation, the mi and xi have been aggregated into
broad import and export categories):

(10) ∆ ( CAD/Y )t − ∆ ( iγ ∗t ) + ∆ ( i − i∗ ) ( F d /Y )t  + ∆ ( NT /Y )t

+ [ σ x (1 + ε e ) − σ m (1 + ηe )] eˆ + ( σ m − σ x ) ( π − π∗ ) + σ m η y g
− σ m ε y ∗g∗ + σ m pˆ ∗m − σ x pˆ ∗x − ( σ m − σ x ) ( g + π ) ,

where σm and σx are ratios of imports and exports to GDP; ηe < 0 and εe >
0 are the price elasticities of imports and exports, respectively; ηy and ε*y
are the elasticities of imports and exports with respect to domestic and
foreign income, respectively; g and g* represent rates of real GDP growth
at home and in the rest of the world, respectively; π and π* are domestic
and world inflation, respectively; p̂*m and p̂*x are the rates of change in
international prices of imports and exports, respectively; and ê is the
rate of change of the real exchange rate. From this equation it follows
that, in order for a real depreciation to improve the trade balance (and,
other things equal, the current account), it is required that [σx(1 + εe) −
σm(1 + ηe)] > 0.35
Although equation 10 is not a reduced-form equation, it is useful for
undertaking a number of simulation exercises.36 For example, with equa-
tions 2, 3, 9, and 10, and under assumed values for growth, inflation, and
interest rates and changes in the international terms of trade, it is possible
to analyze how changes in portfolio preferences will affect the trajectories
of the current account and the RER.

Simulation Results
The barebones model developed above may be used to compute the
current account and RER adjustments consistent with shifts in portfolio

35. Under balanced initial trade, this expression becomes the traditional Marshall-
Lerner condition.
36. In equation 10 I have assumed that di = di* = 0. Since α and αjj are exogenous, this
assumption does not affect the behavior of the RER. Later in the paper I discuss the way in
which changes in interest rates and other variables such as the international terms of trade
affect the results.
Sebastian Edwards 235

preferences by foreign and domestic investors, including a reduction in

the home bias in portfolio investment decisions.37 A first step in this
analysis is to calibrate the model. Table A-2 in the appendix presents the
parameter values used in the base-case simulation; most of these values
are taken from existing studies of the U.S. and world economies. I have
selected the values of ψ and κ that best track the actual dynamics of the
U.S. current account between 1996 and 2004; the best results are obtained
for ψ = 0.30 and κ = 0.20. I also assume that foreigners’ demand α for
U.S. assets increased gradually from 0.205 to 0.300 between 1996 and
2004 (see the values for αHistorical, and αInitial in table A-2). As may be seen
from the top left and bottom right panels of figure 5, for these parameter
values the model tracks actual current account and RER behavior for
1996–2004 quite closely.38
One limitation of this type of simulation exercise is that it is difficult to
forecast how foreign investors’ net demand for U.S. assets will behave in
the future. It is precisely for this reason that a number of authors have
avoided the issue and have instead computed the RER adjustment
“required” to eliminate the current account deficit completely.39 Here I
take a different approach: instead of assuming that the current account
deficit has to be reduced to zero or some other arbitrary number, I analyze
the dynamic of the current account under alternative assumptions regard-
ing foreigners’ net demand for U.S. assets. I am particularly interested in
understanding what is likely to happen under an optimistic scenario where
foreigners’ demand for U.S. assets continues to grow. What makes this
approach particularly interesting is the finding that, even under such a
scenario, it is highly likely that, in the not-too-distant future, the U.S. cur-
rent account will undergo a significant reversal.
As table A-2 shows, in these simulation exercises I assume a gradual
portfolio adjustment over the next five years. I assume that α increases
from its current value of 0.30 to 0.40 by 2010 and that αjj falls from 0.73

37. In fact, there are indications that the process of international capital market integra-
tion will continue in the future, as some of the largest emerging economies, including
China, are increasingly allowing their nationals to invest abroad. See, for example, “China
to Seek Full Currency Conversion,” Financial Times, February 28, 2005, p. 6.
38. To obtain the best possible historical fit for the model, I incorporated into the his-
torical simulation changes in the terms of trade that track what was observed in 1996–2004.
39. Obstfeld and Rogoff (2000, 2004). For similar approaches see Mussa (2004) and
Blanchard, Giavazzi, and Sa (this volume).
Figure 5. Results of Base-Case Simulations

Current account deficit Trade account deficit

Percent of GDP Percent of GDP

5
6 Simulated
4 Trade deficit
4 3

Actual 2
2
1

Net U.S. international liabilities Real exchange rate

Percent of GDP Index

Actual
60
120
50 Simulated

40 110

30
100
20

2004 2010 2016 2022 2004 2010 2016 2022

Source: Author’s calculations using model described in the text.

Sebastian Edwards 237

to 0.71 during the same period. This adjustment implies a reduction in

home bias both in the rest of the world and in the United States. In the
base-case scenario the assumed portfolio adjustment is equivalent to for-
eigners doubling their net demand for U.S. assets to an amount equal to
60 percent of U.S. GDP. This is a very large number. Indeed, it implies
that, under the assumptions that g = 0.03 and π = 0.023, during the five
years from 2005 to 2010 the U.S. NIIP will deteriorate by a further
$5.7 trillion.
Before proceeding, the following assumptions made in the base-case
scenario deserve some comment (see table A-2 for details). First, I
assume that the United States and the rest of the world grow at the same
rate (g = g*). This is consistent with the idea that, while the United States
will grow faster than Europe and Japan, the rest of the world—including
China and India—will continue to grow very rapidly. In a number of
alternative simulations I consider different values for growth. A second
assumption concerns the values of the key elasticities, which have been
taken from existing studies on the U.S. and global economies.40 Two
important characteristics of these elasticities are that the income elasticity
for U.S. imports exceeds that for imports by the rest of the world (the so-
called Houthakker-Magee phenomenon), and that the RER elasticity of
U.S. imports exceeds (in absolute terms) that of U.S. exports. Finally, in
the base-case scenario I assume that the adjustment has no effect on the
m = p̂*
international terms of trade ( p̂* x = 0); in alternative simulations I con-
sider the case where the terms of trade change.
BASE-CASE SIMULATIONS. Figure 5 presents the results of this base-
case exercise. In these simulations, 2005 should be interpreted as the “ini-
tial” period; the previous eight years (the shaded area) represent recent
history. The figure presents simulation results for the current account, the
trade account, net U.S. assets held by foreigners, and the trade-weighted
RER index (and, for the eight historical years, the actual RER index). The
most salient features of the base-case simulation are the following:
—Under the deliberately optimistic assumption of a significant increase
in foreign net demand for U.S. assets, the current account deficit contin-
ues to increase until it peaks at 7.3 percent of U.S. GDP. From that point
onward the deficit declines toward its new steady state of 3.2 percent of
GDP.

40. See Hooper, Johnson, and Marquez (2000).

 238 Brookings Papers on Economic Activity, 1:2005

—Once the current account deficit reaches its peak, the reversal is
quite sharp. In the base-case scenario, during the first three years of
adjustment the deficit is reduced by 3.1 percent of GDP. The reversal of
the trade deficit is even sharper. The reason is that, with a larger net U.S.
debtor position, net payments (interest and dividends) to foreign investors
increase significantly relative to GDP.
—As the bottom right panel shows, once the process of current account
reversal begins, the trade-weighted RER index falls rapidly: during the
first three years of adjustment, the accumulated real depreciation is 13.3
percent. By the time the new, sustainable current account deficit is
reached, the accumulated depreciation amounts to 22.5 percent. This
result is roughly in line with other studies (table A-1). In alternative sim-
ulations in which the valuation effect of dollar depreciation on the U.S.
net foreign asset position is ignored, the resulting real depreciation is
larger: for example, in the first three years of adjustment the accumulated
depreciation is 16.8 percent.
—This simulation also indicates that the new steady state is associated
with a sharp depreciation: the RER falls to 19.1 percent below its initial
(2005) level.
Alternative assumptions regarding growth, inflation, interest rates, the
terms of trade, elasticities, and other key parameters will, of course, affect
the results. Except when the changes in the assumptions are extreme,
however, the main qualitative result holds: even under very optimistic
assumptions regarding foreigners’ net demand for U.S. assets, the current
account deficit is likely to go through a large reversal in the not-too-dis-
tant future.
ALTERNATIVE PORTFOLIO CHOICES. An important question is how
sensitive these results are to portfolio choices. To explore this issue,
I report in figure 6 results from a second simulation exercise, which
assumes that, after increasing their net holdings of U.S. assets to 60 per-
cent of U.S. GDP by 2010, foreign investors make a new portfolio adjust-
ment and gradually reduce their desired holdings of U.S. assets to
50 percent of U.S. GDP. (The bottom left panel of the figure depicts the
trajectory of net foreign holdings of U.S. assets in this simulation.) As
the figure shows, in this case the current account reversal is signifi-
cantly more abrupt, as is the depreciation in terms of the trade-weighted
RER index. In the first three years of the adjustment, the current account
deficit declines by 5.3 percent of GDP, and the accumulated depreciation
Figure 6. Results of Simulations Using Alternative Assumptions

Current account deficit Trade account deficit

Percent of GDP Percent of GDP

6 Simulated 4

4 2

2 Actual
0

Net U.S. international liabilities Real exchange rate

Percent of GDP Index
Actual
50 120

40 110

30 100 Simulated

20 90

2004 2010 2016 2022 2004 2010 2016 2022

Source: Author’s calculations using model described in the text.

 240 Brookings Papers on Economic Activity, 1:2005

is 23.7 percent. Moreover, as the top right panel of the figure shows, by
the third year of the adjustment (2011 in the simulation), the trade deficit
has turned into a surplus. It is important to keep in mind that this simula-
tion still assumes that the long-run net demand by foreigners for U.S.
assets is still significantly higher—20 percent of GDP higher, to be pre-
cise—than today. Because of space considerations, I have not presented
the results of more pessimistic scenarios in which foreigners reduce their
net demand for U.S. assets below the current level. Suffice it to say that in
those scenarios the current account reversal is even more pronounced, as
is the concomitant real depreciation.
DOES ADJUSTMENT NEED TO BE ABRUPT? The results presented in fig-
ures 5 and 6, and in particular the abrupt current account reversal that
takes place after the deficit peak is reached, depend on the assumptions
made about parameters ψ and κ; different values of these parameters
would result in different dynamics. For instance, if in the future the
dynamic of the adjustment process changes, such that ψ declines while κ
increases, this would result in a more gradual convergence of the current
account deficit to its new, sustainable level. To take a concrete example,
values of ψ = 0.20 and κ = 0.35 would result in an accumulated compres-
sion of the current account of 1.9 percent of GDP during the first four
years of the adjustment process. This is a significantly less drastic adjust-
ment than the 6 percent of GDP obtained in figure 6, and it shows that an
abrupt collapse in the deficit is not unavoidable. Furthermore, in this sim-
ulation the current account deficit would peak at 6.2 percent of GDP
(results not shown) rather than at 7.3 percent as in figure 6.
The process of net accumulation of U.S. assets by foreigners may also
differ from what I have assumed in both simulations. For instance, if they
slow their accumulation of U.S. assets, or if they stretch the process over
a longer period, the eventual adjustment would be less abrupt than is
depicted in figures 5 and 6. The real depreciation of the trade-weighted
dollar might also be less pronounced. This would be the case, for exam-
ple, if U.S. saving were to increase in the next few years, moving closer
to its historical average. In that case expenditure reduction would play
a more significant role in the adjustment, and expenditure switching
(through dollar depreciation) would be less important.
The simulations discussed above assumed an exogenously given rate
of growth of GDP. This, of course, need not be the case in reality. It is
likely, in fact, that current account reversals of the type and magnitude
Sebastian Edwards 241

suggested by the simulation results will have an effect on real economic

activity.41 In the next section I use a new, comparative cross-country data
set to investigate the real consequences of current account reversals in
the world economy since 1971. This comparative analysis will help give
some idea of the possible effects of a U.S. current account reversal like
that in the simulations in figures 5 and 6.

How Costly Are Current Account Reversals?

An International Comparative Analysis

The main message of the simulation exercises just presented is that,

even under optimistic scenarios where foreigners’ demand for U.S. assets
increases significantly, the U.S. current account experiences a significant
reversal in the not-too-distant future. But what will be the nature of the
adjustment process? I address this issue here by analyzing the interna-
tional experience with current account reversals in the period 1971–2001.
Although the U.S. case is unique, both because of the size of its economy
and because the dollar is the world’s main vehicle currency, an analysis
of the international experience will shed some light on the likely nature of
the adjustment. A particularly important question is whether this adjust-
ment will entail real costs in the form of slower (or negative) growth
and higher unemployment. Previous studies have generated conflicting
results: after analyzing the evidence from a large number of countries,
Gian Maria Milesi-Ferretti and Assaf Razin conclude that major current
account reversals have not been costly: “reversals,” they claim, “are not
systematically associated with a growth slowdown.42 Jeffrey Frankel and
Eduardo Cavallo, on the other hand, conclude that sudden stops of capi-
tal inflows (a phenomenon closely related to reversals) have resulted in
growth slowdowns.43

41. See the pioneering study on current account reversals by Milesi-Ferretti and Razin
(2000). See also Edwards (2004).
42. Milesi-Ferretti and Razin (2000, p. 303).
43. Frankel and Cavallo (2004).
 242 Brookings Papers on Economic Activity, 1:2005

In what follows I analyze several aspects of current account reversals,

including44
—their incidence
—the relationship between reversals and sudden stops of capital
inflows
—the relationship between reversals and depreciation
—the factors determining the probability of a country experiencing a
reversal, and
—the costs, in terms of slower growth, of reversals.
In analyzing these issues I rely on two complementary statistical
approaches: First, I use nonparametric tests to analyze the incidence and
main characteristics of current account reversals. Second, I use panel
regression-based analyses to estimate the probability of a country experi-
encing a reversal, and the cost of such a reversal in terms of a short-term
decline in output growth. Although the data set covers all regions of the
world, in an effort to shed light on the U.S. case, I emphasize the experi-
ence of large countries.45

Current Account Reversals during 1971–2001:

The International Evidence
I use two definitions of current account reversals: I define a type I
reversal as a reduction in the current account deficit of at least 6 percent of
GDP within a three-year period, and a type II reversal as a reduction in the
current account deficit of at least 4 percent of GDP in one year. (In both
cases the reversal is recorded as occurring in the year when the episode
ends. For example, if a country’s current account deficit declined by
7 percent of GDP between 1980 and 1982, the episode is recorded as
having taken place in 1982. Also, for an episode to count as a current

44. In Edwards (2004) I used a smaller data set to investigate reversals in emerging
economies. In that paper, however, I did not consider the experience of large or industrial
countries with reversals. I also used a very simple framework for analyzing growth. In con-
trast, in this section I use a two-step dynamic of growth approach.
45. Croke, Kamin, and Leduc (2005) recently analyzed the nature of current account
adjustments in industrial economies. Their analysis differs from mine in several respects.
First, they concentrate on milder current account adjustments; second, their sample
includes only industrial countries; and, third, they are interested in analyzing whether there
is likely to be a “disorderly” adjustment, defined as a situation of financial disruption.
Sebastian Edwards 243

account deficit reversal, the initial balance has to be indeed a deficit.46)

Thus, in a type I reversal, the magnitude of the adjustment is more pro-
nounced than in a type II reversal but is distributed over a longer period.47
Table 6 presents data on the incidence of both types of reversal for
the complete sample of countries as well as for each of the six groups
of countries considered in the previous section. For the overall sample the
incidence of type I reversals is 9.2 percent, and that of a type II reversal
is 11.8 percent. The incidence of reversals among the industrial countries
is much smaller, however, at 2.7 percent and 2.0 percent, respectively.
Indeed, the Pearson χ2 and F tests reported in table 6 indicate that the
hypothesis of equal incidence of reversals across regions is rejected
strongly.
The industrial countries that experienced type I current account rever-
sals during the period are Finland (in 1978 and 1994), Greece (1988),
Ireland (1984), New Zealand (1977–78 and 1988–89), Norway
(1979–80, 1989, and 2000), and Portugal (1979 and 1984–85). Those
that experienced type II reversals are Austria (1982), Canada (1982),
Greece (1986), Iceland (1983 and 1986), Ireland (1975), Italy (1975),
Malta (1997), New Zealand (1978), Norway (1989), and Portugal
(1982–83 and 1985). With the exception of Italy and Canada, all of these
countries are economically very small, underlining the point that there
are no historical precedents of large countries undergoing profound cur-
rent account adjustments. As pointed out above, this implies that the
results reported here on current account reversals should be interpreted
with a grain of salt and should not be mechanically extended to the case
of the United States.
The analysis presented above distinguished countries by their stage
of development and world region. An alternative way of dividing the
sample, and one that is particularly relevant for deriving lessons for
the United States, is by economic size. I define “large” countries as
those whose GDP placed them in the top 25 percent of the sample distribu-
tion in 1995 (by this criterion there are forty-one “large” countries in the
sample). For the period 1971–2001 the incidence of type I reversals

46. These definitions differ somewhat from those used in other studies, including
Freund (2000), Milesi-Ferretti and Razin (2000), Edwards (2002), and Guidotti, Villar, and
Sturzenegger (2003).
47. Notice that it is possible for a country to have experienced both a type I and a type
II reversal during the same historical episode.
 244 Brookings Papers on Economic Activity, 1:2005
Table 6. Incidence of Current Account Reversals by World Region, 1970–2001
Percenta
Type I reversal Type II reversal
Region or country group No reversal Reversal No reversal Reversal
Industrial countries 97.3 2.7 98.0 2.0
Latin America and Caribbean 92.0 8.0 87.7 12.3
Asia 88.3 11.7 87.7 12.3
Africa 88.3 11.7 83.4 16.6
Middle East 86.6 13.4 85.0 15.0
Eastern Europe 90.7 9.3 88.9 11.1
All countries 90.8 9.2 88.2 11.8
Summary statistics
Uncorrected Pearson χ2 (5) 37.31 67.42
Design-based F test (5, 12,500) 7.46 13.08
P-value 0.00 0.00
Source: Author’s calculations using data from World Bank, World Development Indicators, various years.
a. Number of reversal episodes divided by the product of all countries in the group and all years, times 100.

among large countries is 5.3 percent, and that of type II reversals is

6.8 percent.

Current Account Reversals and Sudden Stops of Capital Inflows

In the last few years a number of authors have analyzed episodes of
sudden stops of capital inflows into a country.48 From an analytical per-
spective, sudden stops and current account reversals should be closely
related phenomena, but there is no reason for them always to occur
together. Indeed, because of changes in international reserves, it is per-
fectly possible for a country to suffer a sudden stop without simultane-
ously experiencing a current account reversal. However, in countries with
floating exchange rates, changes in international reserves tend to be rela-
tively small, and, at least in principle, the relationship between sudden
stops and reversals should be stronger.
To investigate formally the relationship between these two phenom-
ena, I define a “sudden stop” episode as an abrupt and major reduction in
capital inflows to a country that until that time had been receiving large
volumes of foreign capital. Specifically, I impose the following criteria:
capital inflows into the country in question during the two years preceding

48. See Calvo, Izquierdo, and Mejia (2004); Edwards (2004).

Sebastian Edwards 245

the episode must have been larger (relative to GDP) than those of three
quarters of the countries in its region; and net capital inflows must have
declined by at least 5 percent of GDP in the year of the episode itself.49
Table 7 presents summary statistics, for three country samples, on the
coincidence of sudden stops and current account deficit reversals (under
both definitions of the latter). The first sample consists of large countries,
defined, as stated above, as those whose GDP is in the top quartile of the
sample distribution; the second consists of industrial countries only; and
the third is the complete sample. The bottom panel of the table shows, in
the first column, that 21.1 percent of all countries experiencing a sudden
stop also faced a type I current account reversal, and 15.0 percent of those
with type I reversals also experienced (in the same year) a sudden stop.
The bottom panel also shows, in the second column, that 51 percent of all
countries subjected to a sudden stop faced a type II current account rever-
sal, and that 26.7 percent of those experiencing a type II reversal also suf-
fered (in the same year) a sudden stop. The χ2 tests indicate that in both
cases the hypothesis of independence between reversals and sudden stops
is rejected. The data for the industrial countries show that the joint inci-
dence of type I reversals and sudden stops is rather low for this group. In
fact, according to the χ2 test, the null hypothesis of independence between
the two phenomena cannot be rejected. The relationship between sudden
stops and type II reversals, however, is somewhat stronger than for type
I reversals among this group: the hypothesis of independence is rejected
(χ2 = 23.6; p = 0.00). The results for large countries are similar to those
for industrial countries.
An analysis of the lead-lag structure of reversals and sudden stops
suggest that sudden stops tend to occur either before or at the same time
(during the same year) as current account reversals. Indeed, a series of
nonparametric χ2 tests rejects the hypothesis that current account rever-
sals precede sudden stops (results not shown).

Current Account Reversals and the Exchange Rate

An important policy question—and one that is particularly relevant for
the current policy debate in the United States—is whether current account

49. To check the robustness of the results, I also used two alternative definitions of sud-
den stops, which considered a reduction in inflows of 3 and 7 percent of GDP in one year.
Detailed results using these definitions are not reported here.
 246 Brookings Papers on Economic Activity, 1:2005
Table 7. Conditional Incidence of Current Account Reversals and Sudden Stops
of Capital Inflows, 1970–2001
Percent
Country sample and eventa Type I reversal Type II reversal
Large countries
Reversal Sudden stop 9.3 25.5
Sudden stop Reversal 7.0 15.6
χ2(1) 1.3 27.5
P-value 0.26 0.00
Industrial countries
Reversal Sudden stop 5.0 18.2
Sudden stop Reversal 7.1 28.6
χ2(1) 0.4 23.6
P-value 0.51 0.00
All countries
Reversal Sudden stop 21.1 51.0
Sudden stop Reversal 15.0 26.7
χ2(1) 26.6 262.5
P-value 0.00 0.00
Source: Author’s calculations from data in World Bank, World Development Indicators, various years.
a. x|y denotes the probability of x occurring given the occurrence of y in the same year.

reversals have historically been associated with unusually large depreci-

ations. The starting point for my analysis of this issue is the construction
of an index of “external pressures” along the lines suggested by Barry
Eichengreen and others:50

(11) I t = ∆E E − ( σ E σ R )( ∆R R ) ,

where ∆E/E is the rate of change of the nominal exchange rate, ∆R/R is
the rate of change of international reserves, and σE and σR are the standard
deviations of changes in the RER and in international reserves, respec-
tively. Traditional analyses define a crisis to have occurred when It
exceeds the mean of the index plus k standard deviations. My crisis indi-
cator Ct thus takes a value of 1 (crisis) or zero (no crisis) according to the
following rule:51
50. Eichengreen, Rose, and Wyplosz (1996).
51. The pioneering work here is that by Eichengreen, Rose, and Wyplosz (1996), who
suggested that the index (equation 11) also include changes in domestic interest rates. The
original index, however, has limited use in broad comparative analyses, because most
emerging and transition economies do not have long time series on interest rates. For this
reason, most empirical analyses are based on a restricted version of the index such as
equation 11.
Sebastian Edwards 247

(12) Ct = { 1, if I t ≥ mean ( I t ) + kσ 1
0, otherwise.

Using equation 12, I define two currency crisis indicators: The first (crisis
type A) uses the traditional crisis index and assigns Ct a value of 1 when
k ≥ 3. The second (crisis type B) looks to the nominal exchange rate to
determine the value of Ct. In this case It = ∆E/E, and Ct = 1 if It ≥ mean(It)
+ kσE and 0 otherwise.
In this case the country experiences a large depreciation without a
major loss in international reserves. This indicator is more relevant for the
case of floating exchange rate countries, where changes in international
reserves are usually minimal.
I computed a number of two-way frequency tables similar to table 7
using both crisis definitions and both definitions of current account rever-
sals. I also performed χ2 tests for independence of occurrence of these
phenomena. Table 8 presents data on the shares of current account rever-
sals of both types that are accompanied by crises. Results are presented
for the same three samples as above: large countries, industrial countries,
and all countries, under three different lag structures (no lag between
reversal and crisis, crisis lagged one period, and crisis lagged two periods
into the reversal).52 The results suggest that, historically, current account
reversals and currency crises have occurred jointly in a large proportion
of cases. Consider, for example, the case of crisis type A and reversal type
I for the sample of large countries: 26.7 percent of countries with a type
I reversal experienced a contemporaneous type A crisis; 43.1 percent
experienced such a crisis in the second year of the reversal episode; and
34.5 percent of the reversals were accompanied by a crisis in the third
(and final) year of the reversal episode. Table 8 also shows that industrial
countries with reversals tended to experience currency crises during the
initial year of the reversal episode. The table also reports p-values for
χ2 tests of the independence of reversals and currency crises; in most
cases the null hypothesis that the two are independent is rejected at con-
ventional levels.

52. Data on the percentage of crises that also correspond to reversals are available from
the author on request. The results of the χ2 tests confirm those discussed above. I also used
GDP distributions for other years to define large countries and obtained similar results.
 248 Brookings Papers on Economic Activity, 1:2005

Table 8. Incidence of Current Account Reversal Associated with Currency Crisisa

Percent of countries with reversal episodes
Reversal and Crisis follows Crisis follows
crisis are reversal by reversal by
contemporaneous one year two years
Type of reversal and Type A Type B Type A Type B Type A Type B
country sample crisis crisis crisis crisis crisis crisis
Type I reversal
Large countries 26.7 16.1 43.1 17.2 34.5 13.8
(0.09) (0.01) (0.00) (0.00) (0.00) (0.05)
Industrial countries 6.7 0.0 25.0 12.5 50.0 12.5
(0.49) (0.43) (0.16) (0.10) (0.00) (0.11)
All countries 21.2 9.1 25.6 10.3 22.2 9.8
(0.10) (0.38) (0.00) (0.08) (0.01) (0.09)
Type II reversal
Large countries 31.2 18.2 42.9 15.6 29.5 12.8
(0.00) (0.00) (0.00) (0.00) (0.01) (0.04)
Industrial countries 28.6 14.3 35.7 0.0 26.7 6.7
(0.09) (0.07) (0.01) (0.43) (0.11) (0.67)
All countries 20.2 10.0 23.8 11.5 16.7 8.2
(0.05) (0.03) (0.00) (0.00) (0.86) (0.47)
Source: Author’s calculations using data from World Bank, World Development Indicators, various years.
a. Numbers in parentheses are p-values of the χ2 test.

Table 9 presents data on the distribution of exchange rate changes for

countries with type I current account reversals.53 The top panel reports
results for the nominal exchange rate (relative to the dollar; here a posi-
tive number indicates a depreciation), and the bottom panel for the trade-
weighted RER index. These changes are calculated as the cumulative
exchange rate change for the period from three years before the reversal
to the year of the reversal. For comparison I have also included the dis-
tribution of three-year nominal exchange rate changes for a control group
of countries that did not experience a current account reversal during
1970–2001. The results in the top panel indicate that countries experi-
encing reversals have tended to have significantly larger nominal depre-
ciations than the control group. Consider, for example, the case of large
countries: the average depreciation associated with a reversal episode in

53. Data on countries experiencing type II reversals are not reported here, but the
results are similar and are available from the author on request.
Sebastian Edwards 249
Table 9. Mean Cumulative Changes in Exchange Rates Following Type I
Current Account Reversalsa
Percent
Countries Countries not
experiencing experiencing Kruskal-Wallis test
Country sample type I reversal reversal (p-value)b
Nominal exchange rate
Large countries 33.1 9.2 0.00
Industrial countries 18.9 3.2 0.19
All countries 27.5 9.5 0.00
Real exchange ratec
Large countries −1.4 0.04 0.12
Industrial countries 9.3 1.6 0.55
All countries −4.0 3.6 0.00
Source: Author’s calculations.
a. Data are cumulative changes over the three years beginning with the year of the current account reversal.
b. The null hypothesis is that the data from the two samples have been drawn from the same population.
c. A positive number indicates a real appreciation.

those countries that suffered reversals is 33 percent versus only 9.2 per-
cent for the control group. To test formally whether the nominal exchange
rate behaved differently in reversal and control countries, I performed a
series of nonparametric Kruskal-Wallis χ2 tests on the equality of the
distribution of the cumulative depreciation. The null hypothesis is that
the data for the reversal countries and those for the control group have
been drawn from the same population. As table 9 shows, in the majority
of cases (two out of three) the null hypothesis is rejected at conventional
levels.
The bottom panel of table 9 presents data for the cumulative change in
the RER for the reversal and the control groups. Large countries experi-
enced a rather small real depreciation on average (1.4 percent) in the
period surrounding a current account adjustment, a result that is not statis-
tically different from that for the control group (p = .12). For the complete
sample the χ2 test indicates that the treatment and the control groups are
drawn from different populations. Perhaps surprisingly, for the industrial
countries the cumulative average change in the RER is an appreciation,
not a depreciation.
The average accumulated depreciations (both nominal and real) in
the reversal countries reported in table 9 are very small compared with
the “required” depreciations calculated in a number of studies, includ-
 250 Brookings Papers on Economic Activity, 1:2005

ing the simulations reported earlier in this paper. Obstfeld and Rogoff,54
for example, estimate that eliminating the U.S. current account deficit
would require a real depreciation of between 16 and 36 percent. Blanch-
ard, Giavazzi, and Sa have done estimates that indicate a required depre-
ciation of the trade-weighted dollar of 40 percent or more.55 One of many
possible reasons for these differences is that the United States is a very
large country, whereas the countries that have experienced reversals are
much smaller. Also, the elasticities may be different for the United
States than for the average reversal country. Yet another possibility has
to do with the level of economic activity and aggregate demand. Most
recent models of the U.S. current account assume that the economy
stays on a full-employment path. It is possible, however, that countries
that have experienced reversals have also gone through economic slow-
downs, and that a reduction in aggregate demand contributed to the
adjustment effort.

The Probability of Experiencing a Current Account Reversal

To better understand the forces behind current account reversals, I esti-
mated a number of equations on the probability of experiencing a rever-
sal, using panel data and the following empirical model:

(13) ρ jt =  1, if ρ*jt > 0
 0, otherwise.
(14) ρ*jt = α
jt + ε jt .

Variable ρjt takes a value of 1 if country j experienced a current

account reversal in period t, and zero if it did not. Whether the country
experiences a current account reversal is assumed to be the result of an
unobserved latent variable ρ*, jt which in turn is assumed to depend linearly
on vector
jt. The error term εjt is given by a variance component model:
εjt = vj + µjt, where vj is independent and identically distributed (i.i.d.) with
zero mean and variance σ v2 and µjt is normally distributed with zero mean
and variance σµ2 = 1. The data set used covers eighty-seven countries for
the 1970–2000 period; data are not available for every country for every

54. Obstfeld and Rogoff (2004).

55. Blanchard, Giavazzi, and Sa (this volume).
Sebastian Edwards 251

year, however. See appendix table A-3 for exact definitions and data
sources.
In determining the specification of this probit model, I followed the lit-
erature on external crises and included the following covariates:56 the one-
year-lagged ratio of the current account deficit to GDP; a “sudden stop”
dummy that takes the value of 1 if the country experienced a sudden stop
in capital inflows in the previous year; an index of the occurrence of sud-
den stops in the same region in the same year (to capture the effect of
regional contagion); the one-year-lagged ratio of gross external debt to
GDP;57 the one-year-lagged rate of growth of domestic credit; the one-
year-lagged ratio of the country’s fiscal deficit to GDP; and the logarithm
of the country’s initial GDP per capita. The regressions were performed
with and without the fiscal deficit variable for both measures of current
account reversal.
Table 10 presents the results of estimating this variance-component
probit model for a sample of large countries, defined as before. The vast
majority of the coefficients have the expected sign, and most are signifi-
cant at conventional levels. The results may be summarized as follows: A
larger current account deficit increases the probability of a reversal in the
following year, as does a sudden stop of capital inflows. Countries with
higher GDP per capita have a lower probability of a reversal. The results
do not provide strong support for the contagion hypothesis: the variable
that measures the incidence of sudden stops in the county’s region is sig-
nificant in only one of the four equations (although its sign is always pos-
itive). There is evidence that an increase in a country’s gross external debt
increases the likelihood of a reversal, and that larger public sector deficits
increase the probability of a type II reversal. Countries with looser mone-
tary policy, as measured by growth in domestic credit, also have a higher
probability of experiencing a reversal. Although the United States is a
very special case, the results reported in table 10 provide some support for
the idea that, during the last few years, the probability of the United States
experiencing a current account reversal has increased. Indeed, the United
States has experienced a steady increase in some important determinants

56. See, for example, Frankel and Rose (1996), Milesi-Ferretti and Razin (2000), and
Edwards (2002).
57. Ideally, one would want to have data for net debt; however, data on net liabilities
are unavailable for most countries.
 252 Brookings Papers on Economic Activity, 1:2005
Table 10. Probability of Current Account Reversals in Large Countries:
Random-Effects Probit Regressionsa
Type I reversal Type II reversal
b
Independent variable 10-1 10-2 10-3 10-4
Ratio of current account deficit 0.05 0.05 0.19 0.19
to GDP (1.65)* (1.63)* (5.46)*** (5.53)***
Occurrence of sudden stop 0.82 0.83 0.93 0.83
in country (2.06)** (2.08)** (2.46)** (2.24)**
Index of sudden stops in region 0.78 0.80 1.42 1.64
(0.66) (0.68) (1.54) (1.84)*
Ratio of external debt to GDP 0.01 0.01 0.001 0.001
(2.81)*** (2.88)*** (0.29) (0.32)
Domestic credit growth 0.001 0.001 0.0002 0.0003
(2.50)** (2.52)** (1.65)* (1.71)*
Ratio of fiscal deficit to GDP −0.004 0.05
(0.12) (1.85)*
Initial GDP per capita −0.28 −0.29 −0.15 −0.16
(2.19)** (2.23)** (1.57) (1.66)*
No. of observations 545 582 557 597
No. of countries 36 37 36 37
Source: Author’s regressions.
a. Results obtained from estimating the model in equations 13 and 14 in the text on unbalanced panel data for the sample of large
countries. Numbers in parentheses are z statistics (in absolute value); all equations include country dummy variables, results for
which are not reported. Asterisks indicate statistical significance at the ***1 percent, **5 percent, and *10 percent levels.
b. All independent variables are lagged one period.

of reversals, such as its gross international debt, its fiscal deficit, and the
current account deficit itself.

Current Account Reversals and Growth

I investigate next the relationship between current account reversals
and real economic performance, with particular attention to the following
issues: whether, historically, abrupt current account adjustments have had
an effect on GDP growth; whether sudden stops and current account
reversals have had similar impacts on growth; and whether the effects of
reversals depend on the structural characteristics of the country in ques-
tion, including its economic size, its openness to trade, and the extent to
which it restricts capital mobility. In addressing these issues, I emphasize
the case of large countries; as a comparison, however, I also provide
results for the complete sample of large and small countries.
Previous analyses of the real effects of current account reversals have
reached different conclusions. Milesi-Ferretti and Razin, for example,
Sebastian Edwards 253

used before-and-after analyses as well as cross-country regressions to

address this issue, concluding that “reversal events seem to entail sub-
stantial changes in macroeconomic performance between the period
before and the period after the crisis but are not systematically associated
with a growth slowdown.”58 On the other hand, in a previous paper I used
dynamic panel regression analysis and concluded that major current
account reversals had a negative effect on investment, and that they had
“a negative effect on GDP per capita growth, even after controlling for
investment.”59
GROWTH EFFECTS OF CURRENT ACCOUNT REVERSALS AND SUDDEN
STOPS: AN ECONOMETRIC MODEL. The point of departure for the empiri-
cal analysis is a two-equation formulation for the dynamics of real growth
of GDP per capita in country j in period t. Equation 15 is the long-run
GDP growth equation, and equation 16 captures the growth dynamics:
(15) g
j = α + x jβ + rj θ +
j .
(16) ∆g jt = λ  g
j − g jt −1  + ϕv jt + γu jt + ε jt .
I use the following notation: g̃j is the long-run rate of real growth in GDP
per capita in country j; xj is a vector of structural, institutional, and policy
variables (identified below) that determine long-run growth; rj is a vector
of regional dummies; α, β, and θ are parameters to be estimated; and
j
is an error term assumed to be heteroskedastic. In equation 16, gjt is
the rate of growth of GDP per capita in country j in period t. The terms vjt
and ujt are shocks, assumed to have zero mean and finite variance and
to be mutually uncorrelated. Specifically, vjt is assumed to be an external
terms-of-trade shock, whereas ujt captures other shocks, including current
account reversals and sudden stops of capital inflows. εjt is an error term,
which is assumed to have a variance component form, and λ, ϕ, and γ are
parameters that determine the particular characteristics of the growth
process. Equation 16 has the form of an equilibrium correction model
and states that the actual rate of growth in period t will deviate from the
long-run rate of growth because of the existence of three types of shocks:
vjt, ujt, and εjt. Over time, however, the actual rate of growth will tend to

58. Milesi-Ferretti and Razin (2000, p. 303, emphasis added).

59. Edwards (2002, p. 52). In a recent paper, Guidotti, Villar, and Sturzenegger (2003)
consider the role of openness in an analysis of import and export behavior in the aftermath
of a reversal. See also Frankel and Cavallo (2004).
 254 Brookings Papers on Economic Activity, 1:2005

converge toward its long-run value, with the rate of convergence given by
λ. Parameter ϕ in equation 16 is expected to be positive, indicating that an
improvement in the terms of trade will result in a (temporary) accelera-
tion in the rate of growth, and that negative terms-of-trade shocks are
expected to have a negative effect on gjt.60 From the perspective of the
present analysis, a key issue is whether current account reversals and
sudden stops reduce growth; that is, whether coefficient γ is significantly
negative. In estimating equation 16 I used dummy variables for sudden
stops and reversals. An important question, addressed in detail below, is
whether the effects of different shocks on growth are different for coun-
tries with different structural characteristics, such as the degree of trade
and capital account openness.
Equations 15 and 16 are estimated using a two-step procedure. In the
first step I estimate the long-run growth equation 15 using a cross-country
data set. These data are averages for 1974–2001, and the estimation cor-
rects for heteroskedasticity. These first-stage estimates are then used to
generate long-run predicted growth rates to replace g~j in the equilibrium
error correction model (equation 16). In the second step I estimate equa-
tion 16 using the generalized least squares (GLS) method for unbalanced
panels; I use both random effects and fixed effects estimation proce-
dures (only the former are reported here). The data are annual data for 157
countries for 1970–2000; data are not available for every country for
every year, however. (See appendix table A-3 for exact data definitions
and sources.)
In estimating equation 15, I followed the standard literature on growth,
as summarized by Robert Barro and Xavier Sala-i-Martin, Jeffrey Sachs
and Andrew Warner, and David Dollar, among others.61 I assume that
the long-run rate of growth of GDP g̃j depends on a number of structural,
policy, and social variables: the equation includes the logarithm of ini-
tial GDP per capita, the investment ratio, the secondary education cover-
age rate (as a proxy for human capital), an index of the degree of
openness of the economy to trade and capital flows, the ratio of govern-
ment consumption to GDP, and regional dummies. The results obtained
from these first-stage estimates are not reported but are available upon
request.

60. See Edwards and Levy-Yeyati (forthcoming) for details.

61. Barro and Sala-i-Martin (1995); Sachs and Warner (1995); Dollar (1992).
Sebastian Edwards 255

Table 11 presents the results of the second-stage estimation of the

growth dynamics equation (equation 16), using random effects. The top
panel reports results for the sample of large countries, and the bottom
panel for the complete sample. The equations whose results are reported
in columns 11-1 and 11-2 include the type I and type II reversal dummies,
respectively. Column 11-3 includes the sudden stops indicator and neither
reversal dummy. Columns 11-4 and 11-5 include both the sudden stops
indicator and the type I or the type II reversal variable, respectively, as
regressors.62
The results in table 11 may be summarized as follows: The estimated
coefficient on the growth gap is, as expected, positive, significant, and
smaller than 1. The estimates are on the high side (between 0.66 and
0.72), suggesting that, on average, deviations between long-run and actual
growth get eliminated rather quickly. For instance, according to the
results in column 11-1, approximately 85 percent of a shock to real
growth in GDP per capita will be eliminated within three years. Also,
as expected, the estimated coefficients on the terms-of-trade shock are
always positive and statistically significant, indicating that an improve-
ment in the terms of trade results in an acceleration, and a deterioration
in a deceleration, in the rate of growth of real GDP per capita. As may
be seen from columns 11-1 and 11-2, the coefficients on both the cur-
rent account reversal variables are significantly negative, indicating
that reversals result in a deceleration of growth. For large countries
these results suggest that, on average, a type I reversal is associated with
a reduction of GDP growth by 2.1 percentage points. This effect is elim-
inated gradually as g converges toward g̃j. In the case of type II rever-
sals, the estimated negative effect on GDP growth is even larger, at
−4.1 percentage points. The results in column 11-3 show that countries
that have experienced a sudden stop of capital inflows have also tended
to experience a reduction in GDP growth: for large countries the point
estimate is −2.4 percentage points. This is the case whether or not the
country in question has also suffered a current account reversal. The equa-
tions reported in the last two columns in table 11 include both the current
account reversal (type I or type II) and sudden stop indicators. The
results in column 11-5 suggest that higher costs of adjustment have been

62. In the analysis that follows, and in order to focus the discussion, I concentrate on
the effects of current account reversals.
 256 Brookings Papers on Economic Activity, 1:2005
Table 11. Impact of Current Account Reversals and Sudden Stops on Economic Growth:
Random-Effects GLS Regressionsa
Sample and
independent variable 11-1 11-2 11-3 11-4 11-5
Large countries
Growth gapb 0.67 0.72 0.68 0.66 0.71
(21.20)*** (25.33)*** (22.82)*** (20.54)*** (24.60)***
Change in terms of 0.09 0.10 0.08 0.08 0.10
trade (7.88)*** (10.30)*** (7.99)*** (7.34)*** (9.52)***
Type I current −2.12 −2.11
account reversal (3.94)*** (3.89)***
Type II current −4.13 −3.74
account reversal (9.34)*** (7.94)***
Sudden stop of −2.36 −2.39 −1.37
capital inflows (3.99)*** (3.99)*** (2.36)**
Constant −0.28 −0.21 −0.31 −0.18 −0.18
(2.10)** (1.70)* (2.36)** (1.36) (1.39)
No. of observations 799 846 811 764 810
No. of countries 41 41 41 41 41
Adjusted R2 0.41 0.50 0.45 0.42 0.50

All countries
Growth gapb 0.82 0.82 0.81 0.82 0.82
(40.26)*** (42.10)*** (40.18)*** (38.93)*** (40.76)***
Change in terms of 0.07 0.08 0.07 0.07 0.08
trade (11.77)*** (12.65)*** (11.31)*** (11.10)*** (12.18)***
Type I current −1.04 −0.73
account reversal (3.00)*** (2.03)**
Type II current −2.01 −1.80
account reversal (6.64)*** (5.50)***
Sudden stop of −1.23 −1.02 −0.53
capital inflows (2.82)*** (2.28)** (1.19)
Constant −0.30 −0.15 −0.27 −0.26 −0.14
(2.26)** (1.16) (2.62)*** (2.33)** (1.32)
No. of observations 1,723 1,821 1,641 1,546 1,635
No. of countries 90 90 81 81 81
Adjusted R2 0.48 0.49 0.51 0.52 0.51
Source: Author’s regressions.
a. Results obtained from estimating the model in equations 15 and 16 in the text on unbalanced panel data. The dependent variable is
the change in the growth rate of GDP per capita (in percentage points). Numbers in parentheses are t statistics (in absolute value); all
regressions include country dummy variables, results for which are not reported. Asterisks indicate statistical significance at the ***
1 percent, **5 percent, and *10 percent levels.
b. Difference between estimated long-run and actual annual growth rates of real GDP, in percentage points.

associated with type II reversals: the coefficient on the type II dummy

variable is more than twice as large (in absolute terms) as that on the sud-
den stop variable in the same equation. According to this equation, coun-
tries that have experienced both a reversal and a sudden stop experienced,
Sebastian Edwards 257

on average, a decline in growth in GDP per capita of 5.1 percentage

points.
To summarize, the results presented in table 11 are revealing and cast
some light on the likely costs of a future current account reversal in the
United States. Historically, large countries that have suffered such rever-
sals have experienced deep reductions in GDP growth. These estimates
indicate that, on average, and with other factors unchanged, the decline in
growth in GDP per capita has been in the range of 2.1 to 4.1 percentage
points in the first year of the adjustment. Three years after the initial
adjustment, GDP growth will still be below its long-run trend.
EXTENSIONS, ENDOGENEITY, AND ROBUSTNESS. Here I discuss some
extensions of the model and examine the robustness of the estimates,
including possible endogeneity bias. Specifically, I address the role of
countries’ structural characteristics in determining the costs of adjustment,
present results from instrumental variables GLS regressions with random
effects, and consider the effects of changes in the terms of trade.
Openness and the costs of adjustment. Recent studies on the econom-
ics of external adjustment have emphasized the role of openness to trade.
Guillermo Calvo, Alejandro Izquierdo, and Luís-Fernando Mejia,
Frankel and Cavallo, and I, among others, have found that countries that
are more open to international trade tend to incur a lower cost of adjust-
ment to a current account reversal.63 These studies, however, do not dis-
tinguish between large and small countries or between openness in the
trade account and openness in the capital account. To investigate whether
openness has historically affected the cost of external adjustment in large
countries, I added two interactive regressors to equation 16: the first inter-
acts the reversal indicator with trade openness, and the second with an
index of the country’s degree of international capital mobility. Trade
openness is proxied by the fitted value of the ratio of imports plus exports
to GDP obtained from a gravity model of bilateral trade.64 The index on
international capital mobility is one that I developed in a previous paper;65
the index ranges from 0 to 100, with higher numbers denoting greater cap-
ital mobility. The results, presented in table 12, show that the coefficients

63. Calvo, Izquierdo, and Mejia (2004); Frankel and Cavallo (2004); Edwards (2004).
64. The use of gravity trade equations to generate instruments in panel estimation was
pioneered by Jeffrey Frankel. See, for example, Frankel and Cavallo (2004).
65. Edwards (forthcoming).
 258 Brookings Papers on Economic Activity, 1:2005
Table 12. Impact of Trade Openness and Capital Mobility on Growth in Large Coun-
tries: Random-Effects GLS Regressionsa
Independent variable 12-1 12-2 12-3 12-4
Growth gapb 0.67 0.67 0.68 0.68
(21.17)*** (21.12)*** (22.35)*** (22.40)***
Change in terms of trade 0.09 0.09 0.09 0.09
(7.78)*** (7.83)*** (8.77)*** (8.79)***
Type I current account reversal −3.48 −3.84
(1.98)** (4.42)***
Type I current account reversal × 0.27 0.27
trade openness indicator (2.47)** (2.55)**
Type I current account reversal × −0.007
capital mobility indicator (0.24)
Type II current account reversal −1.92 −4.12
(1.83)* (7.94)***
Type II current account reversal × −0.02 −0.04
trade openness indicator (0.58) (1.27)
Type II current account reversal × −0.05
capital mobility indicator (1.70)*
Constant −0.28 −0.29 −0.16 −0.16
(2.14)** (2.19)** (1.26) (1.27)
No. of observations 794 793 793 793
No. of countries 41 41 41 41
Adjusted R2 0.38 0.38 0.43 0.43
Source: Author’s regressions.
a. Results obtained from estimating the model in equations 15 and 16 in the text on unbalanced panel data, with the addition of
the variables interacting current account reversals with trade openness and capital mobility indicators. The dependent variable is
the change in the growth rate of GDP per capita (in percentage points). Numbers in parentheses are t statistics (in absolute value);
all regressions include country dummy variables, results for which are not reported. Asterisks indicate statistical significance at
the ***1 percent, **5 percent, and *10 percent levels.
b. Difference between estimated long-run and actual annual growth rates of real GDP, in percentage points.

on the reversal indicators continue to be significantly negative, as they

were in table 11. The coefficient that interacts trade openness with the
presence of a type I reversal is significantly positive in columns 12-1 and
12-2 in table 12. The point estimate in both is 0.27, indicating that exter-
nal adjustment is less costly in countries with higher trade ratios. How-
ever, the coefficient that interacts trade openness and the dummy for
reversals is not significant when the type II reversal indicator is used. The
coefficient that interacts capital account openness and reversal is not sig-
nificant in any of the regressions.
Endogeneity and instrumental variables estimates. The results dis-
cussed above were obtained using a random-effects GLS procedure for
unbalanced panels, and under the assumption that the reversal variable is
exogenous. However, whether a reversal takes place may be affected by
Sebastian Edwards 259

the country’s growth performance, and thus endogenously determined. To

deal with this issue I reestimated equation 16 using an instrumental vari-
ables GLS panel procedure. The following instruments were used: the
one- and two-period-lagged ratio of the current account deficit to GDP; a
lagged sudden stop dummy, which takes the value of 1 if the country
experienced a sudden stop in the previous year; the same regional conta-
gion variable used in the previous analysis; the one-year-lagged ratio of
external gross debt to GDP; the one-year-lagged ratio of net international
reserves to GDP; the one-year-lagged rate of growth of domestic credit;
and the logarithm of initial GDP per capita. As the results in table 13
show, the coefficients on the reversal indicators are significantly negative,
confirming that, historically, current account reversals have had an
adverse effect on growth. The absolute values of the estimated coeffi-
cients, however, are larger than those obtained with the random-effects
GLS procedure (top panel of table 11).
Terms-of-trade effects. The estimation that yielded the results in
table 11 controlled for terms-of-trade changes. That is, the coefficients on
the type I and II reversal variables capture the effect of a current account
reversal with the terms of trade held constant. As discussed above, how-
ever, external adjustment in large countries is very likely to affect the
terms of trade. The exact nature of that effect will depend on a number of
factors, including the relevant elasticities and the extent of home bias in
consumption. To get an idea of the effect of current account reversals
when international prices are allowed to adjust, I reestimated equation 16
excluding the terms-of-trade variable for the sample of large countries.
The full results are not reported here, but the estimated coefficients on
the reversal variables were smaller in absolute terms than those in table
11: −2.43 versus −2.12 in table 11 for type I reversals, and −3.63 versus
−4.13 for type II reversals. These results suggest that, for this sample,
external adjustment has been associated, on average, with an improve-
ment in the international terms of trade.
Robustness tests and other extensions. To test the robustness of the
results, I also estimated several alternative versions of equation 16 for the
sample of large countries. In one of these exercises I introduced lagged
values of the reversal indicators as additional regressors. The results (not
reported here) indicated that lagged values of these indexes were not sig-
nificant at conventional levels. I also varied the definition of “large coun-
tries,” but this likewise did not dramatically affect the results.
 260 Brookings Papers on Economic Activity, 1:2005
Table 13. Impact of Current Account Reversals on Growth in Large Countries:
Instrumental Variables Regressionsa
Independent variable 13-1 13-2
Growth gapb 0.86 0.89
(18.50)*** (20.50)***
Change in terms of trade 0.06 0.11
(3.87)*** (6.86)***
Type I current account reversal −9.40
(4.55)***
Type II current account reversal −12.24
(7.40)***
Constant 0.24 0.38
(1.27) (1.95)*
No. of observations 514 538
No. of countries 34 34
Adjusted R2 0.41 0.40
Source: Author’s regressions.
a. Results obtained from estimating the model in equations 15 and 16 in the text on unbalanced panel data using an instru-
mental variables GLS procedure. The instruments used were the one- and two-period-lagged ratio of the current account deficit
to GDP; a lagged sudden stop dummy equal to 1 if the country experienced a sudden stop in the previous year; the regional sud-
den stop index used in table 10; the one-year-lagged ratio of external gross debt to GDP; the one-year-lagged ratio of net inter-
national reserves to GDP; the one-year-lagged rate of growth of domestic credit; and the logarithm of initial GDP per capita.
Numbers in parentheses are t statistics (in absolute value); all regressions include country dummy variables, results for which
are not reported. Asterisks indicate statistical significance at the ***1 percent, **5 percent, and *10 percent levels.
b. Difference between estimated long-run and actual annual growth rates of real GDP, in percentage points.

Concluding Remarks

The results reported in this paper illustrate the uniqueness of the cur-
rent U.S. external situation. Never before in modern economic history has
a large industrial country run persistent current account deficits of the
magnitude posted by the United States since 2000. This development can
be explained in the context of a portfolio model of the current account,
where, for a number of reasons—the end of the Cold War, the Internet
revolution, and the liberalization of international capital movements in
most countries—foreign investors have increased their (net) demand for
U.S. assets. Indeed, by increasing their holdings of U.S. assets to 30 per-
cent of their wealth, foreigners have provided American residents with
sufficient funds to run the large current account deficits of the last few
years.
The future of the U.S. current account—and thus of the dollar—
depends on whether foreign investors will continue to add U.S. assets to
their investment portfolios. As a way of sharpening the discussion, I have
deliberately made a very optimistic assumption, namely, that during the
Sebastian Edwards 261

next five years foreigners’ net holdings of U.S. assets, as a proportion of

U.S. GDP, will double from their current level. The simulation model
indicates that, even under this optimistic assumption, in the not-too-
distant future the United States is likely to go through a significant exter-
nal adjustment. Indeed, one cannot rule out a scenario in which the U.S.
current account deficit shrinks abruptly by 3 to 6 percent of GDP. Accord-
ing to the simulations, this type of adjustment would imply a cumulative
real depreciation of the trade-weighted dollar in the range of 13 to 23 per-
cent during the first three years of the adjustment.
To obtain an idea of the possible consequences of this type of adjust-
ment, I analyzed the international evidence on current account reversals.
The results of this empirical investigation indicate that major current
account reversals have been associated with large declines in GDP
growth: In large countries, with other factors unchanged, the decline in
growth in GDP per capita has averaged in the range of 2.1 to 4.1 percent-
age points in the first year of the adjustment. Three years after the initial
adjustment, GDP growth is still below its long-run trend.
The results presented in this paper are revealing and suggest that the
United States is likely to experience a major adjustment in the not-too-
distant future. However, many questions are still unresolved and will
require additional research. These include the following:
—How does the behavior of foreign central banks, including their
future demand for U.S. assets, affect the likelihood and magnitude of a
U.S. current account reversal? A particularly important question involves
the appropriate international reserves policy for central banks in a world
where most exchange rates have at least some flexibility. A number of
analysts are concerned that the Asian central banks will reduce their
demand for U.S. assets, unleashing an abrupt collapse in the value of the
dollar.
—How exactly does the adjustment process work in large countries?
Although I have concentrated on a group of countries that I defined as
“large,” in fact all of the countries in my sample that have experienced
current account reversals have much smaller economies than the United
States. In particular, more analysis is needed of the consequences for
global interest rates of a major U.S. current account adjustment.
—How do nominal exchange rates behave in a current account adjust-
ment episode? Most models of the U.S. current account imbalance,
including the portfolio model presented here, have focused on the RER.
 262 Brookings Papers on Economic Activity, 1:2005

However, estimating the way in which the adjustment process will affect
nominal exchange rates is not a trivial matter. The actual adjustment in
nominal exchange rates will depend on pass-through coefficients, as well
as on the exchange rate policies followed by some important U.S. trade
partners, including China, Japan, and other Asian countries.
—Finally, how does private sector saving, and in particular household
saving, behave in the United States? To the extent that household saving
increases—as a result of a decline in home prices, for instance—the cur-
rent account deficit will decline without putting pressure on the value of
the dollar. Likewise, if saving in foreign countries declines, the current
account surplus in the rest of the world will also tend to decline, helping
to achieve global balance. Of course, what matters for current account
behavior is aggregate national saving. Therefore the behavior of public
sector saving is a fundamental variable for future current account and real
exchange rate behavior in the United States and the rest of the world.
APPENDIX A
Table A-1. Comparison of Selected Studies of U.S. Current Account Adjustment and the Dollar, 1999–2005
Author(s) Methodology
Mann (1999) Model tracks U.S. NIIP through
time.
Analyzes trajectory of NIIP under
three scenarios and asks whether
trajectories are sustainable.
Elasticities-based adjustment
mechanism.
Considers two scenarios for global
growth.
Obstfeld and Rogoff (2000) Develops and calibrates optimizing
model of small open economy
with two goods: tradable and
nontradable.
Output is exogenous; prices are
assumed to be flexible; monetary
policy stabilizes price level.
Analyzes effect on RER of exoge-
nous shock that results in reduc-
tion of current account deficit
by 4.4 percent of GDP.
O’Neill and Hatzius (2002) Analyzes trajectory of NIIP as per-
centage of GDP.
Main assumptions
Income elasticity of imports (1.7)
exceeds that of exports (1.0).
Base-case scenario assumes no
RER adjustment for dollar.
Scenario with dollar adjustment
assumes real depreciation of
25 percent.
Structural adjustment scenario
assumes that exports’ elasticity
increases to 1.3.
Elasticity of substitution between
tradables and nontradables is
assumed equal to 1.
Assumes 6 percent annual nominal
interest rate and NIIP equal to
20 percent of GDP.
Tradables output is assumed to be
25 percent of GDP.
Assumes that full employment is
maintained.
Analyzes rates of return obtained
by foreign owners of U.S.
assets.
Main results
In base-case scenario NIIP
becomes increasingly negative
and current account is unsus-
tainable in medium run.
In real depreciation scenario cur-
rent account deficit is within sus-
tainable range even in a 10-year
horizon.
In structural adjustment scenario
current account deficit is 3 per-
cent of GDP in10-year horizon,
if global economy performs well.
In base case, elimination of current
account deficit implies 16 per-
cent real depreciation, and 12
percent nominal depreciation of
the dollar.
Scenario assuming tradables share
of GDP is 15 percent results in
real depreciation of 20 percent.
Effect on nominal value of dollar
could be even higher if reduc-
tion in current account deficit is
very rapid.
Analysis finds it unlikely that U.S.
will continue to attract foreign
buyers for its assets at observed
(continued)
Table A-1. Comparison of Selected Studies of U.S. Current Account Adjustment and the Dollar, 1999–2005 (continued)
Author(s) Methodology
Argues that at the observed levels
of current account deficits, the
NIIP is moving toward the lev-
els of Canada, Australia, and
New Zealand. It is difficult to
believe that this is possible for a
large country such as the U.S.
Estimates “required” RER depre-
ciation in order to bring current
account deficit to 2 percent and
NIIP not to surpass 40 percent.
Wren-Lewis (2004) Calibrates partial equilibrium
model to obtain set of bilateral
RERs consistent with attaining
specified (exogenous) current
account deficits.
No attempt is made to determine
sustainable level of U.S. current
account.
Considers effect of U.S. fiscal
shock and of U.S. technological
shock.
Main assumptions
Argues that, with exception of
FDI, these rates of return have
been modest.
Shows that FDI has declined sig-
nificantly as source of current
account deficit financing.
To determine initial conditions,
estimates “underlying” (or cycli-
cally adjusted) current account
balances.
Considers three possible long-
term scenarios corresponding
to current account deficits of 1,
2, and 3 percent of GDP.
Uses partial equilibrium model of
small economy with three goods
(including nontraded good).
Elasticities and other parameter
values taken from regression
analysis and from OECD data
set.
Main results
low rates of return, and thus its
current account deficit is clearly
unsustainable.
Return to sustainable deficit
(2 percent of GDP) will imply
real depreciation of as much as
43 percent.
Current account deficit of 2 percent
of GDP is consistent with nomi-
nal exchange rates of 88 yen to
dollar and 1.18 dollars to euro.
Under positive technological
shock, “sustainable” current
account deficit may be larger
and consistent with nominal
exchange rates of 89–100 yen to
dollar rate and 1.11–1.19 dollars
to euro.
Estimates that if China has
current account surplus of
1 percent of GDP, nominal
exchange rate would be
6.71 renminbi to dollar.
Benassy-Quere and others Estimates RER path consistent
(2004) with equilibrium in nontradable
goods market.
RER is assumed to depend on
country’s net foreign asset posi-
tion and on relative productivity.
Mussa (2004) Analyzes trajectory of NIIP and
argues that it is unlikely to con-
tinue to grow at current pace. If
it did, it would reach 100 percent
of GDP.
Argues that challenge is for RER
adjustment to be gradual and
not disrupt growth.
Argues that U.S. fiscal adjustment
is necessary for smooth correc-
tion of imbalances.
No attempt is made to calculate
“outer limit” of U.S. NIIP.
Analyzes RER adjustment com-
patible with a gradual reduction
of current account deficit to
2 percent of GDP and NIIP
between 40 and 50 percent.
Model is estimated simultaneously
for fifteen currencies.
Data on net foreign assets obtained
from Lane and Milesi-Ferretti
(2004) and relative productivi-
ties obtained as ratio of con-
sumer to producer price index.
No attempt made to impose exter-
nal equilibrium condition.
Results provided for two cases:
using dollar as numeraire and
using euro as numeraire.
On basis of results from large
econometric models, assumes
that 1-percentage-point reduc-
tion of U.S. current account
deficit is associated with 10 per-
cent real depreciation.
Extent of currency misalignment
depends on how broad is adjust-
ment.
Using dollar as numeraire, esti-
mates that euro was undervalued
by between 1.2 and 7.6 percent
in 2003, and yen by between
14.3 and 22.1 percent in 2001.
Calculates that further real depre-
ciation of 20 percent relative to
mid-2004 values is needed to
achieve long-term current
account deficit of 2 percent of
GDP.
Discusses policies that would
assist adjustment process: fiscal
consolidation in the U.S. to help
keep U.S. demand growth
below pace of output growth,
and more expansive monetary
policy in Europe and Japan.
Concludes that “some” international
policy cooperation is likely to
help the adjustment process.
(continued)
Table A-1. Comparison of Selected Studies of U.S. Current Account Adjustment and the Dollar, 1999–2005 (continued)
Author(s) Methodology
O’Neill and Hatzius (2004) Update of O’Neill and Hatzius
(2002) model.
Analyzes trajectory of NIIP as per-
centage of GDP and finds that
path is not sustainable.
Introduces role of productivity
gains into original framework.
Analyzes composition of capital
flows into U.S.
Incorporates role for valuation
effects.
Obstfeld and Rogoff (2004) Extends Obstfeld-Rogoff (2002)
model to two-country world.
Terms of trade are now endoge-
nous.
Incorporates valuation effects of
exchange rate changes on NIIP.
Assumes elimination of current
account deficit, that is, reduction
equal to 5 percent of GDP.
Main assumptions
Estimates trade balance equation
and uses resulting coefficients
to compute real depreciation
“required” to achieve different
current account adjustment
targets.
Trade equation also includes for-
eign and U.S. demand growth.
Ratio of current account deficit to
tradables is 25 percent; current
account deficit is 5 percent of
GDP.
Output is exogenously given in
both countries.
NIIP is 20 percent of GDP.
Domestic country produces 22
percent of world tradables.
Simulation is done for alternative
values of elasticities and under
different assumptions regarding
changes in tradables output and
military spending.
Main results
Reduction of current account
deficit to 3 percent of GDP
would imply real depreciation
on order of 21.6 to 23.6 percent.
Reduction of current account
deficit to 2 percent would imply
real depreciation on order of
32.1 to 34.1 percent.
Elimination of current account
deficit would imply real depre-
ciation on order of 53 to 55 per-
cent (significantly greater than
estimated by Obstfeld and
Rogoff, 2004).
Assuming constant output, elimi-
nation of current account deficit
implies real depreciation
between 14.7 and 33.6 percent.
If tradables output increases by 20
percent, required real deprecia-
tion ranges from 9.8 to 22.5
percent.
If permanent increase in military
expenditure occurs, required real
depreciation ranges from 16.0 to
36.1 percent.
Roubini and Setser (2004) Uses macro aggregate model to
project U.S. current account.
Imposes exogenous assumptions
on RER and analyzes current
account path.
Blanchard, Giavazzi, Sa Uses portfolio model to analyze
(this volume) U.S. current account behavior.
Assumes changes in portfolio pref-
erences in world economy.
Sources: Literature cited.
First scenario considers a
constant RER for dollar.
Second scenario considers con-
stant trade deficit at 5 percent of
GDP and real depreciation of
approximately 7 percent.
Third scenario considers faster
growth rate of exports and sub-
stantial (50 percent) depreciation
and assumes gradual elimination
(by 2012) of fiscal deficit.
Considers dynamics of
adjustment.
Considers valuation effects of
changes in dollar.
Simulates model under certain
assumptions for values of key
parameters (elasticities, portfolio
shares, and others).
Asks what real depreciation of
dollar is required to eliminate
current account deficit?
In first scenario current account
deficit is 13 percent of GDP in
2012.
In second scenario current account
deficit is 9 percent of GDP in
2012.
In third scenario NIIP stabilizes at
approximately 55 percent of
GDP and current account deficit
declines gradually, reaching
4.3 percent of GDP in 2012.
Estimates range for required real
depreciation (today). After
incorporating valuation effects,
this range is estimated at
between 40 and 90 percent.
 268 Brookings Papers on Economic Activity, 1:2005

Table A-2. Parameter Values for Variables Used in the Simulations

Parameter
Variable value Definition and comments
Portfolio adjustment model
WWorld
Initial $80 trillion World wealth in 2005
WUS
Initial $36 trillion U.S. wealth in 2005
αinitial 0.300 Foreigners’ demand for U.S. assets in (early)
2005a
αjj, initial 0.730 U.S. residents’ demand for U.S. assets in
(early) 2005
αfinal 0.400 Foreigners’ portfolio allocation to U.S. assets
in 2010. An alternative simulation
assumes that, after reaching 0.40, α
declines gradually to 0.365 in 2014.
αjj, final 0.710 U.S. residents’ demand for U.S. assets in
(early) 2010. An alternative simulation
assumes that, after reaching 0.71, αjj rises
to 0.72 in 2014.
αhistorical 0.205 Foreigners’ demand for U.S. assets in (early)
1996. The move to the “initial” current value
of 0.30 is assumed to have been gradual.
αjj, historical 0.800 U.S. residents’ demand for U.S. assets in
(early) 1996
λ 3.0 Wealth-to-GDP ratio
γ*initial 0.290 Value of [αθ − (1 − αjj)]λ in (early) 2005
(see text)
γ*final 0.600 Value of [αθ − (1 − αjj)]λ in 2010
γ*historical 0.150 Value of [αθ − (1 − αjj)]λ in 1996

Transfer problem
g 0.03 Assumed long-term sustainable annual rate of
growth of U.S. GDP
g* 0.03 Annual growth rate of rest-of-world GDP
(including emerging economies as well as
Europe and Japan)
π 0.023 Long-term annual rate of U.S. inflation
π* 0.023 Long-term annual rate of foreign inflation;
some simulations used a value of 0.03.
i 0.043 Long-term real U.S. interest rate; some simula-
tions used a value in the range 0.05 to 0.065.
i* 0.053 Long-term real rest-of-world interest rate;
some simulations used values in the range
0.06 to 0.075.
ηe −1.10 Price elasticity of U.S. imports; this is
slightly below the consensus value; a range
of values was used in other simulations.
(continued)
Sebastian Edwards 269
Table A-2. Parameter Values for Variables Used in the Simulations (continued)
Parameter
Variable value Definition and comments

Real exchange rate elasticity of U.S. exports

εe (approximately the consensus value); sensi-
0.50 tivity analyses used a range of 0.2 to 0.6.
Consensus value for income elasticity of U.S.
ηy imports
1.50 Consensus value for income elasticity of U.S.
εy exports
1.20 Share of imports as a fraction of U.S. GDP in
σm 2004
0.14 Share of exports in U.S. GDP in 2004
σx Rate of change in world price of imports; in
p̂*
m 0.09 alternative simulations a range of −0.05 to
0 −0.10 was used.
Rate of change in world price of exports; in
p̂*
x alternative simulations a range of 0.05 to
0 0.07 was used.
Coefficient measuring rate of adjustment
ψ from actual to demanded asset stock; value
0.30 chosen to obtain best possible fit for
1996–2004.
Coefficient measuring rate of adjustment of
κ absorption to change in income; value cho-
0.20 sen to obtain best possible fit for
1996–2004 period.
Source: Author’s model described in the text.
a. The adjustment period for α and αjj is assumed to be five years.
 270 Brookings Papers on Economic Activity, 1:2005
Table A-3. Variable Definitions and Data Sources
Variable Definition
Type I current account Reduction in an existing cur-
reversal rent account deficit by at
least 6 percent of GDP
over three years.
Type II current account Reduction in an existing cur-
reversal rent account deficit by at
least 4 percent of GDP in
one year
Sudden stop Reduction in net capital
inflows by at least 5 per-
cent of GDP in one year.
The country must have
received an inflow of cap-
ital larger than its region’s
third quartile during the
preceding two years.
Type A currency crisis Dummy variable equal to 1
when an index of external
pressures exceeds its mean
by 3 standard deviations
Type B currency crisis Dummy variable equal to 1
when an index of external
pressures exceeds its mean
by 3 standard deviations
exclusively through
changes in the nominal
exchange rate
Nominal exchange rate Dollars per local currency
unit
Real exchange rate Bilateral real exchange rate
calculated using consumer
price indexes in both
countries
Source
Author’s determination based
on data from World Bank,
World Development Indi-
cators, various years
Author’s determination based
on data from World Bank,
World Development Indi-
cators, various years
Author’s determination based
on data from World Bank,
World Development Indi-
cators, various years
Author’s determination based
on international reserves
and nominal exchange rate
data from International
Monetary Fund, Interna-
tional Financial Statistics,
various years
Author’s determination based
on nominal exchange rate
data from International
Monetary Fund, Interna-
tional Financial Statistics,
various years
International Monetary Fund,
International Financial
Statistics, various years
Author’s calculations using
nominal exchange rate and
consumer price index data
from International Mone-
tary Fund, International
Financial Statistics
(continued )
Sebastian Edwards 271
Table A-3. Variable Definitions and Data Sources (continued)
Variable Definition Source
Change in terms of trade Change in capacity to World Bank, World Develop-
import for a given amount ment Indicators
of exports, in terms of
constant local currency
Ratio of reserves to GDP Net international reserves World Bank, World Develop-
divided by GDP ment Indicators
Domestic credit growth Growth rate of domestic World Bank, World Develop-
credit in percent a year ment Indicators
Ratio of external debt to Total external debt divided World Bank, World Develop-
GDP by GDP ment Indicators
Ratio of fiscal deficit to Overall government budget World Bank, World Develop-
GDP deficit divided by GDP ment Indicators
GDP per capita Real GDP per capita in 1995 World Bank, World Develop-
dollars ment Indicators
Index of capital mobility Index from 0 to 100, with Edwards (forthcoming)
higher values indicating
greater capital mobility
Trade openness indicator Exports plus imports divided World Bank, World Develop-
by GDP ment Indicators
Comments and
Discussion

Kathryn M. E. Dominguez: The U.S. current account deficit at the end

of 2004 reached 5 percent of GDP, a remarkably high number and far
outside the experience of any other large developed country. This paper
by Sebastian Edwards examines the factors that have led to such a large
imbalance, attempts to forecast how long deficits of this magnitude can be
sustained, and analyzes the likely near-term consequences for the U.S.
economy of a reversal of the current account balance.
Current account deficits have been the norm for the United States for
some twenty-five years, just as surpluses have been the norm for many
developing countries and many of the rest of the world’s developed coun-
tries. In theory a deficit, even if persistent, is not necessarily cause for
concern. A country can finance deficits only if the world perceives it to be
a good credit risk. Indeed, the reason many developing countries are forced
to run surpluses is that they lack access to deficit financing. So, if the world
has been willing to finance U.S. current account deficits for over a quarter
of a century, why the concern?
Edwards makes the case that the reason for concern is that the U.S. cur-
rent account deficit is not sustainable: the world will not be willing to con-
tinue to finance U.S. deficits on the current scale into the future. He argues
further that, even if net demand for U.S. assets continues to increase, a
current account reversal is inevitable, which in turn will result in a signif-
icant reduction in U.S. growth. If Edwards’s predictions are accurate, the
implications for the U.S. economy are quite bleak.
I will begin by discussing some of the underlying factors that have led
to the recent large U.S. current account deficits, and which are likely to
influence the future path of global imbalances. Next I will raise some dis-
agreements with some of the assumptions made in the portfolio balance
272
Sebastian Edwards 273

model used by Edwards to simulate U.S. current account dynamics into the
future. I will then discuss whether evidence from smaller countries regard-
ing the links between current account reversals and economic growth is
relevant for the U.S. situation. Finally, I will return to Edwards’s predictions
about the likely downturn for the U.S. economy in light of the evidence from
the data and the model.
Edwards’s figure 1, which graphs the U.S. current account balance and the
dollar real exchange rate since 1973, provides some historical context and
shows the close association of dollar appreciations and U.S. current account
deficits. His table 1 shows that the sources of financing for these deficits
have changed significantly over the past few years. In particular, foreign
direct investment and other equity flows, which were important in the
1990s, have been replaced with net fixed-income flows, consisting largely
of purchases of U.S. Treasury securities by foreign central banks. This shift
in financing has had the favorable consequence (from the point of view of
the U.S. income account) of U.S. investors receiving higher returns on the
foreign assets they hold than foreigners have received on their U.S. assets.
In addition, because the bulk of U.S. liabilities held by foreigners are denom-
inated in dollars, whereas the bulk of foreign assets held by U.S. investors
are denominated in foreign currency, the recent dollar depreciation has
led to positive valuation effects, which, in turn, have improved the U.S.
net international investment position.
An examination of the factors driving the movements in these data is war-
ranted, especially if these factors are expected to persist. In this context it is
interesting to contrast the role of exchange rate policies in the 1980s with that
in the more recent period. In the mid-1980s several foreign countries joined
with the United States in coordinated interventions to bring down the value
of the dollar and correct global imbalances. In the more recent period there
has been no such coordinated attempt on the part of the United States or the
rest of the world to intervene against the dollar. Quite the opposite: central
banks in Asian countries have been intervening to support the value of the
dollar relative to their own currencies, by building dollar reserves to the tune
of about $2 trillion. China alone holds around $610 billion in dollar reserves
and Japan $840 billion. Their purchases of low-return U.S. Treasury securi-
ties have sustained and indeed amplified global imbalances by serving to
both finance the U.S. deficit and maintain a high value of the dollar.
Edwards discusses this role of foreign central banks in sustaining the
U.S. deficit in the context of the United States’ increasing vulnerability to
 274 Brookings Papers on Economic Activity, 1:2005

a change in sentiment toward the dollar on the part of these few key play-
ers. An alternative view is that, if the Asian central banks come to believe
that it is in their best interests to help correct global imbalances, they can
look to the policies followed by central banks in the 1980s to help bring
about an orderly change in the value of the dollar.
Another issue that is little discussed in the paper, but is clearly a driving
force behind the current global imbalances, is differences in economic
growth rates. The U.S. economy in the 1990s sustained the longest expan-
sion in its recorded history. At the same time, Europe and Japan largely
experienced at best lackluster growth. These growth differentials affected
the U.S. current account balance in two ways. First, faster relative U.S.
growth led to higher aggregate demand in the United States for both domes-
tic products and imports, relative to foreign demand for U.S. products.
Second, because the U.S. economy was booming in both absolute and com-
parative terms, global investors were attracted to U.S. assets, which in turn
helped finance the deficit and maintain a strong dollar. Any future changes
in growth differentials are similarly likely to affect the size and sustain-
ability of the U.S. current account deficit.
Edwards’s figure 4 shows U.S. investment and saving rates since 1970.
What is striking is how different today’s saving rates look relative to his-
torical norms. In the past decade and especially in the last few years, the net
household saving rate has been unusually low. This, combined with the
dramatic fall in net public saving in recent years, has reduced total U.S.
saving to its lowest level in a quarter of a century. Investment in recent
years has also been relatively low, although well within the range of his-
torical norms during periods of slow growth. A country’s current account
deficit, of course, equals the difference between saving and investment, and
if we believe that the U.S. saving rate will eventually revert to historical
norms, this provides additional reason for optimism that the current account
deficit will improve.
Edwards introduces a partial equilibrium version of a simple portfolio
balance model of the current account to simulate how potential changes in
the world’s appetite for U.S. assets will influence current account and real
exchange rate adjustment. A number of the simplifying assumptions implicit
in the model are likely to have important implications for the results. In
particular, relative asset allocation shares (the shares of their wealth that
foreign and U.S. investors allocate to foreign and domestic assets) are
assumed to be exogenously determined and subject to home bias. The model
Sebastian Edwards 275

is then used to analyze how an exogenous change in these portfolio allo-

cation preferences will influence the current account. In his base-case
simulation, Edwards assumes that foreign net holdings of U.S. assets rise
to 60 percent of U.S. GDP by 2010 (starting from 30 percent at the end of
2004), that the U.S. economy and the rest of the world economy grow at
the same rate, that the U.S. saving rate remains at its current low level,
that the income elasticity for U.S. imports is higher than for rest-of-world
imports, and that the international terms of trade remain unchanged. The
key result is that, even given the assumed doubling in foreign demand for
U.S. assets, the U.S. current account deficit will eventually decline to a
steady state of 3.2 percent of GDP, with a relatively sharp reversal occur-
ring in 2007, bringing about a 3 percent of GDP reduction in the deficit in
three years. The accumulated real dollar depreciation over this period is
22.5 percent.
The virtue of Edwards’s model is its simplicity. It allows the reader to
easily follow the mechanics of how a change in portfolio preferences influ-
ences all the other key variables (net assets held by foreigners, the trade
balance, the real exchange rate, and the international wealth transfer associ-
ated with these changes). The problem with the model is that its assump-
tions more or less guarantee the main result, namely, that even with an
increase in foreign holdings of U.S. assets, the U.S. current account will
experience a sharp reversal. Edwards briefly discusses how a relaxation of
some of his assumptions would likely influence (and in most cases soften)
his most dire predictions, but the main message remains that a reversal is
inevitable.
The final section of the paper uses cross-country evidence to estimate the
likely costs of a U.S. current account reversal in terms of growth and employ-
ment. The analysis involves examining the incidence of current account
reversals since 1970 for various categories of countries, the correlation of
reversals and sudden stops in capital inflows, the correlation of reversals
and exchange rate changes, and, finally, the correlation of reversals and
GDP growth rates. Edwards is candid about a serious problem with this
analysis: since the United States is unique both because of its size and
because of the role of the dollar in the global economy, it is not clear that
the cross-country evidence marshaled here has any relevance for the United
States. Indeed, Edwards finds no historical examples of large countries
that have experienced a large current account reversal along the lines pre-
dicted by his portfolio balance model.
 276 Brookings Papers on Economic Activity, 1:2005

Yet the statistical information provided is interesting in its own right,

even if its relevance to the U.S. experience is uncertain. The international
evidence suggests that current account reversals, especially when combined
with sudden stops in capital inflows, result in significant reductions in rates
of GDP growth. It is certainly hard to argue that a dramatic fall in the cur-
rent account deficit could occur without major economic dislocation in any
country, including the United States. What is less clear is whether the United
States is likely to experience a dramatic current account reversal or a sudden
stop in capital inflows. Indeed, the fact that many of the factors driving
the large current account deficit are unusual relative to historical (or inter-
national) norms suggests that predicting future dynamics based on recent
experience may not be appropriate.
Does an exit strategy exist for the United States that would allow it to
forestall a sharp current account reversal? This question evokes a whole
series of related questions that seem worth considering before coming to
any conclusions about the future course of the U.S. economy. How might
a rise in U.S. interest rates (which would likely both attract more foreign
investment and decrease the U.S. income account) muddy the waters? Might
higher interest rates, in turn, influence household saving rates? Will the
U.S. government continue to run large budget deficits, and, if so, might
U.S. households turn Ricardian and save more? Might an internationally
coordinated intervention strategy to gradually lower the value of the dollar
work to improve global imbalances? How might an increase in rest-of-world
growth rates (or a fall in U.S. growth rates) influence trade and investment
patterns?
In summary, Edwards has provided a stimulating paper that argues plau-
sibly that the U.S. current account deficit is both unsustainable and likely to
lead to a fall in U.S. economic growth. This conclusion relies on the
assumption that many of the key U.S. and foreign macroeconomic variables
will continue along their course of the past few years. Time will tell
whether or not this assumption is a valid one. Given the rather bleak impli-
cations of Edwards’s analysis, my hope is that the data prove him wrong.

Pierre-Olivier Gourinchas: Sebastian Edwards has written an ambitious

paper on an important topic. The paper starts with a thirty-year perspective
on U.S. current account developments. It introduces and calibrates a port-
folio balance model designed to help in understanding the developments of
recent years. It then uses the model to project the adjustment path for the
Sebastian Edwards 277

dollar’s real exchange rate and the U.S. current account. Finally, it discusses
the growth implications of current account adjustment in light of the expe-
rience of other countries.
I cannot stress enough the importance of the topic: few macroeconomic
questions are as pressing today as that of the sources and the implications
of growing world external imbalances,1 and few issues are receiving as much
attention from academics and policymakers alike. In that respect the paper
stands resolutely on the alarmist side of the current debate, claiming that
the U.S. external position is not sustainable. Even under optimistic assump-
tions about foreigners’ appetite for U.S. assets, the paper finds, “the [U.S.]
current account will have to go through a significant adjustment in the
not-too-distant future.” This adjustment would not be immediate, but it
would be dramatic when it arrives: the benchmark projection finds that the
current account deficit first increases to 7.3 percent by 2009, then experi-
ences an abrupt reversal that brings it down to 3.2 percent by about 2018.
Such a reversal in the current account would be accompanied by a sharp
real depreciation of the dollar of 22.5 percent once the correction begins.
Finally, the paper argues, a current account reversal of this magnitude is
also typically associated with a significant slowdown in economic activity.
READING THE TEA LEAVES. The paper starts with a thorough analysis
of the buildup in U.S. external imbalances. Like other papers before it,2 it
emphasizes two important elements of the current situation. First, the source
of financing of the current account deficit has shifted away from foreign
private investors to foreign central banks, and away from equity and direct
investment to fixed-income vehicles, especially U.S. Treasury securities.
Second, the deficit has been associated in recent years with a decline in U.S.
national saving, especially household and public saving. Since the financ-
ing of the deficit by foreign central banks does not reflect market forces, it
follows that U.S. households and government are artificially living beyond
their means. This interpretation, shared by many commentators, puts the
blame squarely on U.S. domestic factors. The willingness of Asian central
banks, concerned about the value of their currency against the dollar, to
finance the deficits only serves to maintain the gravity-defying properties
of the deficits and of the dollar exchange rate and makes the U.S. position
more vulnerable.

1. See Rajan (2005).

2. For example, Roubini and Setser (2004).
 278 Brookings Papers on Economic Activity, 1:2005

Although this is a compelling story, it is possible to construct a differ-

ent reading of the situation. First, the decline in national saving can be an
equilibrium outcome, without any shift in the U.S. saving schedule. In fact,
if the main impetus behind the current imbalances were a decline in U.S.
saving, those imbalances should be associated with high world real interest
rates. This was the case in the early 1980s in the United States, the period
of so-called twin deficits.
Yet what do we observe? According to the Federal Reserve, the ten-year
yield on Treasury inflation-indexed securities has declined from 2.29 percent
in January 2003 to 1.63 percent in February 2005. An appealing solution
to what has become known as Greenspan’s conundrum was put forth by
Federal Reserve Governor Ben Bernanke.3 He observes that the combina-
tion of large deficits and low real interest rates could be the result of a shift
in the net saving schedule of the rest of the world, through either reduced
foreign investment or increased foreign saving (the saving glut hypothesis).
This shift could be a consequence perhaps of the recent current account
reversals in Asia as well as the growth slowdown in Europe.
It is also instructive to analyze the sources of financing of the U.S. deficits.
As mentioned above, a common interpretation, shared by this paper, is
that the current account is increasingly financed through reserve accumu-
lations. I want to argue that the picture is quite different when one consid-
ers gross flows together with net flows (table 1). Edwards emphasizes that
net foreign direct investment and equity flows for the United States turned
negative in 2003–04. Hence the net financing had to have come from net
reserves and net debt accumulation. Yet consider gross equity and direct
investment liability flows in the table. After a decline between 2000 and
2003, these flows increased again in 2004, from $37.3 billion to $56.2 bil-
lion for equities and from $39.9 billion to $115.5 billion for direct invest-
ment. What this means is that the negative net flows come from the larger
U.S. gross purchases of equity ($93 billion in 2004) and direct invest-
ment ($248.5 billion). The increase in gross U.S. foreign liabilities totaled
$1.4 trillion in 2004. This total capital inflow helped finance a $666 billion
current account deficit and $818 billion in foreign asset acquisition by U.S.
investors (with a $52 billion statistical discrepancy). There is really no mean-
ingful way in which the $358 billion in net reserve accumulation has to have
served to finance the current account deficit. One could just as accurately

3. Bernanke (2005).
Sebastian Edwards 279
Table 1. Net and Gross Financial Flows of the United States
Billions of dollarsa
Item 1998 1999 2000 2001 2002 2003 2004
Reserves, net −26.7 52.3 42.5 23.1 110.3 250.1 358.1
Foreign private purchases of 28.6 44.5 −70.0 −14.4 100.4 113.4 108.1
U.S. Treasuries
Currency 16.6 22.4 5.3 23.8 21.5 16.6 14.8
Securities, netb 32.1 182.6 338.0 309.2 301.4 178.6 323.2
Debt, gross assets 22.8 1.9 15.2 −24.5 −33.5 −28.1 −2.2
Debt, gross liabilities 110.7 185.9 267.4 274.4 229.3 213.7 357.9
Net debt 87.8 184.0 252.2 298.9 262.8 241.8 360.0
Equity, gross assets 101.4 114.3 106.7 109.1 17.6 100.4 93.0
Equity, gross liabilities 45.6 112.9 192.5 119.5 56.2 37.3 56.2
Net equity −55.7 −1.4 85.8 10.4 38.6 −63.2 −36.8
Foreign direct 36.4 64.5 162.1 24.7 −62.4 −133.9 −133.0
investment, net
Assets 142.6 224.9 159.2 142.3 134.8 173.8 248.5
Liabilities 179.0 289.4 321.3 167.0 72.4 39.9 115.5
Claims reported by −15.1 −21.5 31.9 57.6 32.6 55.1 −41.5
nonbanks, net
Claims reported 4.2 −22.0 −31.7 −7.5 66.1 65.2 −15.6
by banks, net
Other −1.2 −2.1 −1.8 −1.6 −0.9 −2.5 −0.2
Net financing 75.0 231.7 476.3 415.0 569.0 542.7 614.0
Memoranda:
Current account balance 209.6 296.8 413.4 385.7 473.9 530.7 665.9
Statistical discrepancy 134.6 65.1 −62.8 −29.3 −95.0 −12.0 51.9
Source: Bureau of Economic Analysis, U.S. International Transactions, table 1.
a. Items may not sum to totals because of rounding.
b. Excluding U.S. Treasury securities.

say that it served to finance the $248.5 billion in foreign direct investment
acquisitions by U.S. firms. In fact, as my figure 1 shows, although the share
of official flows into the United States has recently increased dramatically,
from about zero to 25 percent of foreign-owned assets in the United States,
this follows just as dramatic a collapse, from 25 percent to –5 percent
between 1996 and 1998.
Figure 1 highlights that official purchases represent—except for the few
years of the equity bubble—a relatively stable share of gross liability flows.
As a rough estimate, foreign official assets provide at most only a quarter
of the total capital inflows into the United States.
CURRENT ACCOUNT REVERSALS AND ASSET SUBSTITUTION. The portfolio
balance model developed in the paper predicts large but delayed current
 280 Brookings Papers on Economic Activity, 1:2005

Figure 1. Share of Official Assets in Total Foreign Purchases of U.S. Assetsa

Percent

40
Excluding bank and
30 nonbank claims

20

10 All claims

1992 1994 1996 1998 2000 2002

Source: Bureau of Economic Analysis, International Transactions Accounts.

a. Figure reports official purchases as a share of all foreign claims in the United States and as a share of all foreign claims
excluding bank and non bank claims. Bank and nonbank asset and liability claims roughly offset each other.

account and exchange rate adjustments. I must admit that I was initially
puzzled by the difference between the estimates of this paper and those
by Olivier Blanchard, Francesco Giavazzi, and Filipa Sa in this volume.
After all, both papers use the same model (in Edwards’s paper, equations 1
through 4) and a similar calibration of the model’s parameters. Both mod-
els emphasize the sort of valuation effects that have received considerable
attention recently.4 Yet Blanchard, Giavazzi, and Sa predict a gradual
adjustment of the current account (and a gradual depreciation of the dollar,
cumulating to 54 percent), whereas, as already noted, Edwards predicts that
the current account will keep worsening for the next four years before
reversing sharply between 2009 and 2012.
The answer lies in the specification of the asset demand side of both mod-
els. Whereas Blanchard and his coauthors assume that investors’ demand
for assets depends upon the expected excess return (the expected rate of
dollar depreciation when local currency returns are constant and equal),
Edwards assumes that the portfolio shares are exogenous. This is an extreme

4. See Gourinchas and Rey (2005) and Lane and Milesi-Ferretti (2004b).
Sebastian Edwards 281

assumption: it implies that there is no substitution between domestic and

foreign assets, regardless of their relative returns.
Edwards assumes that the degree of home bias in investment decreases
exogenously between 2004 and 2010. Crucially, it decreases more for foreign
investors (the share of U.S. assets in foreign portfolios increases from 0.3
to 0.4) than for U.S. investors (the share of foreign assets in U.S. portfolio
increases from 0.27 to 0.29). Abstracting from the short-term dynamics
represented by equation 10, this decline in relative home bias is what sus-
tains the increase in current account deficits for the next four years (top left
panel of Edwards’s figure 5) and the continued appreciation of the dollar
over the same period (bottom right panel of figure 5).
But consider what happens when foreigners stop increasing the share
of U.S. assets in their portfolios. At that point the drying up in foreign
financing requires a drastic current account reversal, which can only be
triggered by a real dollar depreciation. Hence the predictions of the cur-
rent account reversal in terms of timing and magnitude come mostly from
the assumptions about the path of relative asset shares. The sharp dollar
depreciation that occurs when the adjustment begins delivers crushingly
low returns to foreigners on their U.S. investment. They lose 22 percent in
three years if local currency returns are equal and constant. One would expect
that, faced with such negative excess returns, investors would demand less
U.S. assets today. But a lower demand for U.S. assets today would pre-
cipitate the adjustment in the exchange rate and the current account. This
indicates that the assumption of exogenous portfolio shares, although use-
ful to calibrate and simplify the model, is eventually too extreme.
Besides the assumptions on portfolio shares, the assumption of a com-
mon growth rate in the United States and the rest of the world matters
also.5 Consider, for instance, what would happen if the rest of the world is
expected to grow faster than the United States. Given stable portfolio shares,
foreigners will want to purchase more U.S. assets as they grow richer. Hence
the supply of foreign capital will not collapse abruptly.

5. This is especially important when looking at the steady state of the model. In the
benchmark calibration, while the trade deficit stabilizes, the dollar keeps depreciating in
real terms. What is happening is a consequence of the Houthakker-Magee paradox. If the
income elasticity of imports exceeds that of exports, the real depreciation must occur along
a balanced growth path with common growth rates at home and abroad and a stable ratio of
the trade balance to GDP. It is possible to have a stable real exchange rate if the rest of the
world grows faster than the United States (Krugman’s 45-degree rule; Krugman, 1989).
 282 Brookings Papers on Economic Activity, 1:2005

CURRENT ACCOUNT REVERSALS AND GROWTH. The last part of the paper
emphasizes that a current account reversal is likely to trigger a growth
slowdown in the United States. This section presents a rich set of empiri-
cal results. This is an area where Edwards has made numerous important
contributions, and I will limit myself to two brief observations. The first is
that his projected real depreciation of the dollar does not allow for the
possibility of a U.S. growth slowdown. Yet it is possible that a decline in
output growth would improve the current account, and as imports drop with
domestic income, the residual decline in the real exchange rate should be
comparatively smaller than in the absence of a slowdown.
More important, although I agree with the general conclusion of this
section—indeed, it is hard to see how a current account reversal of the
size that the paper projects could occur without major disruptions in the
U.S. economy—it is important to keep in mind that the U.S. situation is
unprecedented in many ways besides the size of the external deficit. In
particular, never before in peacetime has the center country of the interna-
tional monetary system accumulated net liabilities on such a scale. It is
reasonable to analyze current account reversals for other economies—most
of which are small relative to gross world product—using the small-country
theoretical apparatus. It is more questionable to do the same for the United
States, as Edwards acknowledges. General equilibrium considerations—
already essential in understanding the source of the current account deficit
in the first place—are crucial in understanding the possible rebalancing
that needs to occur. World asset and good prices cannot be taken as given.
To conclude, it is the great merit of this paper to offer an ambitious
contribution to the debate on the sustainability and ultimate adjustment in
U.S. external imbalances. The paper’s emphasis on the portfolio alloca-
tion problem is also most welcome.

General discussion: Panel members discussed Edwards’s model’s predic-

tion of a sharp depreciation of the dollar leading to current account reversal
and possible crisis. Olivier Blanchard observed that these predictions
result from the absence of valuation effects and zero substitutability between
domestic and foreign assets in the model. With valuation effects, a smaller
depreciation of the dollar would reduce the share of dollar assets in foreign
portfolios to the level consistent with foreign investors’ preferences. And,
with some substitutability between domestic and foreign assets, the depre-
ciation would likely be much more gradual, starting as soon as investors
Sebastian Edwards 283

anticipated the possibility of a change in portfolio preferences. Indeed, it

would already have started. Richard Cooper believed the model’s assump-
tion that foreign investors would target a certain share of U.S. assets in
their portfolios was too rigid. Although the assumption of a preference
shift to more dollar assets can explain the recent accumulation, in the sim-
ulations, when the new, preferred portfolio is achieved, the demand for
U.S. assets falls abruptly, leading to a crisis. Cooper suggested that a
more natural baseline might call for some fraction of rest-of-world saving
to be invested in the United States for an extended period, in which case
the model would predict a gradual slowdown in demand for U.S. assets
and a much softer landing.
Peter Garber observed that Edwards’s baseline model does not explain
how the rest of the world would absorb $500 billion of added saving that
would arise from a reversal in the U.S. current account equal to 4 percent
of U.S. GDP. He conjectured that the added saving would feed back into
foreign investors’ behavior, possibly changing the model’s predictions in
a meaningful way. Gian Maria Milesi-Ferretti said he found it difficult to
judge how far the present, decade-long trend toward greater global diver-
sification of portfolios would go or how it might end. The world may be
transitioning to a new steady state, but we have little guidance about where
that new steady state will be. This implies that the history of current account
reversals is not a good benchmark for the present U.S. situation, but one
cannot rule out the possibility that some shock will lead to a crisis and
return investors’ behavior to what it was in the 1990s.
Barry Eichengreen questioned the robustness of the paper’s empirical
analysis, comparing it to a similar recent analysis by Hilary Croke, Steven
Kamin, and Sylvain Leduc. Their sample includes more industrialized coun-
tries and a less restrictive definition of reversal, which leads them to pre-
dict sizable current account reversals in some large countries, including
France, the United Kingdom, and the United States. Unlike Edwards, how-
ever, they do not find large negative output effects associated with reversals,
possibly indicating that large countries are able to adjust more smoothly.
Sebastian Edwards replied that the differences between his results and
those of Croke, Kamin, and Leduc arise from the very mild episodes of
exchange rate adjustment that qualify as reversals in their analysis. He found
it unsurprising that such gradual adjustments are not disruptive.
Hélène Rey professed puzzlement at Edwards’s finding that a sudden stop
or reversal of capital inflows has more pronounced effects the higher the
 284 Brookings Papers on Economic Activity, 1:2005

degree of trade openness. Theory, as embodied in the model by Obstfeld and

Rogoff in this volume, holds that a more closed economy requires a larger
relative price adjustment, which suggests that the disruption to the economy
will be more substantial. Believing there might be an omitted-variable bias
in Edwards’s results, she suggested including the degree of liability dollar-
ization, which has been a robust explanatory variable in the sudden stop
literature, as an independent variable. Edwards agreed that his results were
surprising, but he added that, in the sudden stop literature, liability dollar-
ization plays no role in the impact of a sudden stop or reversal, but only
on the probability of their happening. Milesi-Ferretti conjectured that the
reason the empirical analysis found current account reversals not to be
associated with large exchange rate swings is that changes in the terms of
trade are often associated with current account reversals. As examples he
noted that some countries in the sample, such as Norway and New Zealand,
have highly volatile terms of trade. If the price of their exports doubles, their
current account balances improve without a weakening of their currencies.
Several panelists stressed the importance of valuation effects in pre-
venting the U.S. net debt position from increasing substantially in recent
years. Cooper pointed out that, in 2003, the U.S. current account deficit
reached $530 billion, yet the U.S. net asset position declined by only $98
billion. The valuation effects that account for this difference are due both
to exchange rate effects and to differing returns on foreign and domestic
gross portfolios. Richard Portes observed that the depreciation that con-
tributed to these valuation effects cannot go on forever. Milesi-Ferretti added
that, to the extent valuation effects from differences in asset returns hold
up the dollar, the trade deficit will grow faster, requiring a larger depre-
ciation eventually.
Sebastian Edwards 285

References
Ades, Alberto, and Federico Kaune. 1997. “A New Measure of Current Account
Sustainability for Developing Countries.” New York: Goldman Sachs Emerg-
ing Markets Economic Research.
Barro, Robert J., and Xavier Sala-i-Martin. 1995. Economic Growth. New York:
McGraw-Hill.
Benassy-Quere, Agnes, Pascale Duran-Vigneron, Amina Lahreche-Revil, and
Valerie Mignon. 2004. “Burden Sharing and Exchange Rate Misalignments
within the Group of Twenty.” In Dollar Adjustment: How Far? Against What?
edited by C. Fred Bergsten and John Williamson. Washington: Institute for
International Economics.
Bergsten, C. Fred, and John Williamson, eds. 2003. Dollar Overvaluation and the
World Economy. Special Report 16. Washington: Institute for International
Economics (February).
________. 2004. Dollar Adjustment: How Far? Against What? Washington:
Institute for International Economics.
Bernanke, Ben S. 2005. “The Global Saving Glut and the U.S. Current Account
Deficit.” The Sandridge Lecture, Virginia Association of Economics, Rich-
mond, March 10. Available at www.federalreserve.gov/boarddocs/speeches/
2005/200503102.
Blanchard, Olivier, Francesco Giavazzi, and Filipa Sa. 2005. “The U.S. Current
Account and the Dollar.” Working Paper 11137. Cambridge, Mass.: National
Bureau of Economic Research (February).
Caballero, Ricardo, Emmanuel Farhi, and Mohamad L. Hammour. 2004. “Specula-
tive Growth: Hints from the U.S. Economy.” Working Paper 10518. Cambridge,
Mass.: National Bureau of Economic Research (May).
Calvo, Guillermo A., Alejandro Izquierdo, and Luis-Fernando Mejia. 2004. “On
the Empirics of Sudden Stops: The Relevance of Balance-Sheet Effects.”
Working Paper 10520. Cambridge, Mass.: National Bureau of Economic
Research (May).
Choi, Chi-Young, Nelson Mark, and Donggyu Sul. 2004. “Unbiased Estimation
of the Half-Life to PPP Convergence in Panel Data.” Working Paper 10614.
Cambridge, Mass.: National Bureau of Economic Research (July).
Cooper, Richard. 2004. “America’s Current Account Deficit Is Not Only Sustain-
able, It Is Perfectly Logical Given the World’s Hunger for Investment Returns
and Dollar Reserves.” Financial Times, November 1.
Corden, W. Max. 1994. Economic Policy, Exchange Rates, and the International
System. University of Chicago Press.
Croke, Hilary, Steven B. Kamin, and Sylvain Leduc. 2005. “Financial Market
Developments and Economic Activity during Current Account Adjustments in
Industrial Economies.” International Finance Discussion Paper 827. Washington:
Board of Governors of the Federal Reserve System (February).
 286 Brookings Papers on Economic Activity, 1:2005

Dollar, David. 1992. “Outward-Oriented Developing Economies Really Do Grow

More Rapidly: Evidence from 95 LDCs, 1976–1985.” Economic Development
and Cultural Change 40, no. 3: 523– 44.
Dooley, Michael P., David Folkerts-Landau, and Peter Garber. 2004a. “The
Revived Bretton Woods System: The Effects of Periphery Intervention and
Reserve Management on Interest Rates and Exchange Rates in Center Coun-
tries.” Working Paper 10332. Cambridge, Mass.: National Bureau of Economic
Research (March).
________. 2004b. “Direct Investment, Rising Real Wages and the Absorption of
Excess Labor in the Periphery.” Working Paper 10626. Cambridge, Mass.:
National Bureau of Economic Research (July).
Edwards, Sebastian. 1995. Crisis and Reform in Latin America: From Despair to
Hope. Oxford University Press for the World Bank.
________. 1999. “Crisis Prevention: Lessons from Mexico and East Asia.” Working
Paper 7233. Cambridge, Mass.: National Bureau of Economic Research (July).
________. 2002. “Does the Current Account Matter?” In Preventing Currency
Crises in Emerging Markets, edited by Sebastian Edwards and Jeffrey A.
Frankel. University of Chicago Press.
________. 2003. “Debt Relief and the Current Account: An Analysis of the HIPC
Initiative.” World Economy 26, no. 4: 513–31.
________. 2004. “Thirty Years of Current Account Imbalances, Current Account
Reversals and Sudden Stops.” International Monetary Fund Staff Papers 51 (spe-
cial issue): 1–49 (January).
________. Forthcoming. “Capital Controls, Sudden Stops and Current Account
Reversals.” In International Capital Flows, edited by Sebastian Edwards. Uni-
versity of Chicago Press.
Edwards, Sebastian, and Eduardo Levy-Yeyati. Forthcoming. “Flexible Exchange
Rates as Shock Absorbers.” European Economic Review.
Eichengreen, Barry, Andrew K. Rose, and Charles Wyplosz. 1996. “Contagious
Currency Crises.” Working Paper 5681. Cambridge, Mass.: National Bureau of
Economic Research (July).
Frankel, Jeffrey A., and Eduardo A. Cavallo. 2004. “Does Openness to Trade
Make Countries More Vulnerable to Sudden Stops, or Less? Using Gravity to
Establish Causality.” Working Paper 10957. Cambridge, Mass.: National Bureau
of Economic Research (December).
Frankel, Jeffrey A., and Andrew K. Rose. 1996. “Currency Crashes in Emerging
Markets: An Empirical Treatment.” Journal of International Economics 41, no.
3–4: 351–66.
Freund, Caroline L. 2000. “Current Account Adjustment in Industrialized Countries.”
International Finance Discussion Paper 692. Washington: Board of Governors
of the Federal Reserve System (December).
Sebastian Edwards 287

Gourinchas, Pierre-Olivier, and Hélène Rey. 2005. “International Financial

Adjustment.” Working Paper 11155. Cambridge, Mass.: National Bureau of
Economic Research (February).
Guidotti, Pablo E., Agustin Villar, and Federico Sturzenegger. 2003. “Aftermaths
of Current Account Reversals: Export Growth or Import Compression.” Paper
presented at the Eighth LACEA Meeting, Puebla, Mexico, October.
Hooper, Peter, Karen Johnson, and Jaime Marquez. 2000. “Trade Elasticities for
G-7 Countries.” Princeton Studies in International Economics 87. Princeton
University.
Kraay, Aart, and Jaume Ventura. 2002. “Current Accounts in the Long and Short
Run.” Working Paper 9030. Cambridge, Mass.: National Bureau of Economic
Research (June).
Krugman, Paul R. 1989. “Differences in Income Elasticities and Trends in Real
Exchange Rates.” Working Paper 2761. Cambridge, Mass.: National Bureau of
Economic Research.
Lane, Philip R., and Gian Maria Milesi-Ferretti. 2001. “The External Wealth of
Nations: Measures of Foreign Assets and Liabilities for Industrial and Developing
Countries.” Journal of International Economics 55, no. 2: 263–94.
________. 2004a. “International Investment Patterns.” Discussion Paper 4499.
London: Centre for Economic Policy Research (July).
________. 2004b. “Financial Globalization and Exchange Rates.” CEPR Discus-
sion Paper 4745. London: Centre for Economic Policy Research (November).
Mann, Catherine L. 1999. Is the U.S. Trade Deficit Sustainable? Washington:
Institute for International Economics.
________. 2003. “How Long the Strong Dollar?” In Dollar Overvaluation and
the World Economy, edited by C. Fred Bergsten and John Williamson. Special
Report 16. Washington: Institute for International Economics (February).
________. 2004. “The U.S. Current Account, New Economy Services, and Impli-
cations for Sustainability.” Review of International Economics 12, no. 2 (special
issue): 262–76 (May).
Milesi-Ferretti, Gian Maria, and Assaf Razin. 2000. “Current Account Reversals
and Currency Crises: Empirical Regularities.” In Currency Crises, edited by
Paul Krugman. University of Chicago Press.
Mussa, Michael. 2004. “Exchange Rate Adjustments Needed to Reduce Global
Payments Imbalance.” In Dollar Adjustment: How Far? Against What? edited by
C. Fred Bergsten and John Williamson. Washington: Institute for International
Economics.
Obstfeld, Maurice, and Kenneth Rogoff. 2000. “Perspectives on OECD Eco-
nomic Integration: Implications for U.S. Current Account Adjustment.” In
Global Economic Integration: Opportunities and Challenges. Kansas City,
Mo.: Federal Reserve Bank of Kansas City (August).
 288 Brookings Papers on Economic Activity, 1:2005

________. 2004. “The Unsustainable U.S. Current Account Position Revisited.”

Working Paper 10869. Cambridge, Mass.: National Bureau of Economic
Research (November).
O’Neill, Jim. 2003. “Features of a Dollar Decline.” In Dollar Overvaluation and
the World Economy, edited by C. Fred Bergsten and John Williamson. Special
Report 16. Washington: Institute for International Economics (February).
O’Neill, Jim, and Jan Hatzius. 2002. “US Balance of Payments: Still Unsustain-
able.” Global Economics Paper 70. New York: Goldman Sachs (March).
________. 2004. “US Balance of Payments: Unsustainable, But. . . .” Global
Economics Paper 104. New York: Goldman Sachs (March).
Rajan, Raghuram G. 2005. “Global Current Account Imbalances: Hard Landing
or Soft Landing.” Speech at the Credit Suisse First Boston Conference, March
15. Available at www.imf.org/external/np/speeches/2005/031505.htm.
Roubini, Nouriel, and Brad Setser. 2004. “The U.S. as a Net Debtor: The Sustain-
ability of the U.S. External Imbalances.” Stern School of Business, New York
University, and University College, Oxford (August).
Sachs, Jeffrey D., and Andrew Warner. 1995. “Economic Reform and the Process
of Global Integration.” BPEA, no. 1: 1–95.
Taylor, Alan M. 2002. “A Century of Current Account Dynamics.” Working Paper
8927. Cambridge, Mass.: National Bureau of Economic Research (May).
Tille, Cedric. 2003. “The Impact of Exchange Rate Movements on US Foreign
Debt.” Current Issues in Economics and Finance 9, no. 1: 1–7 (January).
Wren-Lewis, Simon. 2004. “The Needed Changes in Bilateral Exchange Rates.”
In Dollar Adjustment: How Far? Against What? edited by C. Fred Bergsten
and John Williamson. Washington: Institute for International Economics.
DEAN BAKER
Center for Economic and Policy Research
J. BRADFORD DELONG
University of California, Berkeley
PAUL R. KRUGMAN
Princeton University

Asset Returns and Economic Growth

IT is difficult to see how real U.S. GDP growth can be as rapid in the
next half-century as it has been in the last. The baby boom is long past,
and no similar explosion of fertility to boost the rate of labor force growth
from natural increase has occurred since or is on the horizon. The modern
feminist revolution is two generations old: no reservoir of potential female
labor remains to be added to the paid labor force. Immigration will doubt-
less continue—the United States is likely still to have only one-twentieth
of the world’s population late in this century and to remain vastly richer
than the world on average—but can immigration proceed rapidly enough
to make the labor force grow as fast in the next fifty years as it did in the
past fifty? Productivity growth, the other possible source of faster GDP
growth, is a wild card: although we find very attractive the arguments of
Robert Gordon for rapid future productivity growth,1 his is not the con-
sensus view; this is shown most strikingly by the pessimistic projection of
the Social Security trustees that very long run labor productivity growth
will average 1.6 percent a year.2
A slowing of the rate of real economic growth raises challenges for the
financing of pay-as-you-go social insurance systems that rely on a rapidly
expanding economy to provide generous benefits for the elderly at relatively
low tax rates on the young. An alternative way of financing such systems

1. Gordon (2003). Oliner and Sichel (2003) and Kremer (1993) provide additional rea-
sons to be very optimistic about future productivity growth.
2. Board of Trustees of the Federal Old Age and Survivors Insurance and Disability
Insurance Trust Funds (2005; all citations from this report are for the intermediate projec-
tion). Contrast this with the 2.0 percent average annual rate of economy-wide labor pro-
ductivity growth from the fourth quarter of 1989 through the first quarter of 2005.

289
 290 Brookings Papers on Economic Activity, 1:2005

is to prefund them, and for that reason projections of future rates of return
on capital play an important role in today’s economic policy debates. The
solutions to many policy issues depend heavily on whether historical real
rates of return—especially the 6.5 percent or so annual average realized rate
of return on equities—are likely to persist: the higher are likely future rates
of return, the more attractive become policies that, at the margin, shift some
additional portion of the burden of financing social insurance onto the
present and the near future, thus giving workers’ contributions the power
to compound over time.
We believe that the argument for prefunding—that slowing economic
growth creates a presumption that the burden of financing social insur-
ance should be shifted back in time toward the present—is much shakier
than many economists recognize.3 It is our belief that if forecasts of slower
real GDP growth come to pass, then it is highly likely that future real
returns to capital will likewise be significantly below past historical aver-
ages. In our view the links between asset returns and economic growth are
strong: the algebra of capital accumulation and the production function
and the standard macrobehavioral analytical models that economists use
as their finger exercises suggest this; arithmetic suggests this as well, for
we cannot see any easy way to reconcile current real bond, stock dividend,
and stock earnings yields with the twin assumptions that asset markets are
making rational forecasts and that rationally expected real rates of return
will be as high in the future as they have been in the past half-century.
Our basic argument is very simple. Consider a simple chart of the sup-
ply and demand for capital in generational perspective (figure 1). The
supply of capital—the amount of investable assets accumulated by savers—
presumably follows a standard (if probably steeply sloped) supply curve,4
with relative quantities of total saving and thus of capital plotted on the
horizontal axis, and the price of capital—that is, its rate of return—on the
vertical axis. The demand for capital by businesses will, of course, depend

3. An argument challenged, for reasons similar to but not exactly aligned with those we
discuss here, in Cutler and others (1990).
4. Supply is likely to be steeply sloped because of opposing income and substitution
effects. An increase in the rate of return increases the total lifetime wealth of savers, which
presumably increases their consumption when young and so diminishes their saving. An
increase in the rate of return also increases the incentive to save, which presumably increases
saving. The net effect—which we believe to be positive—is likely to be relatively small.
Dean Baker, J. Bradford DeLong, and Paul R. Krugman 291
Figure 1. The Supply and Demand of Capital and the Rate of Return

Rate of return

Supply of saving

Demand under
rapid economic
growth

Demand under slow

economic growth

Capital

on the rate of return demanded by the savers who commit their capital to
businesses: the higher this required rate of return, the lower will be busi-
ness demand for capital—and the more eager will businesses be to substi-
tute labor for capital in production. The demand for capital by businesses
depends on many other factors as well, from which we single out two:
—The rate of growth of the labor force. Labor and capital are comple-
ments. A larger labor force for firms to hire from will raise the marginal
product of capital for any given level of the capital stock, making businesses
more willing to pay higher returns in order to get hold of capital.
—The rate of improvement in the economy’s level of technology. Better
technology—also a complement to capital—will boost business demand.
What is the effect of a slowdown in economic growth—through either a
fall in the rate at which the labor force grows, or a fall in the rate at which
technology and thus equilibrium labor productivity increase—on this equi-
librium? Assume that these changes do not affect the saving behavior of
 292 Brookings Papers on Economic Activity, 1:2005

the accumulating generation:5 then they affect only the demand curve and
not the supply curve. Each of these shocks moves the demand curve left-
ward: having fewer workers reduces the marginal product of capital and
hence firm demand for capital; slower productivity growth does the same.
The equilibrium capital stock falls, and the rate of return that savers can
demand, while still finding businesses willing to invest what they have
saved, falls as well. Slower economic growth brings with it lower real rates
of return.
We make our case as follows. After first laying out what we see as the
major issues to be resolved, we discuss how the algebra of the production
function and capital accumulation suggests that rates of return and rates of
growth are strongly linked. We then analyze the standard, very simple,
macrobehavioral models that economists use to address these issues and
find that they, too, lead us to not be surprised by a strong positive rela-
tionship between economic growth and asset returns. We then turn to the
arithmetic: starting from current bond, stock dividend, and stock earnings
yields, we find it arithmetically very difficult to construct scenarios in
which asset returns remain at their historic average values when real GDP
growth is markedly slowed.
Next we turn to what we regard as the most interesting possibility for
escape from this bind. In the late nineteenth century, slower growth in the
British economy was accompanied by no reduction in returns on British
assets, as Britain exported capital on a scale relative to the size of its
economy never seen before or since. Could the United States follow the
same trajectory? Yes. Is it likely to? Not without a huge boost to national
saving.
Before concluding, we turn to a brief analysis of the equity premium.
Much argument and some analysis of the dilemmas of the U.S. social insur-
ance system point to the large historical value of the equity premium in
America as a potential source of excess returns. We argue, however, that
once one has conditioned on the level of the capital-output ratio, returns on

5. As Gregory Mankiw points out in his comment on this paper, and as we discuss below,
in the standard Ramsey model a reduction in the rate of natural increase does affect the sav-
ing of the accumulating generation—and shifts the saving supply curve inward exactly as
much as investment demand shifts inward, keeping the real rate of return unchanged. This is
due to the powerful bequest motive behind the assumption of an infinitely lived representative
household whose utility for a given level of consumption per capita is linear in the size of the
household.
Dean Baker, J. Bradford DeLong, and Paul R. Krugman 293

balanced portfolios in the long run depend only on the physical return to
capital and the margins charged by financial intermediaries. (However,
attitudes toward risk do affect the long-run capital-output ratio.) They do
not depend on the equity premium or the price of risk.
We conclude that if economic growth over the next century falls as far
as envisioned by forecasts like those in the 2005 Social Security trustees’
report, then it is not very likely that asset returns will match historical
experience. If the stock market today is significantly overvalued and about
to come back to earth, if the distribution of income undergoes a signifi-
cant shift away from labor and toward capital, or if the United States mas-
sively boosts its national saving rate and runs surpluses on the relative scale
of pre–World War I Britain, for more than twice as long as Britain did—
then a real GDP growth slowdown need not entail a significant reduction
in asset returns. But these seem to us to be possible, not probable, scenarios,
and not the central tendency of the distribution of possible futures that is a
real economic forecast.

Issues

The United States is in all likelihood undergoing a minor demographic

transition: from a twentieth century in which the population’s rate of nat-
ural increase was high, to a twenty-first century in which, many suspect,
fertility will be at or below levels consistent with zero population growth.
This will translate into a slowdown in growth in labor input. From 1958 to
2004, total hours worked in the economy grew at 1.5 percent a year as the
entrance of the baby-boomers—male and female—and their successors
into the labor force vastly outweighed a decline in average hours worked.
The Social Security Administration’s 2005 trustees’ report projects that
hours worked will grow at only 0.3 percent a year from 2015 through 2045.6
Meanwhile some economists—although far from all—are projecting a
slowdown in productivity growth.7 The Social Security Administration
foresees economy-wide labor productivity growing at only 1.6 percent a
year in 2011 and thereafter. In contrast, between 1995 and 2004 economy-
wide labor productivity grew at 2.5 percent a year, between 1990 and 2004

6. Board of Trustees (2005).

7. See Oliner and Sichel (2003); Gordon (2003); Nordhaus (2005).
 294 Brookings Papers on Economic Activity, 1:2005

it grew at 2.0 percent a year, and between 1958 and 2004 at 1.9 percent a
year.8 Thus, less than a decade from now, the Social Security forecasters
at least see a significant change in both key factors in economic growth: a
fall of 1.2 percentage points a year in the rate of growth of labor input,
and a fall of between 0.3 and 0.9 percentage point, depending on whether
one takes the long 1958–2004 or the short 1995–2004 baseline, in labor
productivity growth. The total growth slowdown forecast to hit in a decade
or less is thus in the range of 1.6 to 2.2 annual percentage points of real
GDP.
What implications will this growth slowdown—if it comes to pass—
have for asset values and returns? One position, taken implicitly by the
Social Security Administration and explicitly by others,9 is that there is no
reason to expect asset returns to be lower in the future. Whereas U.S. eco-
nomic growth is determined by productivity growth and labor force growth
in the United States, U.S. asset returns are determined by time preference,
the intertemporal elasticity of substitution in consumption, and attitudes
toward risk, all in a global economy. Why should they be connected? Thus,
we hear, past asset performance is still the best guide to future returns.
We take a contrary position. Yes, safe asset returns are equal to the
marginal utility of saving, stock market returns equal safe asset returns plus
the cost of bearing equity risk, and the United States is part of a world
economy. Yes, economic growth is equal to productivity growth plus
labor force growth. But only in the case of a small open economy with
fixed exchange rates are asset returns determined independently of the rate
of economic growth. In a large open economy, they are jointly determined
and will be linked.10

8. An alternative breakdown would distinguish 1958–73, during which economy-wide

labor productivity growth averaged 2.6 percent a year; the productivity slowdown period of
1973–95, when it averaged 1.2 percent a year; and the post-1995 “new economy” period,
when it averaged 2.6 percent a year. Much depends on whether one interprets the 1973–95
productivity slowdown period as an anomalous freak disturbance to the economy’s normal
structure, or as just one of those things one can expect to see every half-century or so.
9. Council of Economic Advisers (2005).
10. Even in a small open economy, real returns on assets and rates of economic growth
will be linked unless the real exchange rate is fixed. Even perfect arbitrage by mammoth
amounts of risk-neutral foreign capital only equalizes expected rates of return at home and
abroad calculated in foreign currency. With a flexibly changing real exchange rate, the rate
of return in foreign currency is not the same as the rate of return in domestic currency.
Dean Baker, J. Bradford DeLong, and Paul R. Krugman 295

Perhaps an analogy will be helpful. In international trade, the trade bal-

ance is the difference between what exporters are able to sell abroad and
home demand for imports. In international finance, the trade balance is
the difference between national saving and national investment. How can
this be? Why should a change in exporters’ success at marketing abroad
change either national saving or national investment? Great confusion has
been caused throughout international economics over how, exactly, to
think of the connection. We believe that claims that national economic
growth is unconnected with asset returns reflect a similar failure to grasp
the whole problem.
This is an important issue to get straight now, because the relative attrac-
tiveness of pay-as-you-go versus prefunded social insurance systems depends
to some degree on the gap between the return on capital r and the rate of
real economic growth n + g, the sum of the rate of growth of employment
n and the rate of growth of labor productivity g. If we are willing to be
simple Benthamites, with a social welfare function that shamelessly makes
interpersonal comparisons of utility, the argument is straightforward. The
higher is the rate of economic growth n + g relative to the return on capital
r, the more attractive do pay-as-you-go social insurance systems become.
When n + g approaches r, pay-as-you-go systems appear to be very cheap
and effective ways of increasing social welfare by passing resources down
from the rich and numerous future to the poorer and less numerous present.
By contrast, the larger is r relative to n + g, the greater are the benefits of
prefunding social insurance systems. Prefunded systems can use high rates
of return and compound interest to reduce the wedge between productivity
and after-contribution real wages. They thus sacrifice the possibility of
raising social welfare by moving wealth from the richer distant future to
the near future and the present, but in return they gain by reducing the social
insurance tax rate and thus its deadweight loss. And whenever we make
utilitarian arguments other than those of pure Pareto-preference for why
one set of policies is superior to another set, we are all, in our hearts, secret
Benthamites.
Thus, to the extent that the political debate over the future of social
insurance in America is conducted in the language of rational policy
analysis, getting the gap between r on the one hand and n + g on the other
hand right is important. Policies predicated on a false belief that r is much
larger relative to n + g than it is will unduly burden today’s and tomor-
row’s young people and will leave many disappointed when returns on
 296 Brookings Papers on Economic Activity, 1:2005

assets turn out to be less than anticipated and prefunding leaves large
unexpected holes in retirement financing. Policies predicated on the belief
that n + g is higher relative to r than it is pass up opportunities to lighten
the overall tax burden and still provide near-equivalent income security
benefits in the long run.

Algebra

Let us begin by distinguishing a number of different rates of return. In

this paper we use r to stand for a physical gross marginal product of capi-
tal, and we assume that it is the product of a Cobb-Douglas production
function:
Y
(1) r =α .
K
We distinguish this physical capital rate of gross profit r from the net rate
of return on a balanced financial portfolio rf and from the net rate of return
on equities re.
Only under the assumptions of constant depreciation rates δ, constant
financial markups, and a constant price and amount of risk is the map-
ping among these three straightforward. Toward the end of this paper we
briefly consider the equity premium, but otherwise we assume that depre-
ciation rates, financial markups, and other factors that could vary the wedges
between r, rf , and re are unimportant. Thus we will move back and forth
between these three different rates of return: things that raise or lower the
return on stocks will also raise or lower the return on bonds and (after the
capital stock has adjusted) the physical marginal product of capital as well.
Robert Solow studied a constant-returns Cobb-Douglas production func-
tion with α as the returns-to-capital parameter, and with Y, K, L, and E as
aggregate output, the capital stock, the supply of labor, and the level of
labor-augmenting technology, respectively:11
(2) Y = K α ( EL )1− α .

11. Solow (1956).

Dean Baker, J. Bradford DeLong, and Paul R. Krugman 297

Assume constant rates of labor force growth n, of labor-augmenting

technical change g, of depreciation δ, and of gross saving s. In the closed-
economy case, in which all of domestic capital K is owned by domestic
residents and in which all of national saving goes into increasing the
domestic capital stock, we know that, along a steady-state growth path of
the economy,
K s
(3) = .
Y n+g+δ
This tells us that, along such a growth path,

 n + g + δ
(4) r = α  .
 s
If permanent shocks that reduce n + g cause the economy to transit from
one steady-state growth path to another, the rate of return on capital falls,
with the change in r being

(5) ∆r = ( α s ) ( ∆n + ∆g ).

As long as α is greater than or equal to s—that is, as long as the economy is

not dynamically inefficient12—the reduction in r will be greater than one
for one. From this algebra we would expect the roughly 1.5-percentage-
point reduction in the rate of real GDP growth forecast by the Social Secu-
rity Administration to carry with it a greater than 1.5-percentage-point
reduction in r.
These are steady-state results. How relevant are they for, say, the seventy-
five-year standard forecast horizon used in analyses of the Social Security
system? In the Solow model the capital-output ratio approaches its steady-
state value at an exponential rate of −(1 − α)(n + g + δ), which, at histori-
cal values, is roughly 3.6 percent a year. That closes half the gap to the
steady-state capital-output ratio in twenty years. After seventy-five years
the capital-output ratio has closed 93 percent of the gap between its initial
and its steady-state value.
In this simple Solow setup, only three things can operate to prevent a per-
manent downward shock to n + g from reducing r. Perhaps the depreciation

12. We have every reason to believe that the economy is dynamically efficient, in that
capital in the steady state exceeds the “golden rule” level. See Abel and others (1989).
 298 Brookings Papers on Economic Activity, 1:2005

rate δ could fall. We have been unable to think of a coherent reason why a
reduction in labor force growth n or labor productivity growth g should
independently carry with it a reduction in δ. (However, the reduction in r
could plausibly carry with it an extension of the economic lives of equip-
ment and buildings, and so bring about a partly offsetting fall in δ that
would moderate the decline in r.) Or perhaps the production function
could shift to increase the capital share of income α.
Last, perhaps a permanent downward shock to n + g could also bring
about a reduction in the saving rate s. If it were the case that

 
(6) ds = − 
s
( dn + dg ) ,
 n + g + δ 
then the rate of return r would be constant. There is a reason to think that
a fall in n would carry with it a reduction in s: an economy with slower
labor force growth is an aging economy with relatively fewer young peo-
ple and, presumably, if the young do the bulk of the saving, a lower sav-
ing rate. (A decline in g, however, would tend to work the other way: the
income effect would tend to raise s.)

Analysis

Are the effects just discussed plausibly large enough to keep the rate of
return on capital constant at the rate of economic growth? To assess that,
we need to model saving decisions, which requires moving from algebra
to model-based analysis.

The Ramsey Model

We move now from Solow to Ramsey-Cass-Koopmans.13 Consider a
version of this Ramsey model in which the representative household has
the following utility function:

∑ (1 + β ) (U (C ))N
∞ −t
1− λ
(7) t t ,
t=0

13. See Romer (2000).

Dean Baker, J. Bradford DeLong, and Paul R. Krugman 299

where β is the pure rate of time preference, Ct is consumption per house-

hold member, and Nt is the number of members of the representative house-
hold, growing according to

(8) N t +1 = (1 + n ) N t ,

where n now measures growth in the size of the household. In the stan-
dard Ramsey model setup as presented by David Romer,14 the parameter
λ equals zero, so that the household utility function becomes

∑ (1 + β ) (U (C ))N .
∞ −t
(9) t t
t=0

This choice drives the result that changes in labor force growth do not
have long-run effects on steady-state capital-output ratios or rates of return.
But, to us at least, this assumption seems artificial. If it is indeed the case
that the utility function is that specified in equation 9, then the more mem-
bers of the household, the merrier: household utility is linear in the number
of people in the household but suffers diminishing returns in consumption
per capita. A household with this utility function, provided it has control
over its own fertility, would choose to grow as rapidly as possible; that
would be the way to make individual units of consumption contribute as
much as possible to total household utility. It seems reasonable to allow λ to
be greater than zero and so have a utility function with diminishing returns
both with respect to household consumption per capita and with respect to
household size.
There is yet another reason to be uncomfortable with the assumption that
λ = 0. If the term “golden rule” were not already taken in the growth theory
literature, we would use it here, for λ = 0 requires that those household mem-
bers making decisions in period t love others (the new household members
joining in period t + 1) as they love themselves. They assemble the house-
hold utility function by treating the personal utility that others receive in the
future from their consumption per capita as the equivalent of their own per-
sonal utility. Since we cannot call this the “golden rule,” we instead call it
perfect familial altruism. If 1 > λ > 0, there is imperfect familial altruism:
those making decisions in period t care about the personal utility of extra
family members in period t + 1, but not as much as they care about their own.

14. Romer (2000).

 300 Brookings Papers on Economic Activity, 1:2005

And if λ = 1, decisionmakers in period t act as if they care only about their

own personal utility. We are comfortable with altruism; we are uncomfort-
able with perfect familial altruism.
To the extent that changes in population growth are due to changes in
rates of international migration, the assumption that λ = 0 is not defensible.
The representative agent in period t would then regard the future-period
utility of unrelated strangers of different nationality who migrate into the
country on an equal footing as her own utility, or the utility of her direct
descendants.15
In this version of the Ramsey-Cass-Koopmans model, the first-order
condition for the representative household’s consumption-saving decision is

(1 + n ) 1− λ

(10) U ´(Ct ) dCt = U ´(Ct +1 ) dCt +1 .

(1 + β )
If the household faces a net rate of return on financial investments of rf,
then
1 + rf
(11) dCt = dCt +1 ,
1+ n
because resources in period t + 1 must be split among more members of the
expanded household.
For log utility we then have

(1 + n ) (1 + rf )
−λ
Ct +1
(12) = .
Ct (1 + β )
Along the economy’s steady-state growth path, with consumption per
worker growing at the rate of labor augmentation g, this becomes

rf = (1 + g ) (1 + n ) (1 + β ) − 1,
λ
(13)

and in the continuous-time limit,

(14) rf = β + g + λn.

15. Approximately 0.3 percentage point a year of the slowdown in labor force growth
projected by the Social Security trustees’ report (Board of Trustees, 2005) is due to a slow-
down in immigration.
Dean Baker, J. Bradford DeLong, and Paul R. Krugman 301

Looking across steady-state growth paths, one sees that reductions in the
rate of output growth per worker g reduce rf one for one in the case of log
utility. (They reduce rf by a multiplicative factor γ of the change in g in the
case of constant-relative-risk-aversion utility: U(Ct) = [(Ct)1 − γ]/[1 − γ].)
Reductions in the rate of labor force growth n also reduce rf except in the
case of λ = 0. If 1 > λ > 0, slower rates of labor force growth reduce rf, but
less than one for one. And if λ = 1, decisionmakers in period t are not
altruistic at all: they act as if they care only about their own personal utility,
and reductions in n reduce rf one for one—the same amount as do reduc-
tions in g.
The Ramsey model converges to a balanced-growth path, and this plus
the assumption of a representative agent is sufficient to nail down the rela-
tionship between economic growth and asset returns. In the steady state,
consumption per capita is growing at rate g, and so the relative marginal
utility of consumption per capita one period into the future is

(1 + β ) (1 + g )
−1 −1
(15)

in the case of log utility. And the rate at which consumption per capita can
be carried forward in time is

(1 + r )(1 + n )
−1
(16) f .

To drive the rate of return on capital rf away from

rf = (1 + g ) (1 + n ) (1 + β ) − 1
λ
(17)

requires that the consumption of those agents who are marginal in making
the consumption-saving decision in period t grow at a rate different from
that of growth in consumption per capita. This requires heterogeneous
agents. And the simplest suitable model with heterogeneous agents is the
Diamond model.

The Diamond Model

In the overlapping-generations model of Peter Diamond,16 each agent
lives for two periods, works and saves when young, and earns returns on

16. Diamond (1965).

 302 Brookings Papers on Economic Activity, 1:2005

capital and spends when old. Thus, for a given generation that is young in
period t, their labor income per worker when young wt, their consumption
per worker when young cyt, their consumption per worker when old cot+1,
the net rate of return on capital rt+1, and the economy’s capital stock per
worker in the second period kt+1 are all linked:

(18) wt = cyt + kt +1

(19) cot +1 = (1 + rt +1 ) kt +1.

With a Cobb-Douglas production function, output per (young) worker when

the period-t generation are young—in period t—is
α
 kt 
(20) yt = E 1− α
  ,
1 + n
t

where E is our measure of the efficiency of labor, growing at proportional

rate g each period, and where (1 + n) appears in the denominator because
n is the rate of population growth per generation. With this production
function, labor income is a constant fraction of output per worker,

(21) wt = (1 − α ) yt ,

and the real return on capital will be the residual, capital income, divided by
the capital stock:

αyt αEt1− α ktα −1

(22) rt = = .
kt (1 + n )α −1
Once again take time-separable log utility for our utility function,

ln ( cot +1 )
(23) )
ln ( cyt +
1+β
and look for steady states in capital per effective worker by requiring
that
(24) kt = Et k * .

From this we get the following steady-state first-order condition:

Dean Baker, J. Bradford DeLong, and Paul R. Krugman 303

1 (1 + r ) 1
(25) = .
cyt (1 + β ) cot +1
The model can be solved by substituting in the budget constraint,

1 (1 + r ) 1
(26) = ,
  kt 
a
 (1 + β ) 1 + r ) kt +1
(
(1 − α ) Et1− α   − kt + 1 
 1 + n 

to get

1 1
(27) = ,
1 − α  k*  α
(1 + β ) k *
   − k * 
1 + g 1 + n 
which leads to
 1 

 (1 − α )  
  1− α 
(28) k* = .
 (1 + g ) (1 + n )α ( 2 + β ) 

Recalling that r = (αk*α−1)/(1 + n)α−1, we have

 α(1 + g)(1 + n)(2 + β) 

(29) r =  .
 (1 − α )
In this equation, the lower the rate of productivity growth g, and the
lower the rate of labor force and population growth n, the lower is the rate
of return on capital r.

Conclusion
Thus, in the Diamond overlapping-generations model as well as in the
Ramsey model and the Solow model, slower economic growth comes with
lower net returns on capital. In the Ramsey model, there is reason to think
that reductions in labor productivity growth have a greater effect on rates
of return than do reductions in labor force growth:
—In the basic Solow algebra, the reduction in gross returns r is pro-
portional to (α/s) times the reduction in growth.
 304 Brookings Papers on Economic Activity, 1:2005

—In the Diamond model, the reduction in net returns rf is equal to

2α/(1 − α) times the reduction in labor productivity growth g and, to a
first approximation, equal to 2α/(1 − α) times the reduction in labor force
growth n.
—In the Ramsey model, the reduction in rf is equal (with log utility) to
the reduction in labor productivity growth g and, to a first order, to λ times
the reduction in labor force growth n (where λ is the degree to which
familial altruism is imperfect).
At some level, the same thing is going on in all three setups. Reductions
in economic growth in these setups are all declines in the rate of growth of
effective labor relative to the capital stock provided by previous investment.
Effective labor becomes relatively scarcer and capital relatively more abun-
dant. The terms of trade move against capital, and so the return to capital
falls.
Why, then, does a fall in labor force growth not reduce rates of return in
the Ramsey model in the case of perfect familial altruism, λ = 0? Because
a reduction in population growth also reduces the utility value of moving
consumption forward in time—an important component of the value of
saving in the Ramsey model with perfect familial altruism comes from the
possibility of dividing the saving among more people in the future and
thus escaping the diminishing marginal utility of consumption. Thus the
marginal household utility of saving falls in the Ramsey model when pop-
ulation growth falls. This fall reduces the effective supply of capital by
as much as the fall in the rate of population growth reduces the effective
supply of labor. To the extent that a slowdown in economic growth is
driven by a reduction in the rate of immigration, this representative-agent
effect in the Ramsey model is not an effect that we want the model to have:
perfect familial altruism is not an assumption that anyone would wish to
make.
These models say that there is some economic reason to believe that a
slowdown in economic growth would carry a reduction in asset returns
with it. These models are the standard models that economics graduate
students and their professors use routinely. They are oversimplified. They
are abstract. They are ruthlessly narrow in their conceptions of human moti-
vation and institutional detail. Are they relevant to the real world? Are
they telling us something that we should hear when we try to forecast the
long-run future?
Dean Baker, J. Bradford DeLong, and Paul R. Krugman 305

Arithmetic

Is it possible to imagine scenarios in which asset returns remain close

to their historical averages even when real GDP growth slows markedly?
Yes. Are any such scenarios plausible forecasts in the sense of being the
central tendency of a distribution of possible futures? We believe not. In
this section we conduct some simple arithmetic exercises to make our case.

Earnings and Returns

Jeremy Siegel believes that stocks are “in the middle range of fair mar-
ket value” and that therefore the current earnings yield of 5.45 percent is
a “good long-term estimate of real returns.”17 The sum of dividend payouts,
net buybacks, and investment financed by net retained earnings must add up
to 5.45 percent of today’s stock values.18 Returns to investors are payouts—
dividends and net buybacks—plus the value of investments financed by net
retained earnings.
Firms, which have traditionally paid out, on average, roughly 60 per-
cent of their accounting profits through dividends and buybacks and rely
on retained earnings to finance a substantial share of any increase in their
capital stock, have little room to boost risk-adjusted returns by massively
expanding payouts, unless they can do so without crippling their earnings
growth—that is, unless a good deal of today’s retained earnings are
wasted. Firms similarly have little room to boost risk-adjusted rates of
return on their equity by cutting back on payouts, unless there are very
large wedges between rates of return on retained and reinvested earnings
and rates of return in the market—that is, unless firms have been mas-
sively underinvesting. Current earnings yields thus suggest that the stock
market is in accord with the logic of our algebra and analysis: it is not
anticipating the average real return on the stock market of 6.5 percent a
year or so realized over the past half century.
But reported accounting earnings are not true Haig-Simons earnings
(that is, equal to the amount that can be consumed from earnings without

17. Siegel (2005, p. 8).

18. There is a wedge of 0.3 percentage point a year between the GDP deflator and the
CPI. Siegel’s estimated real rate of return becomes 5.15 percent a year in the CPI-basis
numbers used by the Social Security Administration.
 306 Brookings Papers on Economic Activity, 1:2005

changing wealth).19 There is good reason to believe that returns on retained

earnings are higher than market returns.20 And it is at least plausible that
the wedge between market returns and returns on retained earnings depends
on the rate of economic growth: faster growth means higher demand and
greater profits if returns to scale are increasing. So the argument that earn-
ings yields do not support high expected equity returns needs to be shored
up by an explicit look ahead at how payouts and values might evolve.21

Dividend Yields, Returns, and Growth

Begin with the identity that is the Gordon equation for equity prices:

D
(30) P= ,
re − g

where D are the dividends paid on a stock or an index of stocks, P is the

corresponding price, re is the expected real rate of return on equities, and
g is the expected permanent real growth rate of dividends. This is a stan-
dard way to approach the determinants of equity prices as a whole.22 In
this framework the real rate of return on equities is

D
(31) re = + g.
P
Returns on an index of stocks differ from the current dividend yield plus
the growth rate of economy-wide corporate earnings for two important
reasons:
—First, g will be less than the growth rate of economy-wide corporate
earnings because those earnings are the earnings of newly created compa-
nies that were not in the index last period. Corporate earnings are a return
to entrepreneurship as well as capital; hence the rate of growth of econ-
omy-wide earnings will in general outstrip that of the earnings of the
companies represented in a stock index.
—Second, dividends are not the only way firms pump cash to share-
holders. Stock buybacks decrease the equity base and thus push up the

19. Haig (1921).

20. See Hubbard (1998).
21. See Baker (1997) for the first argument along these lines of which we are aware.
22. Campbell and Shiller (1988).
Dean Baker, J. Bradford DeLong, and Paul R. Krugman 307

rate of growth of the earnings on the index (as opposed to the earnings of
the companies in the index).
It is convenient to think of both of these factors as affecting the payout
ratio rather than the growth rate, and to replace equation 31 with

D+B
(32) re = + g,
P
where B is net share buybacks (buybacks less initial public offerings), and
g is now the growth rate of D + B.23
The 2005 report of the Social Security trustees projects a long-run real
GDP growth rate of 1.8 percent a year on a GDP deflator basis.24 It projects
that labor and capital shares will remain constant in the long run.25 With a
long-run gap of 0.3 percentage point between the consumer price index
(CPI) and the GDP deflator,26 and with an auxiliary assumption that capi-
tal structures are in balance, this is an implicit forecast that the variable g
in the Gordon equation will be 1.5 percent a year. Current dividend yields
on the Standard and Poor’s (S&P) 500 index are 1.9 percent a year. Current
net stock buybacks are 1.0 percent a year. The sum of these is 4.4 percent
a year, which is thus the expected real rate of return r in the Gordon equa-
tion. That is significantly lower than the 6.5 percent real rate of return that
is the historical experience of the American stock market.

Possible Ways Out

Are there ways to escape from this arithmetic of earnings and payouts?
Yes. The U.S. economy is not on a steady-state growth path. Three potential
ways out seem most worth exploring:
—Perhaps the stock market is currently overvalued and will decline,
significantly raising payout yields.

23. Subtracting initial public offerings ensures that the ratio of total economy-wide
earnings to the earnings of companies in the index does not grow. Adding gross buybacks
takes account of the antidilution effects of narrowing the equity base of companies cur-
rently in the index.
24. Board of Trustees (2005, table V.B2).
25. The assumption of a constant income share follows from the derivation of real
wage growth from productivity growth, which is discussed on pages 85–88 of Board of
Trustees (2005).
26. Board of Trustees (2005, table V.B1).
 308 Brookings Papers on Economic Activity, 1:2005

—Perhaps payout growth will be unusually rapid in the near term before
slowing to its long-term forecast trend rate of 1.5 percent a year.
—Perhaps the distribution of world investment will shift in a way that
allows U.S. companies to earn greater and greater shares of their profits
abroad.
Diamond argues for the first possibility.27 A decline in the stock mar-
ket, relative to the economy’s growth trend, of 40 percent would carry pay-
out yields up to the 5.0 percent consistent with a long-run real return of
6.5 percent a year and real profit and dividend growth (on a CPI basis) of
1.5 percent a year. Such a scenario is certainly possible: it was the stock
market’s experience between the late 1960s and the early 1980s. But we
have a hard time seeing it as the central tendency of the distribution of
possible futures.28
The second possibility requires payouts—both dividends and net stock
buybacks—to grow rapidly in the near term to validate a subsequent real
growth rate of 1.5 percent a year and a current expected real return of
6.5 percent a year. If such growth were to be concentrated in the next
decade, the real payouts of the companies in the S&P index would have to
grow at an average of 8.6 percent a year. Over the past fifty years the earn-
ings on the S&P index have grown at an average rate of 2.1 percent a year.
It could happen: perhaps we are in the middle of a permanent shift in the
distribution of income away from labor and toward capital. But, once again,
we regard these as unlikely scenarios, not as the central tendency of the
distribution of possible futures that is a rational forecast.
The third way out is the one that we regard as the most interesting pos-
sibility. We take it up in the next section.

The Open-Economy Case

In any open economy the steady-state Gordon equity valuation equation

is as before, except that the rate of growth is not that of the domestic cor-

27. Diamond (2000).

28. Certainly no investment adviser who anticipates that real equity returns will aver-
age −0.6 percent a year over the next decade has any business advising clients to shift their
portfolio in the direction of equities today. That is true even when the U.S. government is
the adviser, and the relatively young future beneficiaries of Social Security are the clients.
Dean Baker, J. Bradford DeLong, and Paul R. Krugman 309

porate capital stock g but that of the capital stock owned by American
companies, gk:

D+B
(33) re = + gk .
P

If foreign companies, on net, invest in America—that is, if the United States

on average runs a current account deficit—then the rate of growth of the
earnings of American companies in our domestic stock market index will
be slower than the rate of growth of economy-wide earnings and of real
GDP. The open economy will then deepen rather than resolve the problem
of combining slow expected growth with high expected returns. If instead
it is American companies that, on net, invest abroad, then the rate of growth
of the capital stock, and thus of the earnings of companies in the index,
will exceed the rate of growth of the domestic economy g.
How much larger? If we look over spans of time long enough for adjust-
ment costs in investment not to be a major factor, the value of the capital
stock will be proportional to the size of the capital stock.29 If we assume in
addition that companies maintain stable debt-equity ratios, we have

Y 
(34) gk = g + x   ,
 K

where x is that component of the current account surplus (as a share of GDP)
that corresponds to American companies’ net investments abroad,30 and
Y/K is the ratio of current output to corporate capital.

29. We here dismiss the possibility that investments overseas might provide higher
risk-adjusted rates of return in the long run than domestic investments: Tobin’s q = 1 both
here and abroad. The Bureau of Economic Analysis reports that as of the end of 2003 the
market value of foreign-owned assets in the United States is about $10.5 trillion, compared
with foreign assets held by U.S. residents of about $7.9 trillion, yet the associated income
flows are about the same. We attribute this difference to a difference in risk. The experience
of nineteenth-century British investors with such landmarks of effective corporate gover-
nance as the Erie Railroad suggests that, although there are supernormal returns to be
earned in the course of rapid economic development, people with offices separated by
oceans are unlikely to be the ones who reap them.
30. The phrase “corresponds to American companies’ net investment abroad” is needed
to abstract from current account deficits that finance net government consumption or net
household consumption.
 310 Brookings Papers on Economic Activity, 1:2005

Here, again, we return to arithmetic. Our rate of return on equities is

D+B Y 
(35) re = + g + x .
P  K

From the previous section this is

Y 
(36) re = 4.4% + x   .
K

Assuming a capital-output ratio of 3, we then have

(37) x = 3( re − 4.4 percent ).

In words: for any excess of the rate of return on equities over the closed-
economy benchmark case of 4.4 percent a year, three times that figure is
the current account surplus associated with net corporate investment over-
seas needed to produce the higher return.
Note that, for a constant rate of return, the needed surplus grows over
time. In equations 34 through 36, Y/K is not the physical domestic output-
to-capital ratio; it is the ratio of domestic output to total capital owned by
American companies—including capital overseas. As overseas assets mount,
the needed surplus for constant payout yields mounts as well.
Such enormous current account surpluses are possible. Great Britain
had them in the quarter-century before World War I, when it ceased to be
the workshop of the world and became for a little while its financier.31
Slowing economic growth in the late Victorian and Edwardian eras and
reduced investment relative to national saving were cause (or conse-
quence, or possibly both) of the direction of Britons’ saving and of British
companies’ investment overseas. We see no signs that the United States
will undertake a similar trajectory over the next several generations. And
we are impressed by the scales involved: to be consistent with current pay-
out yields, and given a forecast real GDP growth rate of 1.8 percent a year,
to achieve 6.5 percent annual returns on equity the current account surplus
produced by American net corporate investment abroad would have to
begin at 6 percent of GDP and grow thereafter.

31. Edelstein (1982).

Dean Baker, J. Bradford DeLong, and Paul R. Krugman 311

Could such large outward levels of net corporate investment abroad be

consistent with relatively balanced overall trade—in other words, could
they be offset by large net portfolio investment inside the United States?
Not without additional forces at work. The reason is that the open-economy
saving-investment relation,

(38) S − NX ≡ I ,

(where NX is net exports) is an identity. Consider the three uses that such
large inward portfolio investments could have:
—They could be used to purchase securities newly issued by American
businesses to finance investment in the United States. The flow of inward
portfolio investment would add as much to domestic investment as the
outward-directed flow of corporate investment would have subtracted. There
would be no slowdown in the rate of growth of the domestic capital stock.
Thus the rising domestic capital-output ratio would push down rates of return
at home. Since foreigners are making these large portfolio investments in
the United States, this fall in domestic rates of return would be associated
with a similar fall in foreign rates of return as well.
—They could be used to purchase securities newly issued by Ameri-
can businesses to finance investment abroad. In this case, gross foreign
direct investment by domestic firms would have to be large enough not
only to absorb the difference between domestic investment and domestic
saving, and so slow down the rate of growth of the domestic capital stock,
but also to neutralize the portfolio capital inflow. We are thus back to
square one.
—They could be used to purchase already-existing assets from Ameri-
cans, who then do not reinvest the proceeds either in expanding the domes-
tic capital stock or in further funding American investment abroad, but
instead consume the proceeds.32 This means massive dissaving on the part
of those who sell their assets to foreigners: a large fall in S. Once again,
we see a possible scenario but not the central tendency of the distribution
of possible futures that would constitute a forecast.

32. This is the possibility that Mankiw stresses in his comment on this paper: that if
domestic saving rates fall sharply, the reduction in the rate of growth of the domestic capi-
tal stock required to keep rates of return high can be accomplished without a large current
account surplus.
 312 Brookings Papers on Economic Activity, 1:2005

The Equity Premium

Economists do not have a good explanation of the equity premium.

Rajnish Mehra and Edward Prescott titled their well-known paper “The
Equity Premium: A Puzzle,” for good reason.33 Stocks have outperformed
fixed-income assets by more than 5 percentage points a year for as far
back as records go. As Martin Feldstein, former chairman of the Council
of Economic Advisers, has often said, it is as if the market’s attitude toward
systematic equity risk were that of a rich sixty-five-year-old male with a
not-very-healthy lifestyle, whose doctor has told him that he is likely to live
less than a decade. Yet we believe that properly structured markets should—
and can—mobilize a much deeper set of risk-bearers with a much greater
risk tolerance. That they do not appear to have done so is a significant
mystery. We find ourselves persuaded by Mehra that the equity premium
remains a puzzle, unexplained by rational agents in models that maximize
individual utility.34
It is quite possible that a substantial part of the equity premium is a
thing of the past, not the future.35 In the distant past the fear of a recurrence
of the railroad and other “robber baron” scandals, and in the more recent
past the memory of the Great Depression, kept some investors excessively
averse to stocks. In addition, the United States has had remarkably good
economic luck—a point stressed by Robert Shiller.36 And, over time, as
people realized that their predecessors had been excessively fearful of
equity risk, rising price-dividend ratios pushed a further wedge between
stock and bond returns. But today our arithmetic projects stock returns of
4.4 percent a year, for an equity premium of perhaps 2.5 percentage points,
not 5.
To the extent to which this past behavioral anomaly was the result of
an excessive fear of stocks and an excessive attachment to bonds, it is not
clear that its erosion should have an impact on the expected return on a
balanced portfolio. The simplest, crudest, and most extremely ad hoc model
of the equity premium would embed the stock-versus-bond investment
decision in the simplest possible Diamond-like overlapping-generations

33. Mehra and Prescott (1985).

34. Mehra (2003).
35. In conversation, Randall Cohen of the Harvard Business School has been an espe-
cially forceful advocate of this point of view.
36. Shiller (2005).
Dean Baker, J. Bradford DeLong, and Paul R. Krugman 313

model, with the capital stock each period being the wealth accumulated
when young by the old, retired generation. Assume that each generation,
when it saves, invests a share eh of its savings in equities and a share 1 − eh
in bonds. Firms, however, are unhappy with such a capital structure. Unwill-
ing to run a significant risk of bankruptcy, they are unwilling to commit
less than a share ef, where ef > eh, of their payouts to equity. A smaller
cushion—in the sense that a smaller cyclical decline in relative profits
would run the risk of missing bond payments and drawing an appointment
with a bankruptcy court—is simply unacceptable to entrenched managers.
If a physical unit of saving when one is young yields returns to physical
capital r when one is old, the rates of return on equity and debt, re and rd,
respectively, are then calculated as

e 
(39) 1 + re = (1 + r )  f 
 eh 

 1 − ef 
(40) 1 + rd = (1 + r )  ,
 1 − eh 

with the equity premium being

1 + re e (1 − e f )
(41) = f .
1 + rd eh (1 − eh )

In this excessively simple framework, it does seem highly plausible that

(1 + re)/(1 + rd) has fallen with greater household willingness to hold
equity, because of institutional changes (such as revisions of the “prudent
man” rule, the growth of IRAs and 401(k)s, and lower transactions costs
associated with stock trading), the fading memory of 1929, two decades
of fabulous bull markets, and increased financial sophistication on the part
of households.
Thus, even if there were no reasons connected with slowing growth to
expect lower returns on capital, one might well anticipate lower returns
on equity in the future than in the past. And past decades have seen insti-
tutional changes that one would expect, from a behavioral perspective, to
boost the share of financial assets channeled to equities.37

37. Barberis and Thaler (2003).

 314 Brookings Papers on Economic Activity, 1:2005

A lower rate of return on the assets in a balanced portfolio has power-

ful implications for economic policy. A lower equity premium seems, to
us at least, to have powerful implications for one issue, namely, whether the
stock market’s apparent failure to mobilize society’s risk-bearing capacity
is a large-scale market failure, and whether a government-run social insur-
ance scheme can and should attempt to profit from (and thus repair) this
failure to mobilize society’s risk-bearing resources. The government has
the greatest ability of any agent in the economy to manage systematic risk.
If other agents are not picking up their share—and if, as a result, there are
properly adjusted excess returns to be earned by the government’s taking
a direct position itself or assuming an indirect position by reinsuring indi-
viduals’ social insurance accounts—why should the government not do so?
The difference between, broadly speaking, the economists of the coasts
and the economists of the interior is that the first specialize in thinking
up clever schemes to repair apparent market failures, whereas the second
specialize in thinking up clever reasons why apparent market failures are
not really so. Even though we are from the coasts, we find enough rea-
sons to believe that the equity premium will be smaller in the future than
in the past to prefer that attempts to exploit it be implemented slowly and
gingerly.

Conclusion

We see strong reasons to think that, over the long run, rates of return on
assets are correlated and causally connected with rates of economic growth.
We would expect the reduction in asset returns to be greater for a given
reduction in productivity growth than for an equal reduction in labor force
growth. We think that reductions in asset returns could be offset and even
neutralized by other factors—by capital expropriating some of what has
been labor’s share of income, by a failure of today’s stock market values
to soberly reflect likely future returns rather than irrational exuberance, or
by the United States cutting its consumption beneath its production for
generations and following Britain’s pre–World War I trajectory as sup-
plier of capital to the world. But we see these as unlikely (although possi-
ble) scenarios. We do not see any of them as the central tendency of the
distribution of possible futures that is a proper economic forecast. And
although a combination of partial moves in each of the three directions could
Dean Baker, J. Bradford DeLong, and Paul R. Krugman 315

achieve the result, we see no good reason to presume that such a scenario
is likely.
We see the two strands of our argument—our arithmetic demonstration
that equity returns as high in the future as in the past are unlikely, and our
analytical arguments that rates of return and rates of growth are likely to
move together—as reinforcing each other. Returns must be consistent with
the saving decisions of households, the investment decisions of firms, and
the technologies of production. But returns must also equal payout yields
plus capital gains—only in stock market bubbles can capital gains diverge
widely from economic growth, and then only for a little while.
Powerful economic forces work to make sure that what the economy’s
behavioral relationships produce is consistent with its equilibrium flow-of-
funds conditions. That is the logic that applies here: if slower economic
growth reduces the arena for the profitable deployment of capital, rates of
return will fall until less capital is deployed. By how much will they fall?
Until—in steady state—payout yields plus retained earnings are equal to
profits, and retained earnings are no larger than the sustainable growth of
the capital stock permits.
Comments and
Discussion

N. Gregory Mankiw: This paper by Baker, DeLong, and Krugman is really

three papers in one. The first paper is a straightforward review of how
population growth affects the return to capital in standard models of eco-
nomic growth. The second paper is a discussion of what return one should
expect for the stock market in the coming decades, given current measures of
valuation. The third paper offers some ruminations about the equity premium.
What links the three papers is their motivation. President Bush has called
for reform of the Social Security system. According to the Social Security
actuaries, the system faces large unfunded liabilities. That conclusion, how-
ever, is based on a projection that includes much slower labor force growth
(and thus economic growth) than the United States has experienced his-
torically. This raises the question of what rate-of-return projections should
be assumed as the nation considers possible reforms.
When evaluating reform proposals, the Social Security Administration
uses a projected real annual return on equities of 6.5 percent (which, given
the trustees’ assumption about the risk-free rate, implies an equity premium
of 3.5 percent). Paul Krugman has written elsewhere that “a rate of return
that high is mathematically impossible unless the economy grows much
faster than anyone is now expecting.”1 This three-in-one paper began as an
attempt to justify that assertion. I will discuss each of the three papers in turn,
before addressing the policy motivation.
POPULATION AND GROWTH THEORY. The first paper in this paper reviews
several standard neoclassical growth models. The aim is to see what these
models predict for the relationship between population growth and the rate
of return to capital.

1. “Many Unhappy Returns,” New York Times, February 1, 2005.

316
Dean Baker, J. Bradford DeLong, and Paul R. Krugman 317

The Solow growth model gives a clear answer to this question: slower
population growth lowers the rate of return. Because the saving rate is fixed,
slower population growth raises the steady-state capital-labor ratio, which
in turn means a lower marginal product of capital. The Diamond model
gives a similar answer, at least for the functional forms assumed here.
The Ramsey model, however, leads to a very different conclusion. In that
model the saving rate adjusts so that the rate of return is invariant to the pop-
ulation growth rate. This adjustment of the saving rate is economically intu-
itive: if there are going to be fewer people in the future, we need to save less
for the future.
This conclusion is the essence of the analysis presented in a 1990 Brook-
ings Paper called “An Aging Society: Opportunity or Challenge?” written
by David Cutler, James Poterba, Louise Sheiner, and Lawrence Summers.2
They used a standard Ramsey model to argue that, “the optimal policy
response to recent and anticipated demographic changes is almost certainly
a reduction rather than an increase in the national saving rate.” I should note
that national saving is currently low by historical standards, but I will not
suggest that this is necessarily the “optimal policy response” that Cutler
and his coauthors were proposing.
Realizing that the Ramsey model does not support the main contention of
the paper, Baker, DeLong, and Krugman propose a new but unpersuasive
generalization of it. The authors claim that the standard Ramsey model is one
of “perfect familial altruism.” That is not how I would describe it. Even the
standard Ramsey model includes discounting, so that my utility is weighted
more heavily than that of my children and grandchildren. What the proposed
generalization does is make the effective discount rate for future utility
depend on the population growth rate. When population growth slows, the
effective discount rate falls, and this fall in the discount rate blunts the
decline in the saving rate that occurs in the standard Ramsey model.
Is this generalization appealing? Not to me. As the parent of three chil-
dren, I can attest that one of the things parents do when child N is born
is to assure the N − 1 children that they will be loved just as much. The
generalization of the Ramsey model proposed here is, in essence, a denial
of this claim.
In the end it is clear that the tools of modern growth theory lead to an
ambiguous answer about how population growth affects the return to capital.

2. Cutler and others (1990).

 318 Brookings Papers on Economic Activity, 1:2005

One can write down textbook models in which the two variables move
together (the Solow model), and one can write down models in which
they do not (the Ramsey model). The natural response to this theoretical
ambiguity is to muster evidence, either from time-series data or from the
international cross section, about the actual effect of population growth.
This paper, however, presents no evidence one way or the other. Perhaps
that is a subject for a future Brookings Paper.
STOCK MARKET VALUATION. The second paper in this paper discusses
the expected return on the stock market. The authors begin with the obser-
vation that the current average earnings yield is 5.23 percent a year, which
is about a percentage point lower than the historical average. As a result, they
expect future stock returns to be lower than historical averages as well.
I give some weight to this piece of evidence. It is possible that the Social
Security Administration’s assumption of 6.5 percent a year for equity returns
is about a percentage point too high. The risk-free rate assumption of 3 per-
cent a year may also be about a percentage point too high, as judged by
current yields on long-term inflation-indexed bonds. The equity premium of
3.5 percent, however, seems about right.
After observing the earnings yield, the authors consider stock market
valuation from the perspective of the famous Gordon formula, according
to which the expected return on a share of stock equals the current dividend
yield plus the projected growth rate of dividends per share. Although the
Gordon formula has a long and venerable tradition, I don’t think it provides
a particularly edifying approach here. For a neoclassical economist, the
starting point for thinking about the role of dividends in stock valuation is
the classic Modigliani-Miller theorems, which tell us that the dividend pay-
out is irrelevant to the value of the firm. It seems unnatural for purposes of
stock valuation to focus on the level and growth of a variable that, to a first
approximation, does not matter.
If the dividend yield is approximately irrelevant, as Modigliani and Miller
tell us, then it is easy to imagine that it could undergo a major change in the
years to come. Looking ahead, it seems plausible to me that dividend pay-
outs broadly construed could rise significantly. If we are about to experience
a period of slower economic growth because of demographic change, then
firms might well have fewer profitable investment opportunities and, as a
result, might decide to pay out a larger percentage of their earnings. There
are several ways this could occur. One possibility is by increasing normal
dividends or share repurchases. Another is through corporate reorganizations.
Dean Baker, J. Bradford DeLong, and Paul R. Krugman 319

Corporate managers might find cash takeovers and acquisitions more prof-
itable than internal expansion. Cash purchases of other businesses take
money out the corporate sector and are, in essence, a form of share repur-
chases. They are another way to increase dividends, broadly construed.
OPEN-ECONOMY STOCK VALUATION. The authors then consider a related
open-economy issue. Is it possible, they ask, for growth in dividends to
significantly exceed growth in the domestic economy because corporations
are investing and earning profits abroad? They suggest that this is unlikely,
on the apparent ground that it would require something implausible about
capital flows. I am not convinced.
Suppose that General Electric, seeing fewer profitable investments in the
United States, uses some of its earnings to buy a factory in China. That rep-
resents a capital outflow from the United States and a current account sur-
plus, which I think is what the authors have in mind. But consider what the
Chinese former owners of the factory, who now have dollars from the deal,
might do with them. One possibility is that they buy U.S.-produced goods,
which would indeed mean a current account surplus for the United States.
Another possibility is that they buy U.S. assets. They might even buy stock
in General Electric or be given GE stock as part of the transaction. In this
case General Electric can diversify abroad while the United States has bal-
anced trade.
Here is one scenario that seems plausible to me. With much of the rest of
the world, such as China and India, growing so rapidly, U.S. companies will
increasingly find profitable opportunities abroad. At the same time, for-
eigners will increasingly invest in U.S. companies, which will be among
the driving forces behind global growth. Under this scenario an increasing
share of the earnings of U.S. corporations could come from abroad, with-
out any obvious implications for the U.S. current account.
The authors conclude that this is a “possible scenario but not the central
tendency,” but they do not cogently explain how they reach this conclusion.
At least we can agree it would be a mistake to call this scenario “mathe-
matically impossible.”
THE EQUITY PREMIUM. Let me turn now to the last paper in this paper,
which concerns the equity premium. Here the authors give us a model that
is creative, bizarre, or vacuous, depending on your point of view.
Most analysis of the equity premium begins with the premise that it has
something to do with the trade-off between risk and return. Not so in this
model. Here the household sector decides exogenously what fraction of
 320 Brookings Papers on Economic Activity, 1:2005

wealth to put in equities, and the corporate sector decides exogenously

what fraction of the capital return to pay out to equity holders. From these
two exogenously determined shares, the equity premium emerges.
The model reminds me of John Kenneth Galbraith’s worldview. House-
holds are not sufficiently intelligent to make portfolio decisions based on
risk and return. Corporate managers are sufficiently immune to market forces
that they divide up the economic pie however they see fit. If I took this model
seriously, it would do more than inform my view of the equity premium. It
would shake my faith in corporate capitalism!
But there is a less dramatic way to view this part of the paper. Perhaps
the equations presented here should be viewed less as behavioral descrip-
tions and more as accounting identities. If this interpretation is right, then
I am at a loss about what purpose these equations serve. They do not seem
readily adapted to calibration, to gauge how the equity premium has changed
over time. I am comfortable with the authors’ suggestion that the equity
premium may be smaller in the future than it has been in the past because
institutional changes have made the spreading of risk more efficient. But
this model does not shed much new light on this familiar conjecture.
IMPLICATIONS FOR SOCIAL SECURITY REFORM. Let me close with a few
words about Social Security reform. The authors were drawn to this set of
topics because they think it is central to the debate over the president’s
reform proposal. I disagree.
There are two elements to the president’s proposal. First, the president
wants to eliminate the system’s unfunded liabilities by bringing promised
benefits into line with the dedicated payroll tax revenue. Various ideas for
doing this have been put on the table, such as raising the retirement age and
changing indexation rules. Second, the president wants to give workers the
option of converting some of their defined-benefit retirement income from
Social Security into a defined contribution, which would be placed in a
personal account and invested in a broadly diversified portfolio of stocks
and bonds.
Reasonable people can disagree about the merit of these proposals. I made
the case for the president’s proposals as I see it in a recent article in the New
Republic.3 The case for reforming benefits is that the government should
not promise more than it has the wherewithal to pay. The case for moving

3. “Personal Dispute,” The New Republic, March 21, 2005.

Dean Baker, J. Bradford DeLong, and Paul R. Krugman 321

Social Security from a defined-benefit to a defined-contribution structure

is that it gives individuals more choice and control over their retirement
income and the government greater transparency in its finances.
These arguments are not based on any particular estimate of the average
return to capital or of the equity premium. I don’t think the key issue in
the debate over Social Security is whether, over the next century, the risk-
free annual return will be 1 or 3 percent, or whether the equity premium will
be 2 or 4 percent. So even if I agreed with the arguments raised in this paper
and lowered my estimates of rates of return, it would not change my mind
about the need to reform Social Security or the kinds of reforms that are
desirable.
I would guess that, in their hearts, the authors agree with me about this.
To see if I am right, I would like them to ponder the following question:
Suppose that, next week, the stock market falls by 50 percent, so that div-
idend and earnings yields double. Would Baker, DeLong, and Krugman
suddenly change their minds and favor President Bush’s proposal for Social
Security reform? I suspect they would not. If I am right, this suggests that
although the paper raises some interesting questions about the future of
asset returns, as far as the debate over Social Security goes, it is largely a
non sequitur.

William D. Nordhaus: What are the prospective returns on stocks? This is

a question with multi-trillion-dollar stakes, and so it is always of much
interest to many people. It has recently become a political as well as an eco-
nomic question with the proposal by the Bush administration to introduce
private (or personal) accounts as a component of Social Security pensions.
This paper by Baker, DeLong, and Krugman argues that the returns assumed
in the Bush administration’s calculations are too high, that historical returns
do not provide a reliable benchmark for future returns, and that structural
changes to the U.S. economy are likely to lower returns in the future. The
last act of the political-economic drama is missing from this play, however,
for the authors do not discuss in any detail the consequences of their findings
for Social Security.
My comments are primarily directed to the question of estimating the
prospective long-run returns to equity. There are several approaches to this
question. One is to examine historical returns. Estimates here are highly
dependent on the time period. A second approach, which the paper uses, is
based on the Gordon equation. This approach is difficult to apply because
 322 Brookings Papers on Economic Activity, 1:2005

it relies on price-per-share data, ignoring the fact that the number of shares
can change over time.
To provide some perspective, I will look at the “fundamental” return on
stocks by examining the rate of return on corporate capital, using different
approximations. Before going down this road, two caveats should be men-
tioned. First, the returns data customarily examine the returns to nonfinancial
domestic corporations. This component of profits is, unfortunately, a declin-
ing share of corporate profits as measured in the national accounts. After
making up around 85 percent of corporate profits in the decade after World
War II, the share of the domestic nonfinancial sector has declined to slightly
over 50 percent in the last three years. The second shortcoming in using
national accounts data is that there are major accounting differences between
reported book profits and national accounts profits.1 Book profits contain
several inappropriate items (such as gains on pension plans), whereas national
accounts profits are more comparable over time but are limited to income
earned on domestic production. Moreover, neither concept is conceptually
equivalent to a measure of true income.
My table 1 shows several measures of the returns to capital based on
national accounting data on nonfinancial corporations. The first column
presents a first approximation, the real rate of return after tax, measured as
total property return divided by the replacement or market value of real assets.
Total property return includes both interest and profits after corporation
taxes in the numerator, with the current value of fixed capital, software,
inventories, and land in the denominator. According to this first approxima-
tion, the return to capital averaged 6.1 percent a year over the 1960–2004
period. The return in 2004 was slightly above the long-term annual average,
at 6.9 percent.
A second approximation would take into account the cyclical nature of
profits. I have taken a very simple approach, using as a cyclical variable
the difference between the actual unemployment rate and the Congressional
Budget Office’s estimate of the non-accelerating-inflation rate of unemploy-
ment (NAIRU). The cyclical term in the regression is significant and explains
some of the peaks and troughs of the rate-of-return series but does not change
the long-term picture. Most important for my purposes is that the latest year
(2004) shows a cyclically corrected rate of return of 7.0 percent a year, also
slightly above the long-term average (second column in table 1).

1. An excellent discussion of the differences is contained in Petrick (2001).

Dean Baker, J. Bradford DeLong, and Paul R. Krugman 323
Table 1. Alternative Rate-of-Return Measures and Q Ratios, Various Periods,
1960–2004
Percent a year except where stated otherwise
Domestic nonfinancial corporations

Average rate of profita Prospective ten-

year return with Average earnings
Not cyclically Cyclically Average reversion of Q yield on
Period corrected corrected b Q ratioc toward 1a,d S&P 500e
1960–69 7.1 6.8 0.85 7.8 5.7
1970–79 5.3 5.4 0.62 8.7 9.0
1980–89 5.6 6.1 0.55 10.4 8.9
1990–99 6.5 6.6 1.11 5.9 4.8
2000–04 5.9 5.9 1.13 5.2 3.6
1960–2004 6.1 6.2 0.82 7.5 6.7
2000 6.0 5.5 1.65 2.5 3.6
2004 6.9 7.0 0.97 7.2 4.9
Sources: Bureau of Economic Analysis, “Note on the Profitability of Domestic Nonfinancial Corporations, 1960–2001,” Survey
of Current Business, September 2002, pp. 17–20, updated by the author with data on profits and capital from www.bea.gov; Federal
Reserve data from www.federalreserve.gov/releases/z1/current/default.htm; Data Resources, Inc. (DRI); Standard and Poor’s
data from www2.standardandpoors.com/spf/xls/index/SP500EPSEST.XLS
a. Property income after tax (national accounts concept) divided by the current or replacement cost of real assets (capital, land,
inventories).
b. Profits estimated using a cyclical correction derived from regressing the rate of profit in the first column on the difference between
the unemployment rate and the Congressional Budget Office’s estimate of the non-accelerating-inflation rate of unemployment
(NAIRU) using annual data.
c. Ratio of the market value of equities plus net financial assets to the replacement cost of real assets, all for U.S. nonfinancial
corporations (constructed from Federal Reserve data).
d. Estimates constructed by adjusting the cyclically corrected return on the assumption that Q reverts to 1 at an exponential rate
of 10 percent a year.
e. Historical data from DRI, using reported earnings, updated by the author using data from Standard and Poor’s.

A third approximation would take into account that Q (the ratio of the
market value of capital to its replacement cost) may differ from 1. This is
controversial in financial economics, with some economists holding that a
Q ratio that differs from 1 is not possible because of the fine arbitrage of
markets. Readers who believe that can simply skip the discussion of this third
approximation. Baker, DeLong, and Krugman assume that Q (by which I
think they mean average Q) equals 1, but it is worth thinking about what
difference it would make to the results.
Incorporating a nonunitary Q into the prospective returns analysis is com-
plicated, but I discussed the essence of the matter in an earlier Brookings
Paper.2 That paper examined several cases, but the most interesting is the one
in which Q fluctuates, because of animal spirits or irrational exuberance, and

2. Nordhaus (2002).
 324 Brookings Papers on Economic Activity, 1:2005

then reverts to 1 at an exponential rate of 10 percent a year (a rate consistent

with historical data). The third column of table 1 shows an estimate of Q for
nonfinancial corporations, and the fourth column shows the prospective
return under the reversion model of the behavior of Q. The Q-adjusted and
cyclically adjusted prospective return in 2004 was 7.2 percent a year. As
it turns out, this makes very little difference to the estimate using 2004 data,
because Q was very close to 1 for 2004. It would have made a substantial
difference at the peak of the stock market bubble in 2000 or at the trough
of irrational malaise in the 1970s and early 1980s.
A final calculation would look at the Standard and Poor’s earnings-price
ratio (last column of table 1). Ignoring accounting and coverage differences,
the earnings-price ratio should equal the after-tax rate of return on capital
divided by average Q. This relationship is reasonably close for the entire
1960–2004 period. However, the two approaches provide quite different
answers for 2004. The ratio for 2004 was 4.9 percent. The implicit Q in
the Standard and Poor’s returns is seriously inconsistent with the national
accounts estimates just presented: to get the same rate of return would require
a Q ratio of 1.44. I suspect most of the difference is due to differences in
sectoral coverage. Financial corporations have a much higher rate of return
to capital than nonfinancial corporations, and the share of financial corpo-
rations has risen in recent years. As a result the implicit Q will be higher
for the Standard and Poor’s calculations, which includes financial firms.
Overall, this look at the “fundamentals” suggests a prospective return of
around 7 percent a year starting from a base year of 2004.
The above analysis of prospective returns assumes more or less unchanged
underlying conditions. The major thrust of the paper, however, is to exam-
ine the impact of potential structural shifts on the rate of profit and hence on
the rate of return on equities. The authors note correctly that most closed-
economy models would predict a decline in the rate of profit with a
decline in the growth of labor inputs or of labor-augmenting technological
change, all else unchanged.
The authors’ discussion of rates of return in an open economy raises
more questions than I can answer. However, at least two developments
could well raise the return on equities. The first is the impact of correcting
the U.S. current account imbalance. Virtually every study of this issue
(including those in the present volume) suggests that at some point the
dollar is in for a big real depreciation. Such a depreciation would be expected
to raise domestic profits as well as provide capital gains on foreign assets.
Dean Baker, J. Bradford DeLong, and Paul R. Krugman 325

A second and more complicated issue involves the potential opening of

foreign economies to U.S. capital. That process has begun but is by no
means complete. It may turn out that opening foreign markets would raise
the potential return on foreign investment, which would be part of U.S.
stock returns.
The final point I would emphasize concerns uncertainty. The fact is that
we cannot precisely estimate the prospective return on stocks. The authors’
estimates are serious ones. Their analysis concludes that the appropriate esti-
mate for future real returns is 41⁄2 percent a year, less 1 or 2 percentage points
should a slowdown occur. My estimates suggest something more like 7 per-
cent a year, with an uncertain adjustment for future changes. These are cen-
tral tendencies, even before taking into account inherent variabilities. It may
be possible to reduce this factor-of-two uncertainty, but I doubt it.
Given the inherent uncertainty, policies should be robust to a significant
range of possible outcomes. Policies should also be intrinsically robust—
that is, able to withstand a financial meltdown without causing a political
meltdown. I suspect that a pension system where (as is increasingly the
case) most private pensions and a substantial part of public pensions are
defined-contribution plans is not politically robust to two or three decades
of plausibly low returns.
In the end, I may not agree with all of the authors’ numbers and analyses.
But I do agree with their central conclusion, slightly restated: Given the
uncertainty about the prospective equity premium, policymakers should be
very hesitant to base major public policy programs such as Social Security
on the continued existence of a large equity premium and high stock returns
in the future. Those who design policies should be able to answer the fol-
lowing questions: What would be the economic results if returns were at the
low end of the plausible range? Who will compensate those who realize
low or negative returns? And how will that compensation occur?

General discussion: Several panelists commented on the substantial uncer-

tainty surrounding forecasts of population, output, and rates of return for the
long horizons relevant to Social Security. Robert Gordon declared himself
highly skeptical of the prediction by the Social Security trustees, adopted by
the authors, of a substantial long-term decline in U.S. economic growth
caused by a decline in both population and productivity growth. In partic-
ular, he doubted the official assumption that immigration will remain con-
stant, so that immigrants are projected to make up a declining share of
 326 Brookings Papers on Economic Activity, 1:2005

the population. According to historical experience, even without illegal

immigration, immigration as a fraction of the population has increased
steadily since about 1955, and the number of immigrants arriving annu-
ally has been growing at a rate between 3 and 4 percent a year. Moreover,
Gordon expected substantial increases in U.S. income and therefore a
steady increase in the demand for immigrants, as well as no shortage of
supply. He noted that the conservative assumption in his Fall 2003 Brook-
ings Paper made a large difference in the outcome. In particular, that paper
shows that, if annual immigration does no more than rise gradually over
the next twenty years from its current 0.4 percent of the population to
0.5 percent, the total U.S. population seventy-five years from now will be
610 million, instead of 425 million as currently projected. The implied
population growth rate is 1 percent rather than 0.3 percent as Baker, DeLong,
and Krugman assume.
Gordon also believed the trustees’ projected decline in productivity puts
too much weight on the dismal years of growth from 1973 to 1995. Dean
Baker agreed, noting that the trustees use average productivity growth in
the last four completed business cycles to make their projections, thus ignor-
ing the recent strong performance. Gordon observed that productivity growth
in the past twenty, fifty-five, or eighty years has been substantially higher
than the official projections. He also noted that, as long as benefits are
indexed to wages, a reduction in productivity growth automatically reduces
Social Security obligations.
Discussion turned to the authors’ predictions of lower rates of return on
equity investment, conditional on slower growth of population and produc-
tivity. Henry Aaron suggested that, with slightly higher assumptions about
immigration and total factor productivity growth and a modestly lower sav-
ing rate, the rate of return would not fall much below its historical average
of about 6.5 percent a year. He did not think this an implausible combina-
tion of events. Olivier Blanchard agreed that there is substantial uncertainty
about future rates of return and that Aaron’s scenario is possible. But he
thought it wise to worry about the adverse tail of the distribution of outcomes.
He viewed the paper as arguing simply that there is a significant risk that
growth rates will be low and that it is precisely in those cases that investment
in equities, whether in the Social Security trust fund or in private accounts,
will be disappointing.
William Brainard supported the life-cycle model as the framework for
analyzing the effect of demographic changes on saving, much preferring
Dean Baker, J. Bradford DeLong, and Paul R. Krugman 327

it to the assumption of infinitely lived families in the Ramsey model. But,

although he found the authors’ use of the two-period life-cycle model clar-
ifying, he wished they had drawn on the extensive literature that examines
more realistic versions of the model and estimates empirically the effects
of changes in birth rates, life expectancy, and immigration on saving.
Martin Baily thought it important to distinguish between the rate of return
on domestic assets and the return on the U.S. stock market. Foreign earnings
are important to U.S. corporations and the U.S. stock market: in the past
ten years or so, U.S. companies have been quite successful at earning high
rates of return overseas, even though the rate of return on foreign capital
in general is lower than the return on capital in the United States. They are
able to do this because of proprietary technology and better management
methods, advantages that, Baily warned, may not persist. He noted that the
other large, mature industrial regions, Japan and Europe, have even lower
birth rates than the United States, as does China. These dramatic demographic
trends overseas are going to play an important role in the return on capital
within a global context.
William Brainard likewise emphasized the distinction between domes-
tic and national capital stocks and the importance of taking into account
likely changes in global saving. In the authors’ analysis, growth in the
domestic labor force affects the rate of return on the domestic capital stock,
including capital owned by foreigners as well as capital owned by U.S.
investors. U.S. investors, on the other hand, hold claims on the returns on
capital abroad, both through their ownership of U.S. multinational firms
and through their ownership of foreign-headquartered firms. Brainard ampli-
fied Baily’s comment about global trends, noting that the reduction in sav-
ing required to maintain the rate of return on capital in the face of reduced
U.S. labor force growth was similar in magnitude to net foreign investment
in the United States today. Gradual elimination of that capital inflow, in the
absence of increases in national saving, would largely avoid the capital
deepening that, in the authors’ analysis, forces down the return to capital.
Aaron agreed with Gregory Mankiw that Social Security solvency and
Social Security privatization are two distinct issues. Aaron argued, however,
that privatization as proposed by the administration would significantly
aggravate the solvency problem, increasing the projected seventy-five-
year deficit by slightly more than a third from the level projected by the
actuaries. Aaron also believed the plan posed significant risks to individu-
als: In the administration’s proposal, what matters far more than the average
 328 Brookings Papers on Economic Activity, 1:2005

long-range rate of return over an individual’s lifetime is the return over the
final years of the individual’s working life. A significant market decline just
before a worker retires or becomes disabled could be devastating to that
worker. Aaron believed acceptance of so great a risk was inconsistent with
the fundamental purpose of social insurance. Benjamin Friedman commented
that the pros and cons of privatization would make for an interesting debate
regardless of whether Social Security has unfunded liabilities. Although
he agreed with Mankiw that the two issues would best be discussed sepa-
rately, he observed that it was President Bush, not the present authors, who
had chosen to confound them, and he regarded Mankiw’s complaint about
the authors linking them as an implicit criticism of the president.
Dean Baker, J. Bradford DeLong, and Paul R. Krugman 329

References
Abel, Andrew B., N. Gregory Mankiw, Lawrence H. Summers, and Richard J.
Zeckhauser. 1989. “Assessing Dynamic Efficiency: Theory and Evidence.”
Review of Economic Studies 56, no. 1: 1–20.
Baker, Dean. 1997. Saving Social Security with Stocks: The Promises Don’t Add
Up. New York: The Century Foundation.
Barberis, Nicholas C., and Richard H. Thaler. 2003. “A Survey of Behavioral
Finance.” In Handbook of the Economics of Finance, edited by George Con-
stantinides, Milton Harris, and Rene Stulz. Amsterdam: Elsevier.
Board of Trustees of the Federal Old Age and Survivors Insurance and Disability
Insurance Trust Funds. 2005. The 2005 Annual Report of the Board of Trustees
of the Federal Old Age and Survivors Insurance and Disability Insurance Trust
Funds. Government Printing Office.
Campbell, John Y., and Robert J. Shiller. 1988. “Stock Prices, Earnings, and
Expected Dividends.” Journal of Finance 43, no. 3: 661–76.
Council of Economic Advisers. 2005. “Three Questions About Social Security”
(www.whitehouse.gov/cea/three-quest-soc-sec.pdf [February 4]).
Cutler, David M., James M. Poterba, Louise M. Sheiner, and Lawrence H. Summers.
1990. “An Aging Society: Opportunity or Challenge?” BPEA, no. 1: 1–56.
Diamond, Peter A. 1965. “National Debt in a Neoclassical Growth Model.” American
Economic Review 55, no. 5: 1126–50.
_________. 2000. “What Stock Market Returns to Expect for the Future?” Social
Security Bulletin 63, no. 2: 38–52.
Edelstein, Michael. 1982. Overseas Investment in the Age of High Imperialism.
Columbia University Press.
Geanakoplos, John, Olivia S. Mitchell, and Stephen P. Zeldes. 1998. “Would a
Privatized Social Security System Really Pay a Higher Rate of Return?” Working
Paper 6713. Cambridge, Mass.: National Bureau of Economic Research (August).
Gordon, Robert. 2003. “Exploding Productivity Growth: Context, Causes, and
Implications.” BPEA, no. 2: 207–79.
Haig, Robert M. 1921. “The Concept of Income—Economic and Legal Aspects.” In
The Federal Income Tax, edited by Robert M. Haig. Columbia University Press.
Hubbard, R. Glenn. 1998. “Capital-Market Imperfections and Investment.” Journal
of Economic Literature 36, no. 1: 193–225.
Kremer, Michael. 1993. “Population Growth and Technological Change: One
Million BC to 1990.” Quarterly Journal of Economics 108, no. 3: 681–716.
Mehra, Rajnish. 2003. “The Equity Premium: Why Is It a Puzzle?” Working Paper
9512. Cambridge, Mass.: National Bureau of Economic Research (February).
Mehra, Rajnish, and Edward C. Prescott. 1985. “The Equity Premium: A Puzzle.”
Journal of Monetary Economics 15 (March): 145–61.
 330 Brookings Papers on Economic Activity, 1:2005

Nordhaus, William D. 2002. “The Recent Recession, the Current Recovery, and
Stock Prices.” BPEA, no. 1: 199–220.
_________. 2005. “The Sources of the Productivity Rebound and the Manufac-
turing Employment Puzzle.” Working Paper 11354. Cambridge, Mass.:
National Bureau of Economic Research (May).
Oliner, Steven D., and Daniel E. Sichel. 2003. “The Resurgence of Growth in the
Late 1990s: Is Information Technology the Story?” Updated version. Washing-
ton: Federal Reserve Board.
Petrick, Kenneth A. 2001. “Comparing NIPA Profits with S&P 500 Profits.” Survey
of Current Business (April): 16–20.
Romer, David. 2000. Advanced Macroeconomics, 2nd ed. McGraw-Hill.
Shiller, Robert J. 2005. “The Life-Cycle Personal Accounts Proposal for Social
Security: A Review.” Working Paper 11300. Cambridge, Mass.: National
Bureau of Economic Research (May).
Siegel, Jeremy. 2005. “Long Term Returns and the Demographic Crisis.” Wash-
ington: American Enterprise Institute (www.aei.org/doclib/20050310_Siegel.pdf
[March 11]).
Solow, Robert. 1956. “A Contribution to the Theory of Economic Growth.”
Quarterly Journal of Economics 70, no. 1: 65–94.