A Pyrrhic Victory?


Bank Bailouts and Sovereign Credit Risk
1
Viral V. Acharya
2
NYU-Stern, CEPR and NBER
Itamar Drechsler
3
NYU-Stern
Philipp Schnabl
4
NYU-Stern and CEPR
J.E.L. Classification: G21, G28, G38, E58, D62.
Keywords: financial crises, forbearance, deleveraging,
sovereign debt, growth, credit default swaps
First draft: August 2010
This draft: May 2011
1
We are grateful to Isabel Schnabel and Luigi Zingales (discussants), Dave Backus, Mike Cher-
nov, Amir Yaron, Stan Zin, and seminar participants at Austrian Central Bank, AEA Meetings
(2011), HEC Paris and BNP Paribas Hedge Fund Center conference, Bundesbank-ECB-CFS Joint
Luncheon workshop, Federal Reserve Bank of Minneapolis, Oxford Said Business School, Univer-
sity of Minnesota, Indian School of Business, Indian Institute of Management (Ahmedabad), and
Douglas Gale’s Financial Economics workshop at NYU Economics for helpful comments. Farhang
Farazmand and Nirupama Kulkarni provided valuable research assistance.
2
Department of Finance, Stern School of Business, New York University, e-mail:
vacharya@stern.nyu.edu
3
Department of Finance, Stern School of Business, New York University, e-mail: ita-
mar.drechsler@stern.nyu.edu.
4
Department of Finance, Stern School of Business, New York University, e-mail: schn-
abl@stern.nyu.edu.
A Pyrrhic Victory? –
Bank Bailouts and Sovereign Credit Risk
Abstract
We show that financial sector bailouts and sovereign credit risk are intimately linked. A
bailout benefits the economy by ameliorating the under-investment problem of the financial
sector. However, increasing taxation of the corporate sector to fund the bailout may be
inefficient since it weakens its incentive to invest, decreasing growth. Instead, the sovereign
may choose to fund the bailout by diluting existing government bondholders, resulting in
a deterioration of the sovereign’s creditworthiness. This deterioration feeds back onto the
financial sector, reducing the value of its guarantees and existing bond holdings and increas-
ing its sensitivity to future sovereign shocks. We provide empirical evidence for this two-way
feedback between financial and sovereign credit risk using data on the credit default swaps
(CDS) of the Eurozone countries for 2007-10. We show that the announcement of financial
sector bailouts was associated with an immediate, unprecedented widening of sovereign CDS
spreads and narrowing of bank CDS spreads; however, post-bailouts there emerged a signif-
icant co-movement between bank CDS and sovereign CDS, even after controlling for banks’
equity performance, the latter being consistent with an effect of the quality of sovereign
guarantees on bank credit risk.
J.E.L. Classification: G21, G28, G38, E58, D62.
Keywords: financial crises, forbearance, deleveraging, sovereign debt, growth, credit de-
fault swaps
1
1 Introduction
Just two and a half years ago, there was essentially no sign of sovereign credit risk in the
developed economies and a prevailing view was that this was unlikely to be a concern for them
in the near future. Recently, however, sovereign credit risk has become a significant problem
for a number of developed countries. In this paper, we are motivated by three closely related
questions surrounding this development. First, were the financial sector bailouts an integral
factor in igniting the rise of sovereign credit risk in the developed economies? We show that
they were. Second, what was the exact mechanism that caused the transmission of risks
between the financial sector and the sovereign? We propose a theoretical explanation wherein
the government can finance a bailout through both increased taxation and via dilution of
existing government debt-holders. The bailout is beneficial; it alleviates a distortion in
the provision of financial services. However, both financing channels are costly. When
the optimal bailout is large, dilution can become a relatively attractive option, leading
to deterioration in the sovereign’s creditworthiness. Finally, we ask whether there is also
feedback going in the other direction– does sovereign credit risk feedback on to the financial
sector? We explain – and verify empirically – that such a feedback is indeed present, due to
the financial sector’s implicit and explicit holdings of sovereign bonds.
Our results call into question the usually implicit assumption that government resources
are vastly deep and that the main problem posed by bailouts is primarily that of moral
hazard – that is, the distortion of future financial sector incentives. While the moral hazard
cost is certainly pertinent, our conclusion is that bailout costs are not just in the future.
They are tangible right around the timing of bailouts and are priced into the sovereign’s
credit risk and cost of borrowing. Thus, aggressive bailout packages that stabilize financial
sectors in the short run but ignore the ultimate taxpayer cost might end up being a Pyrrhic
victory.
Motivation: The case of Irish bailout. On September 30, 2008 the government of
Ireland announced that it had guaranteed all deposits of the six of its biggest banks. The
immediate reaction that grabbed newspaper headlines the next day was whether such a
policy of a full savings guarantee was anti-competitive in the Euro area. However, there was
something deeper manifesting itself in the credit default swap (CDS) markets for purchasing
protection against the sovereign credit risk of Ireland and that of its banks. Figure 1 shows
that while the cost of purchasing such protection on Irish banks – their CDS fee – fell
overnight from around 400 basis points to 150 basis points, the CDS fee for the Government
2
of Ireland’s credit risk rose sharply. Over the next month, this rate more than quadrupled
to over 100 basis points and within six months reached 400 basis points, the starting level of
its financial firms’ CDS. While there was a general deterioration of global economic health
over this period, the event-study response in Figure 1 suggests that the risk of the financial
sector had been substantially transferred to the government balance sheet, a cost that Irish
taxpayers must eventually bear.
Viewed in the Fall of 2010, this cost rose to dizzying heights prompting economists
to wonder if the precise manner in which bank bailouts were awarded have rendered the
financial sector rescue exorbitantly expensive. Just one of the Irish banks, Anglo Irish, has
cost the government up to Euro 25 bln (USD 32 bln), amounting to 11.26% of Ireland’s Gross
Domestic Product (GDP). Ireland’s finance minister Brian Lenihan justified the propping up
of the bank “to ensure that the resolution of debts does not damage Ireland’s international
credit-worthiness and end up costing us even more than we must now pay.” Nevertheless,
rating agencies and credit markets revised Ireland’s ability to pay future debts significantly
downward. The original bailout cost estimate of Euro 90 bln was re-estimated to be 50%
higher and the Irish 10-year bond spread over German bund widened significantly, ultimately
leading to a bailout of Irish government by the stronger Eurozone countries.
1
This episode is not isolated to Ireland though it is perhaps the most striking case. In
fact, a number of Western economies that bailed out their banking sectors in the Fall of
2008 have experienced, in varying magnitudes, similar risk transfer between their financial
sector and government balance-sheets. Our paper develops a theoretical model and provides
empirical evidence that help understand this interesting phenomenon.
Model. Our theoretical model consists of two sectors of the economy – “financial” and
“corporate” (more broadly this includes also the household and other non-financial parts of
the economy), and a government. The two sectors contribute jointly to produce aggregate
output: the corporate sector makes productive investments and the financial sector invests in
intermediation “effort” (e.g., information gathering and capital allocation) that enhance the
return on corporate investments. Both sectors, however, face a potential under-investment
problem. The financial sector is leveraged (in a crisis, it may in fact be insolvent) and under-
invests in its contributions due to the well-known debt overhang problem (Myers, 1977).
For simplicity, the corporate sector is un-levered. However, if the government undertakes a
1
See “Ireland’s banking mess: Money pit – Austerity is not enough to avoid scrutiny by the markets”,
the Economist, Aug 19th 2010; “S&P downgrades Ireland” by Colin Barr, CNNMoney.com, Aug 24th 2010;
and, “Ireland stung by S&P downgrade”, Reuters, Aug 25th, 2010.
3
“bailout” of the financial sector, in other words, makes a transfer from the rest of the economy
that results in a net reduction of the financial sector debt, then the transfer must be funded
in the future (at least in part) through taxation of corporate profits. Such taxation, assumed
to be proportional to corporate sector output, induces the corporate sector to under-invest.
A government that is fully aligned with maximizing the economy’s current and future
output determines the optimal size of the bailout. We show that tax proceeds that can
be used to fund the bailout have, in general, a Laffer curve property, so that the optimal
bailout size and tax rate are interior. The optimal tax rate that the government is willing to
undertake for the bailout is greater when the financial sector’s debt overhang is higher and
its relative contribution (or size) in output of the economy is larger.
In practice, governments fund bailouts in the short run by borrowing or issuing bonds,
which are repaid by future taxation. There are two interesting constraints on the bailout size
that emerge from this observation. One, the greater is the legacy debt of the government, the
lower is its ability to undertake a bailout. This is because the Laffer curve of tax proceeds
leaves less room for the government to increase tax rates for repaying its bailout-related
debt. Second, the announcement of the bailout lowers the price of government debt due to
the anticipated dilution from newly issued debt. Now, if the financial sector of the economy
has assets in place that are in the form of government bonds, the bailout is in fact associated
with some “collateral damage” for the financial sector itself. Illustrating the possibility of
such a two-way feedback is a novel contribution of our model.
To get around these constraints, the government can potentially undertake a strategic
default. Assuming that there are some deadweight costs of such default, for example, due
to international sanctions or from being unable to borrow in debt markets for some time,
we derive the optimal boundary for sovereign default as a function of its legacy debt and
financial sector liabilities. This boundary explains that a heavily-indebted sovereign faced
with a heavily-insolvent financial sector will be forced to “sacrifice its credit rating” to save
the financial sector and at the same time sustain economic growth.
We then extend the model to allow for uncertainty about the realized output growth of
the corporate sector.This introduces a possibility of solvency-based default on government
debt. Interestingly, given the collateral damage channel, an increase in uncertainty about
the sovereign’s economic output not only lowers its own debt values but also increases the
financial sector’s risk of default. This is because the financial sector’s government bond
holdings fall in value, and in an extension of the model, so do the value of the government
guarantees accorded to the financial sector as a form of bailout. In turn, these channels
4
induce a post-bailout co-movement between the financial sector’s credit risk and that of the
sovereign, even though the immediate effect of the bailout is to lower the financial sector’s
credit risk and raise that of the sovereign.
Empirics. Our empirical work analyzes this two-way feedback between the financial sector
and sovereign credit risk. Our analysis focuses mainly on the Western European economies
during the financial crisis of 2007-10.
In our non-parametric analysis, we examine sovereign and bank CDS in the period from
2007 to 2010 and find three distinct periods. The first period covers the start of the finan-
cial crisis in January 2007 until the bankruptcy of Lehman Brothers. Across all Western
economies, we see a large, sustained rise in bank CDS as the financial crisis develops. How-
ever, sovereign CDS spreads remains very low. This evidence is consistent with a significant
increase in the default risk of the banking sector with little effect on sovereigns in the pre-
bailout period.
The second period covers the bank bailouts starting with the announcement of a bailout
in Ireland in late September 2008 and ending with a bailout in Sweden in late October 2008.
During this one-month period, we find a significant decline in bank CDS across all countries
and a corresponding increase in sovereign CDS. This evidence suggests that bank bailouts
produced a transfer of default risk from the banking sector to the sovereign.
The third period covers the period after the bank bailouts and until 2010. We find that
both sovereign and bank CDS increased during this period and that the increase was larger
for countries with significant public debt ratios. This evidence suggests that the banks and
sovereigns share the default risk after the announcement of banks bailouts and that the risk
is increasing in the relative size of countries’ public debt.
We confirm all of the above results also in our regression analysis linking levels and
changes in financial sector CDS to levels and changes, respectively, of sovereign CDS in
the three periods. We also confirm model’s implications that banks with higher leverage
experience a stronger relationship between sovereign and bank credit risks after the bailouts.
Finally, in support of the collateral damage channel as being potentially relevant for
the co-movement between financial sector and sovereign CDS, we collect bank-level data on
holdings of different sovereign government bonds released as part of the bank stress tests
conducted for European banks in 2010. We document that on average Eurozone banks
stress-tested in 2010 held Eurozone government bond holdings that were as large as one-
sixth of their risk-weighted assets. Consistent with the economic importance of the collateral
damage channel, we show that bank CDS co-move with sovereign CDS on banks’ holdings
5
of government bonds.
The remainder of the paper is organized as follows. Section 2 presents our theoretical
analysis. Section 4 provides empirical evidence. Section 5 discusses the related literature.
Section 6 concludes. All proofs not in the main text are in the online Appendix.
2 Model
There are three time periods in the model: t = 0, 1, and 2. The productive economy consists
of two parts, a financial sector and a non-financial sector. In addition, there is a government
and a representative consumer. All agents are risk-neutral.
Financial sector: The operator of the financial sector solves the following problem, which
is to choose, at t = 0, the amount of financial services to supply at t = 1 in order to maximize
his expected payoff at t = 1, net of the effort cost required to produce these services:
max
s
s
0
E
0
__
w
s
s
s
0
−L
1
+
˜
A
1
+A
G
+T
0
_
×1
{−L
1
+
˜
A
1
+A
G
+T
0
>0}
_
−c(s
s
0
) . (1)
where s
s
0
is the amount of financial services supplied by the financial sector at t = 1. The
financial sector earns revenues at the rate of w
s
per unit of financial service supplied, with w
s
determined in equilibrium. To produce s
0
units, the operator of the financial sector expends
c(s
0
) units of effort. We assume that c

(s
0
) > 0 and c

(s
0
) > 0.
The financial sector has both liabilities and assets on its books. It receives the payoff
from its efforts only if the value of assets exceeds liabilities at t = 1. This solvency condition
is given in equation (1) by the indicator function for the expression −L
1
+
˜
A
1
+A
G
+T
0
> 0.
L
1
denotes the liabilities of the financial sector, which are due (mature) at t = 1. There
are two types of assets held by the financial sector, denoted
˜
A
1
and A
G
. A
G
is the value of
the financial sector’s holdings of a fraction k
A
of outstanding government bonds, while
˜
A
1
represents the payoff of the other assets held by the financial sector.
2
We model the payoff
˜
A
1
, which is risky, as a continuously valued random variable that is realized at t = 1 and
takes values in [0, ∞). The payoff and value of government bonds is discussed below. Finally,
the variable T
0
represents the value of the time 0 transfer made by the government to the
financial sector and is discussed further below.
2
While we refer to government claims principally as government bonds, a broader interpretation can
include claims on quasi-governmental agencies (e.g., Fannie Mae, Freddie Mac) and perhaps also the value
of explicit and implicit government guarantees or support.
6
In case of insolvency, debtholders receive ownership of all financial sector assets and wage
revenue.
3
Non-financial sector: The non-financial sector comes into t = 0 with an existing capital
stock K
0
. Its objective is to maximize the sum of the expected values of its net payoffs,
which occur at t = 1 and t = 2:
max
s
d
0
, K
1
E
0
_
f(K
0
, s
d
0
) −w
s
s
d
0
+ (1 −θ
0
)
˜
V (K
1
) −(K
1
−K
0
)
_
(2)
The function f is the production function of the non-financial sector, which takes as inputs at
t = 0 the financial services it demands, s
d
0
, and the capital stock, K
0
, to produce consumption
goods at t = 1. The output of f is deterministic. Moreover, we assume that f is increasing
in both arguments and concave. At t = 1, the non-financial sector is faced with a decision of
how much capital K
1
to invest, at an incremental cost of (K
1
−K
0
), in a project
˜
V , whose
payoff is realized at t = 2. This project represents the future or continuation value of the
non-financial sector and is in general subject to uncertainty. The expectation at t = 1 of
this payoff is V (K
1
) = E
1
[
˜
V (K
1
)] and, as indicated, is a function of the investment K
1
.
We assume that V

(K
1
) > 0 and V

(K
1
) < 0, so that the expected payoff is increasing but
concave in investment. A proportion θ
0
of the payoff of the continuation project is taxed by
the government to pay its debt, both new and outstanding, as we explain next.
Government: The government’s objective is to maximize the total output of the economy
and hence the welfare of the consumer. It does this by reducing the debt-overhang problem
of the financial sector, which induces it to supply more financial services, thereby increasing
output. To achieve this, the government issues bonds, that it then transfers to the balance
sheet of the financial sector. These bonds are repaid with taxes levied on the non-financial
sector at a tax-rate of θ
0
. In particular, the tax rate θ
0
is set by the government at t = 0
and is levied at t = 2 upon realization of the payoff
˜
V (K
1
). We assume that the government
credibly commits to this tax rate.
We let N
D
denote the number of bonds that the government has issued in the past –
its outstanding stock of debt. For simplicity, bonds have a face value of one, so the face
value of outstanding debt equals the number of bonds, N
D
. The government issues N
T
new
3
Note that we could include the wage revenues in the solvency indicator function, which would provide
an additional channel for wages to feed back into the probability of solvency. Although such a channel
would reinforce the mechanism at work in the model, we choose to abstract from this to avoid the additional
complexity.
7
bonds to accomplish the transfer to the financial sector. Hence, at t = 2 the government
receives realized taxes equal to θ
0
˜
V (K
1
) and then uses them to pay bondholders N
T
+ N
D
.
We assume that if there are still tax revenues left over (a surplus), the government spends
them on programs for the representative consumer, or equivalently, just rebates them to the
consumer. On the other hand, if tax revenues fall short of N
T
+ N
D
, then the government
defaults on its debt. In that case, it pays out all the tax revenue raised to bondholders. We
assume that the government credibly commits to this payout policy.
4
We further assume that default incurs a deadweight loss. In case of default, the sovereign
incurs a fixed deadweight loss of D. Hence, default is costly and there is an incentive to
avoid it.
5
Finally, let P
0
denotes the price of government bonds, which is determined in
equilibrium. This implies that the value of the financial sector’s holding of government
bonds is A
G
= k
A
P
0
N
D
.
The government’s objective is to maximize the expected utility of the representative
consumer, who consumes the combined output of the financial and non-financial sector.
Hence, the government faces the following problem:
max
θ
0
, N
T
E
0
_
f(K
0
, s
0
) +
˜
V (K
1
) −c(s
0
) −(K
1
−K
0
) −1
def
D +
˜
A
1
_
(3)
where s
0
is the equilibrium provision of financial services. This maximization is subject to
the budget constraint: T
0
= P
0
N
T
and subject to the choices made by the financial and
non-financial sectors. Note that 1
def
is an indicator function that equals 1 if the government
defaults (if θ
0
˜
V (K
1
) < N
T
+N
D
) and 0 otherwise.
Consumer: The representative consumer consumes the output of the economy. He solves a
simple consumption and portfolio choice problem by allocating his wealth W between con-
sumption, government bond holdings, and equity in the financial and non-financial sectors.
Since the representative consumer is assumed to be risk-neutral, asset prices equal the ex-
pected values of asset payouts. Let P(i) and
˜
P(i) denote the price and payoff of asset i,
respectively. At t = 0, the consumer chooses optimal portfolio allocations, {n
i
}, that solve
4
We can consider more generally a policy whereby the government pays only a fraction ˜ m of tax revenue
and gives back any remaining tax funds to the consumer. Since it will actually be optimal for the government
to commit to paying bondholders all the tax revenue raised, we restrict ourselves to the case ˜ m = 1.
5
Although D here is obviously reduced-form, one can think of the deadweight cost in terms of loss
of government reputation internationally, loss of domestic government credibility, degradation of the legal
system and so forth. If a country’s reputation is already weak, it will have less to lose from default.
8
the following problem:
max
n
i
E
0
_
Σ
i
n
i
˜
P(i) + (W −Σ
i
n
i
P(i))
_
(4)
The consumer’s first order condition gives the standard result that equilibrium price of an
asset equals the expected value of its payoff, P(i) = E
0
[
˜
P(i)].
3 Equilibrium Outcomes
We begin by examining the maximization problem of the financial sector. Let p(
˜
A) denote
the probability density of
˜
A. Furthermore, let A
1
be the minimum realization of
˜
A
1
for which
the financial sector does not default: A
1
= L
1
−A
G
−T
0
. Then, the first order condition of
the financial sector can be written as:
w
s
p
solv
−c

(s
s
0
) = 0 . (5)
where p
solv

_

A
1
p(
˜
A
1
)d
˜
A, is the probability that the financial sector is solvent at t = 1.
We assume that at the optimal ˆ s
s
0
the first-order condition is satisfied. The second-order
condition of the financial sector’s problem is −c

(s
s
0
) < 0. The parametric choice we will use
below for c(s
0
) is c(s
0
) = β
1
m
s
m
0
where m > 1.
Consider now the problem of the non-financial sector at t = 0. Its demand for financial
services, ˆ s
d
0
, is determined by its first-order condition:
∂f(K
0
, s
d
0
)
∂s
d
0
= w
s
. (6)
Since f is concave in its arguments, the second order condition is satisfied:

2
f(K
0
,s
d
0
)

2
s
d
0
< 0.
Henceforth, we will parameterize f as Cobb-Douglas with the factor share of financial services
given by ϑ: f(K
0
, s
0
) = αK
1−ϑ
0
s
ϑ
0
.
In equilibrium the demand and supply of financial services are the same: ˆ s
d
0
= ˆ s
s
0
. From
here on, we drop the superscripts on s
0
and denote the equilibrium quantity of financial
services simply by s
0
.
9
3.1 Transfer Reduces Underprovision of Financial Services
Taken together, the first-order conditions of the financial sector (5) and non-financial sector
(6) show how debt-overhang impacts the provision of financial services by the financial sector.
The marginal benefit of an extra unit of financial services to the economy is given by w
s
,
while the marginal cost, c

(s
0
), is less than w
s
if there is a positive probability of insolvency.
This implies that the equilibrium allocation is sub-optimal. The reason is that the possibility
of liquidation p
solv
< 1 drives a wedge between the social and private marginal benefit of an
increase in the provision of financial services. The result is that as long as p
solv
< 1, there is
an under-provision of financial services relative to the first-best case (p
solv
= 1). Hence, we
obtain that
Lemma 1. An increase in the transfer T
0
leads to an increase in the provision of financial
services since this raises the probability p
solv
that the financial sector is solvent at t = 1.
3.2 Tax Revenues: A Laffer Curve
Next, to understand the government’s problem, we first look at how expected tax revenue
responds to the tax rate, θ
0
. Let the expected tax revenue, θ
0
V (K
1
), be denoted by T .
Raising taxes has two effects. On the one hand, an increase in the tax rate θ
0
captures a
larger proportion of the future value of the non-financial sector, thereby raising tax revenues.
On the other hand, this reduces the incentive of the non-financial sector to invest in its future,
thereby leading to reduced investment, K
1
. At the extreme, when θ
0
= 1, the tax distortion
eliminates the incentive for investment and tax revenues are reduced to zero. Hence, tax
revenues are non-monotonic in the tax rate and maximized by a tax rate strictly less than 1.
Formally, the impact on tax revenue of an increase in the tax rate is given by:
∂T
∂θ
0
= V (K
1
) +θ
0
V

(K
1
)
dK
1

0
Note that at θ
0
= 0, an increase in the tax rate increases the tax revenue at a rate equal to
V (K
1
), the future value of the non-financial sector. It can be shown that since the production
function V (K
1
) is concave, as taxes are increased the incentive to invest is decreased by the
tax rate, which reduces the marginal revenue of a tax increase. This is given by the second
term on the right-hand side of the expression. To see this, consider the first-order condition
10
for investment of the non-financial sector at t = 1:
(1 −θ
0
)V

(K
1
) −1 = 0 (7)
Since V

(K
1
) < 0, the second-order condition holds. Taking the derivative with respect to
θ
0
by using the Implicit Function theorem gives:
dK
1

0
=
V

(K
1
)
(1 −θ
0
)V

(K
1
)
< 0
which shows that as the tax rate is increased, the non-financial sector reduces investment. In
fact, since we know that at θ
0
= 1 the tax revenue is zero, it must be the case that as the tax
rate is increased, the marginal tax revenue decreases until it eventually becomes negative.
To summarize, tax revenues satisfy the Laffer curve property as a function of the tax rate:
Lemma 2. The tax revenues, θ
0
V (K
1
), increase in the tax rate, θ
0
, as it increases from zero
(no taxes), and then eventually decline.
Henceforth, we parameterize V with the functional form V (K
1
) = K
γ
1
, 0 < γ < 1.
6
As
Appendix A.3 shows, (7) implies that T = θ
t+1
γ
γ
1−γ
(1−θ
t+1
)
γ
1−γ
. It can then be shown that:
Lemma 3. The tax revenue, T , is maximized at θ
max
0
= 1 −γ, is increasing (dT /dθ
0
> 0)
and concave (d
2
T /dθ
2
0
< 0) on [0, θ
max
0
), and decreasing (dT /dθ
0
< 0) on (θ
max
0
, 1).
3.3 Optimal Transfer Under Certainty and No Default
We analyze next the government’s decision starting first with a simplified version of the
general setup. We make two simplifying assumptions: (1) we set to zero the variance of the
realized future value of the non-financial sector, so that
˜
V (K
1
) = V (K
1
), (2) we force the
government to remain solvent. In subsequent sections we remove these assumptions.
If the government must remain solvent, it can only issue a number of bonds N
T
that
it can pay off in full, given its tax revenue. By assumption (1), the tax revenue is known
exactly (it is equal to T ), and hence by assumption (2), N
T
+ N
D
= T . Moreover, since
every bond has a sure payoff of 1, we know that the bond price is P
0
= 1.
6
This functional form is a natural choice for an increasing and concave function of K
1
. Appendix A.2
provides a more structural motivation for this choice based on the calculation of a continuation value un-
der our choice of production function. This calculation suggests that the continuation value implied by a
multiperiod model should take a similar functional form.
11
Under the two simplifying assumptions, we have that the transfer to the financial sector
is T
0
= θ
0
V (K
1
) − N
D
and there is no probability of default, E[1
def
] = 0. Hence, the only
choice variable for the government in this case is the tax rate. Since there is no change in the
non-financial sector’s investment opportunities between t = 0 and t = 1, the government’s
information regarding expected tax revenue is the same at t = 0 as at t = 1, and we can
consider the problem directly at t = 0. Appendix A.4 shows that the first-order condition
for the government can be written as:
dG
dT
+
dL
dT
= 0 (8)
where
dG
dT
=
∂f(K
0
, s
0
)
∂s
0
(1 −p
solv
)
ds
0
dT
0
dL
dT

0
V

(K
1
)
dK
1
dT
which expresses the first-order condition in terms of the choice of transfer size and expected
tax revenue, rather than in terms of the tax rate. As we explain below, this condition is
intuitive since it equates the marginal gain and marginal loss of increasing tax revenue.
Gain From Increased Provision of Financial Services: The term dG/dT in (8) is
the marginal gain to the economy of increasing expected tax revenue. Equations (5) and
(6) show that it reflects the wedge between the social and private benefits of an increase in
financial services. Since, as shown above, ds
0
/dT
0
> 0, the marginal gain from increasing tax
revenue (and hence the transfer) will be large when p
solv
is low, that is, when the financial
sector is at high risk of insolvency and debt-overhang is significant. For illustration, the
graph of dG/dT is given by the solid curve in the top panel of Figure 2.
7
Under-Investment Loss Due to Taxes: The term dL/dT in (8) is the marginal under-
investment loss to the economy due to a marginal increase in expected tax revenue, which
distorts the non-financial sector’s incentive to invest. Equation (7) shows that it reflects
the size of the tax-induced distortion, which is greater than zero as long as the tax rate is
positive. Since dK
1
/dT < 0, then dL/dT < 0. For illustration, a graph of −dL/dT is shown
as the upward-sloping dashed curve in the top panel of Figure 2.
8
7
As the graph indicates, and as the proof to Proposition 1 shows, G is concave in T since the marginal
gain from increasing tax revenues (to increase the transfer) is decreasing.
8
The curve shows that −L is convex as raising additional tax revenues incurs an increasingly large marginal
12
The Optimal Tax Rate and Issuance of Debt: The following proposition, which de-
scribes the solution to the government’s problem under assumptions 1 (certainty), 2 (no
default) and m ≥ 2ϑ, is proven in Appendix A.6.
Proposition 1. There is a unique optimal tax rate,
ˆ
θ
0
, which is strictly less than θ
max
0
. Let
ˆ
T represent the associated tax revenues. Then newly issued sovereign debt has face value
N
T
=
ˆ
T −N
D
and a price of P
0
= 1. Moreover,
1. The optimal tax rate and revenue are increasing in L
1
, the financial sector liabilities,
and in N
D
, the outstanding government debt.
2. The face value of newly issued sovereign debt (the transfer) is increasing in the fi-
nancial sector liabilities L
1
, but decreasing in the amount of existing government debt
N
D
. Moreover, the gross transfer, T
0
+k
A
N
D
, is also decreasing in N
D
.
3. If also m ≤ 2, then the optimal tax rate, revenue, and newly issued sovereign debt, are
increasing in the factor share of the financial sector.
The optimal tax rate is less than θ
max
0
due to the Laffer-curve property of tax revenues,
whereby the marginal underinvestment loss induced by raising revenue becomes infinite as
the tax rate rises to θ
max
0
. In addition, if there is any debt-overhang (i.e., p
solv
< 1), then
the optimal tax rate will be strictly greater than zero, since at a zero tax rate there is a
marginal benefit to having a transfer but no marginal cost.
The optimal government action is to increase the transfer, by increasing tax revenue and
outstanding debt, until the marginal gain from the transfer no longer exceeds the associated
marginal loss due to underinvestment. For illustration, in the top panel of Figure 2, this
is at the intersection point of the two curves, the x-coordinate of which represents
ˆ
T . The
bottom panel of Figure 2 graphs the value of the government’s objective function, whose
slope is given by (8). As the graph illustrates, the objective is concave in T and the unique
optimum occurs at
ˆ
T , corresponding to the intersection point in the top panel.
Next, consider the three parts of Proposition 1.
For any level of transfer, the marginal gain available is greater the more severe is the
debt-overhang, since a lower probability of solvency increases the distortion in the provision
of financial services. This represents an upward shift in the marginal gain curve. Therefore,
underinvestment loss. Furthermore, as shown in Appendix A.5, this marginal loss due to underinvestment
worsens as T is increased, i.e., d
2
L/dT
2
< 0.
13
as (1) and (2) of Proposition 1 state, an increase in L
1
, the financial sector liabilities, leads
to a higher tax rate, more tax revenue, and greater issuance of new sovereign debt to fund
a larger transfer.
If the level of pre-existing government debt (N
D
) is increased, there is again an upward
shift in the marginal gain curve, as shown by the dash-dot curve in the top panel of Figure
2. The reason is that for any level of tax revenue, the effective transfer (T
0
) is smaller, and
therefore the probability of solvency is lower. As is clear from the new intersection point in
the top panel, and as (1) of Proposition 1 states, this pushes the government to increase the
optimal tax rate, tax revenue, and overall amount of sovereign debt.
However, as (2) of Proposition 1 shows, the rate of increase in total sovereign debt is less
than the increase in N
D
. Hence, under the no-default and certainty assumptions, an increase
in pre-existing government debt corresponds to a decrease in newly issued sovereign debt
and a smaller transfer T
0
. The reason for this decrease is that the underinvestment cost of
raising additional tax revenues is increasing.
9
Finally, Proposition 1 shows that, ceteris paribus, a larger factor share of the financial
sector in aggregate production implies that the government will issue a greater amount of
new debt and a larger transfer. Intuitively, if the financial sector’s output is a more important
input into production, then there will be a greater marginal gain from an increase in the
provision of financial services due to the transfer.
10
3.4 Default Under Certainty
Now we allow the government to deviate from the no-default choice of setting N
T
= T −N
D
.
Increasing N
T
above this threshold has both an associated cost and benefit. The benefit is
that this can increase the transfer to the financial sector. Recall that the transfer T
0
equals
P
0
N
T
, where P
0
= max(1, T /(N
T
+ N
D
)) is the price of the government bond. The cost is
that when N
T
> T −N
D
, the government will not be able to fully cover its obligations. In
that case, P
0
< 1 and the government will default, triggering the dead-weight loss of D.
Hence, the government’s decision on how many new bonds to issue, N
T
, splits the pa-
rameter space into two regions:
1. N
T
= T −N
D
and 1
def
= 0 (No Default)
9
Later, we show that the possibility of default or the introduction of uncertainty can alter this result.
10
The factor share is given by ϑ. Note, however, that the comparative static is not simply to vary ϑ, but
to hold total output constant while doing so. Equivalently, we may think about comparing the ratio to total
output of our variable of interest while varying the factor share.
14
2. N
T
> T −N
D
and 1
def
= 1 (Default)
As shown in Appendix A.7, if the choice to default is made, then it is optimal for the
government to issue an infinite amount of new debt in order to fully dilute existing debt (P
0
becomes 0) and hence capture all tax revenues towards the transfer. The resulting situation
is the same as if pre-existing debt N
D
had been set to zero. Therefore, to determine whether
defaulting is optimal, the government evaluates whether its objective function for given N
D
and no default exceeds by at least D (the deadweight default cost) its objective function
with N
D
set to zero.
Formally, let W
no def
denote the maximum value of the government’s objective function
conditional on no-default, W
def
denote the maximum value conditional on default, and W =
max(W
no def
, W
def
). The following lemma characterizes the optimal government action and
resulting equilibrium:
Lemma 4. Conditional on default, it is optimal to set N
T
→ ∞ (and hence P
0
→ 0).
This implies that W
def
= W
no def
¸
¸
N
D
=0
−D. Moreover, if default is undertaken then (1) the
optimal tax rate is lower,
ˆ
θ
def
0
<
ˆ
θ
no def
0
(2) provided that k
A
N
D
<
ˆ
T
def
, the gross transfer
is bigger,
ˆ
T
0
def
>
ˆ
T
0
no def
+ k
A
N
D
, and (3) equilibrium provision of financial services is
higher, ˆ s
0
def
> ˆ s
0
no def
.
Appendix A.7 proves the lemma. We next consider the factors that push the sovereign
towards default.
3.4.1 Default Boundary
Figure 3 displays the optimal default boundary in L
1
×N
D
space along with the No-Default
and Default regions. The following proposition characterizes how a number of factors move
the ‘location’ of the sovereign relative to the default-boundary, or shift the default-boundary
itself.
Proposition 2. Ceteris paribus, the benefit to defaulting is:
1. increasing in the financial sector liabilities L
1
(severity of debt-overhang), the amount
of existing government debt N
D
, and in the factor share of the financial sector
2. decreasing in the dead-weight default cost D, and in the fraction of existing govern-
ment debt held by the financial sector k
A
15
Appendix A.8 provides the proof. Consider a worsening of the financial sector’s health,
leading to a decreased provision of financial services. This increases the marginal gain from
further government transfer, and, in turn, increases the gain to the sovereign from defaulting.
This is represented by a move towards the right in Figure 3, decreasing the distance to the
default-boundary. A similar kind of result holds if the factor share of financial services is
increased, since the marginal gain of further transfer is higher at every level of transfer.
An increase in existing debt implies a bigger spread between the optimal transfer and
optimal tax revenue with and without default. Both the extra transfer and decreased un-
derinvestment represent benefits to defaulting. This is represented by a move upwards in
Figure 3, again decreasing the distance to the default-boundary.
11
It is clear that an increase in the deadweight loss raises the threshold for default. If the
sovereign has a lot to lose from defaulting (think a sovereign with strong domestic credibility
or international reputation) then the net benefit to default will be relatively lower.
Finally, and importantly, an increase in the fraction of existing sovereign debt held by
the financial sector also raises the threshold for default since the act of defaulting, which is
aimed at freeing up resources towards the transfer, causes collateral damage to the financial
sector balance sheet. From the vantage point of Figure 3, both an increase in D and k cause
an outwards shift in the default boundary.
3.4.2 Two-way Feedback
Propositions 1 and 2 indicate that there is a two-way feedback between the solvency situation
of the financial sector and of the sovereign. First, by Proposition 1, a severe deterioration
in the financial sector’s probability of solvency (e.g., an increase in L
1
) leads to a large
expansion in new debt (N
T
) by the sovereign, as it acts to mitigate the under-provision of
financial services. As the marginal cost of raising the tax revenue (dL/dT ) to fund this debt
expansion is increasing, the sovereign is pushed closer to the decision to default (Proposition
2), as well as is its maximum debt capacity (lemma 3). Hence, a financial sector crisis pushes
the sovereign towards distress.
Going in the other direction, by Proposition 1, a distressed sovereign, e.g., one with high
existing debt (N
D
), will have a financial sector with a worse solvency situation. This is
because it is very costly for such a sovereign to fund increased debt to make the transfer to
the financial sector. Hence, a more distressed sovereign will tend to correspond to a more
11
Moreover, since the marginal loss from funding extra debt is increasing, this benefit is convex in N
D
.
16
distressed financial sector (lower post-transfer p
solv
). Strategically defaulting is an avenue for
a distressed sovereign to free debt capacity for additional transfer. However, large holdings
of sovereign debt (k) by the financial sector means that taking this avenue simultaneously
causes collateral damage to the balance sheet of the financial sector, limiting the benefit
from this option (Proposition 2). In this case, a distressed sovereign is further incapacitated
in its ability to strengthen the solvency of its financial sector.
3.5 Uncertainty, Default, and Pricing
We now introduce uncertainty about future output (i.e., growth) by allowing the variance
of
˜
V (K
1
) to be nonzero. Instead of a binary default vs. no-default decision, the government
now implicitly chooses a continuous probability of default when it sets the tax rate and new
debt-issuance. In this case, if raising taxes further incurs a large under-investment loss, the
government can choose to increase debt issuance while holding the tax rate constant. This
dilutes the claim of existing bondholders to tax revenues, thereby generating a larger transfer
without inducing further underinvestment. The trade-off is an increase in the government’s
probability of default and expected dead-weight default loss. In this case, the sovereign
effectively ‘sacrifices’ its own creditworthiness to improve the solvency of the financial sector,
leading to a ‘spillover’ of the financial sector crisis onto the solvency of the sovereign.
Although θ
0
and N
T
, are the variables the government directly chooses, it will turn out
to be more enlightening to look at two other variables that map to them in a one-to-one
fashion. The first variable is T , which again equals θ
0
V (K
1
), the expected tax revenue. The
second variable is:
H =
N
T
+N
D
T
In words, H is the ratio of outstanding debt to expected tax revenue. It measures the
sovereign’s ability to cover its total debt at face value. The government’s problem (3) then
is equivalent to optimally choosing T and H.
12
Note that the no-default and total-default
cases under certainty correspond to setting H = 1 and H →∞, respectively.
To represent uncertainty we write
˜
V (K
1
) = V (K
1
)
˜
R
V
, where
˜
R
V
≥ 0 represents the
shock to
˜
V (K
1
). By construction, E[
˜
R
V
] = 1. We also assume that the distribution of
˜
R
V
12
Formally, the mapping from θ
0
to T is invertible on [0, θ
max
0
] (as before, we can limit our concern to
this region) and given T , the mapping from H to N
T
is invertible. Hence, these alternative control variables
map uniquely to the original ones on the region of interest.
17
is independent of the variables K
1
, θ
0
, and N
T
.
Pricing, Default Probability and the Transfer: Using H we can easily express the
sovereign’s bond price, P
0
, and probability of default, p
def
, as follows:
P
0
= E
0
_
min
_
1,
θ
0
˜
V (K
1
)
N
T
+N
D
__
= E
0
_
min
_
1,
1
H
˜
R
V
__
(9)
p
def
= prob
_
θ
0
˜
V (K
1
) < N
T
+N
D
_
= prob
_
˜
R
V
< H
_
(10)
Note that these quantities depend only on H and do not directly change with T . Next, as
N
T
= (T −N
D
/H)H, we express the transfer in terms of T and H:
T
0
= N
T
P
0
= (T −
N
D
H
)E
0
_
min
_
H,
˜
R
V
__
(11)
The Optimal Probability of Default: Appendix A.9 and A.10 derive the first-order
conditions for T and H, respectively. The first-order condition for T involves the same
transfer-underinvestment trade-off as under certainty, adjusted to account for H. Varying H
involves a new trade-off. As (9)–(11) show, increasing H while holding T constant increases
the transfer, but also increases the probability of sovereign default and decreases the sovereign
bond price. Intuitively, raising H increases the transfer by diluting existing bondholders–it
raises outstanding debt but without increasing expected tax revenue. This captures a greater
faction of tax revenues towards the transfer but raises the probability of default.
The top panel of Figure 4 illustrates the marginal gain (solid line) and loss (dashed line)
incurred by increasing H for a fixed level of T . The dash-dot line represents the marginal
gain curve at a higher level of L
1
than for the solid line. To generate the plots, we need
to assume a specific distribution for uncertainty. For simplicity, we let
˜
R
V
have a uniform
distribution. As the figure shows, and as proven in Appendix A.10, the marginal gain curve
is downwards sloping.This is because, as H increases, the marginal increase in the transfer
due to further dilution decreases. The marginal cost of an increase in H is the rise in
expected dead-weight default cost. This is shown by the dashed green line in Figure 4. For
˜
R
V
uniformly distributed, this cost is a flat function of H until the upper end of the support
of the distribution, falling to zero beyond this upper point. Raising H beyond this point
represents sure default (pdef = 1).
Figure 4 indicates that (with T held constant) there are two potential candidates for the
18
optimal choice of H. The first is the value of H at which the gain and loss curves intersect.
The second is to let H → ∞, representing a total default and full dilution of existing
bondholders. The bottom panel of Figure 4 plots the corresponding value of the government’s
objective as a function of H. The plot shows that for the configuration displayed, a relatively
small value of H achieves the optimum, which is at the intersection of the gain and loss curves
in the top panel. As this optimal H is beyond the left end of the support of
˜
R
V
(which is
the origin in the figure), it corresponds to an optimal non-zero probability of default. Note
that beyond the upper end of the support of
˜
R
V
, the objective function again rises in H.
Th reason for this is that, once debt issuance is large enough that default is certain, it is
optimal for the government to fully dilute existing bondholders to obtain the largest possible
transfer.
Finally, the dash-dot curves in Figure 4 correspond to an increase in L
1
(more severe
debt-overhang) relative to the solid lines. As indicated, this worsening of financial sector
solvency increases the marginal gain from an increase in H, pushing up the marginal gain
curve in the top panel. In the bottom panel, it pushes down the curve because it lowers
overall welfare. As is apparent in the top panel, the optimal response of the sovereign is
to increase the optimal H (by issuing more debt) in order to increase the transfer. This
comes at the cost to the sovereign of a further increase in the probability of default. From
the bottom panel it is apparent that while total default is still suboptimal, it would become
optimal for a sufficiently high level of L
1
.
3.5.1 Comparative Statics Under Uncertainty
The government jointly chooses T and H in an optimal way by comparing the relative
marginal cost and benefit of adjusting each quantity. When the marginal benefit of additional
transfer is large, but the marginal underinvestment loss due to taxation is already high, the
government will optimally begin to ‘sacrifice’ its own creditworthiness to generate additional
transfer. Therefore, a high L
1
(i.e., a financial sector crises) will be associated with both a
high T and a high level of H, up to the point where total default becomes optimal. The
following proposition characterizes how different factors impact
ˆ
H and
ˆ
T , the government’s
optimal choices of H and T in equilibrium:
Proposition 3. If (
ˆ
T ,
ˆ
H) is an interior solution to the government’s problem on a region
of the parameter space, then
ˆ
H is increasing in L
1
, in N
D
and in ϑ, and decreasing in
D. Furthermore,
ˆ
T is also increasing in L
1
.
19
To obtain more precise results we choose a specific distribution for
˜
R
V
. For simplicity,
we again let
˜
R
V
have a uniform distribution. We then plot in Figure 5 comparative statics
of the equilibrium (optimal) values of T , H, T
0
, and P
0
as L
1
and N
D
are varied. The
discontinuities that appears in the plots, as indicated by the dotted lines, represent the
point at which total default becomes optimal.
The top panel of Figure 5 is for L
1
. It shows that T increases monotonically in L
1
,
up to the point where the sovereign chooses total default. The corresponding plot for H
tells a different story. For low levels of L
1
, H is held constant at a low value. This value
corresponds to the lower end of the support of
˜
R
V
, so the probability of sovereign default
remains 0. Correspondingly, the plot shows that in this range, P
0
remains fully valued at
1. For sufficiently high L
1
(e.g., financial crisis), the government chooses to increase H. As
discussed above, it ‘sacrifices’ its own creditworthiness in order to achieve a larger transfer.
The increase in the transfer is apparent in the subplot for T
0
, while the damage to the
sovereign’s creditworthiness is apparent in the plot for P
0
, which begins to decrease once H
begins to rise.
13
.
The plots also show that, when the financial sector’s situation is severe enough (L
1
is
large), the optimal government response can be a total default. The outcome of a total
default is illustrated in the plots at the point of the dotted line. As in the certainty case,
total default fully dilutes existing bondholders, freeing extra capacity for the sovereign to
generate the transfer. As indicated in the plots, this leads to a jump up in T
0
and a jump
down in T . At the same time, P
0
drops to 0.
The bottom panel of Figure 5 shows the comparative statics for N
D
. It is apparent that
for low levels of debt the sovereign keeps H constant at the low end of its support, so there
is no probability of default and P
0
remains at 1. For these values of N
D
, the government
funds the transfer exclusively through increases in tax revenues. Note that in this range the
transfer is decreasing in N
D
, similar to the case of certainty. Once N
D
is sufficiently high, the
underinvestment costs of increasing tax revenue becomes very high and again the sovereign
begins to increase H to fund the transfer. Consequently, the probability of default rises and
P
0
begins to decrease, as shown in the plot. Interestingly, in this range the combination of
increased H and T imply that the transfer is actually increasing in N
D
. The reason for this
is that for large N
D
, the dilution of existing bondholders is an effective channel for increasing
the transfer. Moreover, as the plots show, at high enough N
D
, total default becomes optimal.
13
Note that the transfer increases more rapidly in L
1
once H begins to increase. This occurs because
higher H means a a greater proportion of expected tax revenue is captured towards the transfer
20
3.6 Government ‘Guarantees’
Government guarantees of financial sector debt have been an explicit part of a number of
countries’ financial sector bailouts, notably Ireland. Moreover, it has been common for
sovereigns to step in to prevent the liquidation of banks by guaranteeing their debt, which
strongly suggests that there is an implicit ‘safety net’.
14
.
In this section, we add to the model a simple notion of a government guarantee of financial
sector debt. We do this for two reasons. First, guarantees are a measure that serves to prevent
liquidation of the financial sector by debtholders, which is necessary pre-condition for the
sovereign to act to alleviate debt-overhang and increase the provision of financial services.
Second, guarantees are rather unique in that, by construction, their benefit is targeted at
debtholders and not equity holders. This unique feature is important in helping us identify
empirically a main implication of our model, that there is direct feedback between sovereign
and financial sector credit risk.
In the interest of simplicity, and since debt-overhang alleviation is the central feature of
bailouts in the model, we do not explore the feedback of the guarantees on the transfer and
taxation decisions analyzed above. Instead, we simply set the stage for the implications of
the guarantees for our empirical strategy.
3.6.1 Avoiding Liquidation
A precursor to the government’s actions to increase the provision of financial services is to
prevent liquidation of the financial sector. We model debtholders as potentially liquidating
(or inducing a run on) the financial sector if they are required to incur losses in case of fi-
nancial sector default. To prevent debtholders from liquidating, the government ‘guarantees’
their debt. That is, it pledges to bondholders
L
1

˜
A
1
−T
0
from tax revenues in case of insolvency. This ‘guarantee’ is pari-passu with other claims on
tax revenue. Hence, the ‘guarantee’ has the same credit risk as other claims on the sovereign.
In fact, the ‘guarantee’ is just equivalent to a claim that issues L
1

˜
A
1
−T
0
new government
bonds to debtholders in case of insolvency.
Note that this claim accrues exclusively to debtholders and not to equityholders. This
14
The fallout from the failure of Lehman brothers and the apparent desire to prevent a repeat of this
experience has strongly reinforced this view
21
differentiates it from general assets of the financial sector, such as the asset paying
˜
A
1
or
the transfer, T
0
. Importantly, a change in the value of general assets of the firm changes
the value of equity and debt in a certain proportion, while a change in the value of the
guarantee changes the value of debt but not the value of equity. This implies that, if there
are guarantees, the change in equity value will not be sufficient for determining the change
in debt value. The following proposition gives a formal statement of this, derived under a
uniform distribution for
˜
A
1
(same as used above to generate the figures).
Proposition 4. Let D be the value of debt, E the value of equity, and
˜
A
1
be distributed
uniformly. In the absence of a guarantee, the return on equity is sufficient for knowing the
return on debt. In contrast, in the presence of a guarantee, the return on debt is a bivariate
function of both the return on equity and the return on the sovereign bond price.
This bivariate dependence is approximated by the following relation, which is derived in
the Appendix,
d D
D

(1 −p
solv
)(1 −P
0
)
p
solv
E
D
dE
E
+
(1 −p
solv
)
2
2
L
1
P
0
D
dP
0
P
0
(12)
The term involving the equity return captures the impact on the debt value of any changes
in the value of the general pool of assets of the firm, including expected future profits of
the firm. This corresponds to the canonical model of debt, following Merton (1974), where
changes in the total value of the firm are reflected in both debt and equity values. In the
presence of a guarantee, there is the additional second component, which picks up the change
in the value of debt coming from changes in the value of the government guarantee. The
change in value of the guarantee, which reflects variation in the credit risk of the sovereign,
is concentrated primarily with debt. It is therefore not captured adequately by the return
on equity and requires the second term.
15
3.6.2 Two-way Feedback
As illustrated by Proposition 3 and the discussion of Figure 5, a severe financial sector crisis
‘spills over’ onto the sovereign via an increase in H and deterioration of the sovereign’s
15
Note that there is theoretically also an indirect feedback of sovereign credit risk on financial sector credit
risk. This is because the guarantee represents additional (state-contingent) debt issuance by the sovereign,
which raises expected default costs of the sovereign and potentially reduces the size of the transfer to the
financial sector. As mentioned above, to maintain simplicity we will not incorporate the impact of the
guarantee on the government’s optimal choice of tax revenue or debt issuance. Instead, we will mainly use
the above result to motivate some of our empirical work.
22
creditworthiness. With the introduction of uncertainty about realized tax revenues, the
increase in H varies continuously with the severity of the financial sector crisis (L
1
) and
is reflected continuously in the price of the sovereign’s bonds (equivalently, in CDS levels).
Moreover, as we have shown, there is an upper value where total default becomes optimal.
In addition, as the bottom panel of Figure 5 shows, higher pre-existing sovereign debt is
associated with lower post-transfer sovereign bond prices.
Going in the other direction, higher pre-existing debt corresponds to a smaller transfer
and therefore, ceteris paribus, a weaker post-transfer financial sector (lower post-transfer
p
solv
). Moreover, if the financial sector has substantial holdings of existing sovereign bonds
(substantial k
A
), then any resulting decrease in the sovereign’s bond price due to increasing
H causes collateral damage to the financial sector’s balance sheet. This decreases the net
transfer to the financial sector, leaving it less well-off post transfer. There is, furthermore,
an additional important consequence of increased H that can be seen from the model with
uncertainty. Once H is increased, not only does the probability of default increase, but so
does the sensitivity of the bond’s final payoff to realized growth (and hence tax revenue)
shocks. Since the financial sector’s solvency post-transfer is dependent on this payoff due to
its holdings of transfer and pre-existing government bonds, the increase in H will make the fi-
nancial sector more sensitive to shocks to
˜
R
V
(e.g., growth). This implies that, post-transfer,
there will be increased co-movement between the likelihood of sovereign and financial sector
solvency.
4 Empirics
In this section we present empirical evidence in favor of the main arguments made by this
paper: (1) the bailouts reduced financial sector credit risk but were a key factor in triggering
the rise in sovereign credit risk of the developed countries and (2) there is a two-way feedback
between the creditworthiness of the sovereign and the financial sector.
The setting for our empirical analysis is the financial crisis of 2007-10. We divide the crisis
into three separate periods relative to the bailouts: pre, around, and after. The pre-bailout
period, which culminated in Lehman Brother’s bankruptcy, saw a severe deterioration in
banks’ balance sheets, a substantial rise in the credit risk of financial firms, and a significant
loss in the market value of their equity. This negative shock generated substantial debt
overhang in the financial sector and significantly increased the likelihood of failure of, or
runs on, financial institutions. We interpret this as setting the stage for the initial time
23
period in the model, and the bank bailouts as the sovereign’s response, per the model.
We present our empirical results in two parts. The first part focuses on point (1). We
presents evidence that the bailouts transmitted risk from the banks to the sovereigns and
triggered a rise in sovereign credit risk across a broad cross-section of developed countries.
We then confirm a prediction of the model by documenting the emergence post-bailout
of a positive relationship between sovereign credit risk and government debt-to-gdp ratios.
We also analyze the ability of the pre-bailout credit risk of the financial sector and pre-
bailout government debt-to-gdp ratio to predict post-bailout credit risk. This relationship
is predicted by the model and is supportive of argument that the bailouts induced increases
in sovereign credit risk.
The second part of our analysis focuses on point (2) by testing for the sovereign-bank
two-way feedback. We make extensive use of a broad panel of bank and sovereign CDS data
to carry out tests that establish this channel and show that it is quantitatively important.
A significant challenge in demonstrating direct sovereign-bank feedback is the concern that
another (unobserved) factor directly affects both bank and sovereign credit risk, giving rise
to co-movement between them even in the absence of any direct feedback. We address these
concerns by utilizing a particularly useful feature of government ‘guarantees’–that they are
targeted specifically at bank debt holders–to control for changes in bank fundamentals.
We also gather and exploit data on the sovereign bond holdings of European banks that
was released after the stress tests conducted in the first half of 2010. Using this data we show
that banks’ holdings of foreign sovereign bonds has information about how foreign sovereign
credit risk affects a bank’s credit risk.
We next describe the data construction and provide some summary statistics. This is
followed by the two sections of detailed results and the evidence based on the European bank
stress test data.
4.1 Data and Summary Statistics
We use Bankscope to identify all banks headquartered in Western Europe, the United States,
and Australia with more than $50 billion in assets as of the end of fiscal year 2006. We choose
this sample because smaller banks and banks outside these countries usually do not have
traded CDS. We then search for CDS prices in the database Datastream. We find CDS
prices for 99 banks and match CDS prices to bank characteristics from Bankscope. Next,
we search for investment grade credit ratings using S&P Ratings Express. We find credit
24
ratings for 86 banks and match these data to CDS prices and bank characteristics. Finally,
we match these data to sovereign CDS of bank headquarters and OECD Economic data on
public debt.
Panel A of Table 1 presents summary statistics for all banks with CDS prices and invest-
ment grade credit ratings. As of July 2007, the average bank had assets of $589.3 billion
and equity of $26.8 billion. The average equity ratio was 5.1% and the average Tier 1 ratio
was 8.5%. The average bank CDS was 21.8 basis points and the average sovereign CDS (if
available as of July 2007) was 6.6 basis points.
Panel B of Table 1 present summary statistics of weekly changes in bank CDS and
sovereign CDS for the main bailout periods. We drop all observations with zero changes
in bank CDS or sovereign CDS to avoid stale data. All results presented below are robust
to including these observations. Before the bank bailouts, the average bank CDS was 98.2
basis points. This level of bank credit risk reflected primarily banks’ exposure to subprime
mortgages and related assets. The average sovereign CDS was 12.1 basis points. This low
level of sovereign credit risk suggests that financial market participants did not anticipate
large-scale bank bailouts prior to September 2008.
In the bailout period, we see a significant rise in both bank and credit risk with average
bank and sovereign CDS of 301.3 and 33.6 basis points, respectively. As we discuss below, the
increase in bank credit risk was triggered by the Lehman bankruptcy. After Lehman, most
governments announced bank bailouts aimed at reducing bank credit risk. As a result of
the bailouts, some of the financial sector risk was transferred to sovereigns, which increased
average sovereign credit risk. We note that bank equity values declined sharply during this
period with a negative weekly return of 7.4%.
In the post-bailout period, average bank and sovereign CDS were 194.1 and 90.8 basis
points, respectively. These CDS levels are suggestive of a significant transfer of financial
sector credit risk on sovereign balance sheets. We also find significant variation in sovereign
CDS with a standard deviation of weekly changes of 12.9%. This evidence suggests the
emergence of significant sovereign credit risk after the bank bailouts.
4.2 The Sovereign Risk Trigger
The first bank bailout announcement in Western Europe was on September 30, 2008 in
Ireland. Therefore we define the pre-bailout period as ending on September 29, 2008. We
start it in March 2007, prior to the start of the financial crisis. Note that this period
25
includes the bankruptcy of Lehman Brothers on September 15, 2008, but also the period
immediately afterwards, so that it includes the immediate effect of Lehman’s bankruptcy on
other banks prior to the Ireland announcement. Hence, the pre-bailout period captures both
the prolonged increase in bank credit risk during 2007-2008 and the post-Lehman spike that
occurs before the bank bailouts.
We compute the change in sovereign CDS and bank CDS during this period for all
countries in our data set. For each country we compute the change in bank CDS as the
unweighted average of all the banks with CDS prices. We omit countries for which either
sovereign CDS or banks CDS are not available. Figure 6 summarizes the results for the
pre-bailout period. For each country, the first column depicts the change in sovereign CDS
and the second column depicts the change in bank CDS over the pre-bailout period. The
figure shows that there is a large increase in banks CDS prior to the bank bailouts. For
example, the average bank CDS in Ireland increases by 300 basis points over this period.
However, there is almost no change in sovereign CDS. Overall, the figure shows that the
credit risk of the financial sector was greatly increased over the pre-bailout period but that
there was little impact on sovereign credit risk.
We note that some investors may have expected bank bailouts even before the first official
announcement on September 30, 2008. Such an expectation would reduce the observed
increase in bank CDS and shift forward in time the rise in sovereign CDS. To the extent that
investors held such expectations prior to September 30, 2008, they can explain the small
rise in sovereign CDS that occurs late in the pre-bailout period. However, the fact that the
impact in this period is so small quantitatively suggest that the bank bailouts were a surprise
to the majority of investors.
Almost every Western European country announced a bank support program in October
2008. Most bank support programs consisted of asset purchase programs, debt guarantees,
and equity injections or some combination therefore. Several countries made more than
one announcement during this period. As noted above, the first country to make a formal
announcement was Ireland on September 30, 2010. Many other countries soon followed
Ireland’s example, in part to offset outflows from their own financial sectors to newly secured
financial sectors. As a result, the bank bailout announcements were not truly independent
since sovereigns partly reacted to other sovereigns’ announcements. We therefore define the
around-bailouts period as the period in which the bank bailout announcements occurred.
Figure 7 plots the average change in bank CDS and sovereign CDS over the bailouts
period. As shown in the figure, bank CDS significantly decreased over this one-month period.
26
For example, the average bank CDS in Ireland decreased by about 200 basis points. Similarly,
most other countries had a significant decrease in bank CDS, especially ones that had a
large increase over the pre-bailout period. At the same time, there is a significant increase in
sovereign CDS. For example, the sovereign CDS of Ireland increased by about 50 basis points.
Most other countries exhibit a similar pattern with decreasing bank CDS and increasing
sovereign CDS.
The appearance of this striking pattern across a broad cross-section of countries is di-
rectly in line with the predictions of our model. It shows that the sovereigns responded to
the distress in the financial sector with the bailouts, achieving a substantial reduction in
banks’ credit risk. However, in accordance with our model, this caused a contemporaneous,
immediate increase in the sovereigns’ credit risk. Indeed the figures clearly show that these
sharp increases were aligned tightly with the bailout announcement period.
We define the post-bailout period as beginning after the end of the bank bailouts and
ending in March 2010. We choose March 2010 because this is the date for which the European
bank stress data results were released. The results are robust to using other cut-off dates in
2010. Figure 8 plots the change in bank CDS and sovereign CDS over this post-bailout period.
It shows that both sovereign CDS and bank CDS increased across all countries. Moreover,
it is clear that bank CDS and sovereign CDS move together after the bank bailouts. This
suggests that they are tied together and feedback on each other as our model suggests.
Our model predicts that the bailouts should lead to an increase in sovereign credit risk
and that the post-bailout level of credit risk should depend on pre-bailout debt and the
pre-bailout level of financial sector distress. Moreover, the model suggests that there should
emerge after the bailouts a positive relationship between measures of government debt-to-gdp
even if such relationship appears before.
We test these predictions using data on sovereign CDS, financial sector distress, and gov-
ernment debt-to-gdp ratios. We measure pre-bailout financial sector distress at the country
level by averaging bank CDS on September 22, 2008. We choose this date midway between
Lehman’s bankruptcy and the first bailout announcement. We measure the government
debt-to-gdp ratio as the government gross liabilities as a percentage of gdp. For the post-
bailout date we choose March 31, 2010, the reporting date for the European bank stress
tests.
Table 2 presents the result of our analysis. Column (3) shows the result from regress-
ing post-bailout log sovereign CDS on post-bailout debt-to-gdp. There is a clear, positive
27
relationship and the coefficient is statistically significant.
16
This relationship is shown in
the bottom panel of Figure 9. Column (4) shows the result when the log of the pre-bailout
financial distress variable is also added as a dependent variable. As Column (4) shows the
coefficient on pre-bailout financial distress is large and highly statistically significant. The
coefficient shows that a 1% increase in pre-bailout financial sector distress increases post-
bailout sovereign CDS by 0.965%. The coefficient on debt-to-gdp decreases slightly but
remains marginally significant. The R-squared of the regression is close to 50%.
In contrast, Column (1) of Table 2 shows that in the pre-bailout period there is only
a very weak relationship between debt-to-gdp and sovereign CDS. The coefficient is small
and is statistically insignificant. This is displayed in the upper panel of Figure 9. Column
(3) shows that the coefficient on financial sector distress in the pre-bailout period is also
statistically insignificant.
From these results we can see that there emerged a relationship between debt-to-gdp and
sovereign credit risk that was not present beforehand. Note, moreover, that the emergence
of this relationship coincides with an overall rise in sovereign debt ratios. From the point of
model, the sovereigns have increased debt-ratios into the region where dilution occurs and
there is a negative relationship between debt-ratio and price (see Figure 5).
Columns (5) examines the ability of pre-bailout financial sector distress to predict the
change in government debt-to-gdp from the pre-bailout to the post-bailout period. Consis-
tent with the model, we find that financial sector distress is positively related to the increase
in debt-to-gdp. The coefficient is positive and marginally statistically significant. Column (6)
shows that consistent with the model’s predictions, post-bailout debt-to-gdp is predicted by
pre-bailout debt-to-gdp and pre-bailout financial sector distress. Both coefficients are statis-
tically significant and together the two variables explain 84% of the variation in post-bailout
debt-to-gdp.
4.3 The Sovereign-Bank Feedback
In this section we analyze the two-way feedback between sovereign and bank sector credit
risk. Once the sovereign opens itself up to credit risk due to bailouts, the price of its debt
16
We make note of two points. First, the model predicts that H determines the level of sovereign CDS.
However, the debt-to-gdp ratio corresponds to θ
0
H in the model rather than simply H. Nevertheless, the
prediction of the model carries over to θ
0
H since θ
0
is increasing in financial sector distress. Second, debt-
to-gdp ratios are an imperfect proxy for θ
0
H because H takes into account any future issuance of debt to
pay for current obligations related to the bailouts, whereas debt-to-gdp ratios are lagging. We can address
this to some extent by using leading debt-to-gdp.
28
becomes sensitive to macroeconomic shocks. This leads a second direction of feedback, from
the sovereign to the financial sector. Our model indicates that subsequent changes in the
sovereign’s credit risk should impact the financial sector’s credit risk through three channels:
(i) ongoing bailout payments and subsidies,
17
(ii) direct holdings of government debt, (iii)
explicit and implicit government guarantees.
The feedback channels imply that we should find that changes in sovereign and bank
credit risk are positively correlated. We start by estimating the following relationship in the
post-bailout period:
∆log(Bank CDS
ijt
) = α +β∆log(Sovereign CDS
jt
) +ε
ijt
where ∆log(Bank CDS
ijt
) is the change in the log CDS of bank i headquartered in country
j from time t to time t − 1 and ∆log(Sovereign CDS
jt
) is the change in the log Sovereign
CDS of country j from time t to time t − 1. At the weekly frequency, the estimate of β is
0.47 and is highly statistically signficant. This means that a 10% increase in sovereign CDS
is associated with a 4.7% increase in bank CDS. This result shows that sovereign and bank
CDS exhibit a strong comovement and is consistent with direct sovereign-bank feedback.
However, an obvious concern is that there is another (unobserved) factor that affects
both bank and sovereign credit risk. Such a factor would explain their co-movement without
there necessarily being an underlying direct channel between them. More specifically, we
interpret changes in sovereign credit risk as changes in expectations about macroeconomic
fundamentals, such as employment, growth, and productivity. These fundamentals also have
a direct effect on the value of bank assets such as mortgages or bank loans. Hence, changes
in macroeconomic conditions may generate a correlation between sovereign and bank credit
risk even in the absence of direct feedback. Therefore, establishing that there is indeed direct
feedback between sovereign and financial sector credit risk is a significant challenge.
We take a number of steps to address this concern. In the first step, we add controls
that capture market-wide changes that affect both bank and sovereign risk directly. Our
market-wide controls are a CDS-market index and a measure of aggregate volatility. Our
CDS market index is the iTraxx Europe index, which is comprised of 125 of the most liquid
17
While we model the bailout as happening in one stage, in practice bailouts may occur in mutliple steps.
This may occur when early efforts are deemed to be insufficient. As with the one-stage bailout in our model,
the sovereign’s creditworthiness is important in determining the potential for and magnitude of future bailout
steps. This presents another, dynamic channel through which changes in sovereign creditworthiness impact
both financial sector debt and equity values.
29
CDS names referencing European investment grade credits. The CDS market index captures
market-wide variation in CDS rates caused by changes in fundamental credit-risk, liquidity,
and CDS-market specific shocks.
18
For the volatility index we follow the empirical literature
and use a VIX-like index, the VDAX, which is the DAX-counterpart to the VIX index for
the S&P 500. This captures changes in aggregate volatility which is an important factor in
the pricing of credit risk. In the second step, we include weekly fixed effects. For each week,
the fixed effect captures any variation that is common across all banks. In the third step,
we include bank-specific coefficients on all the control variables and bank fixed effects. This
accommodates potential non-linearities in relationships.
We implement this approach by estimating the following regression:
∆log(Bank CDS
ijt
) = α
i

t
+β∆log(Sovereign CDS
jt
) +γ∆X
ijt

ijt
where ∆X
ijt
are the changes in the control variables from time t to time t − 1, δ
t
are the
weekly fixed effects, and the α
i
are bank fixed-effects
Table 3 shows the results for the pre, around, and post-bailout periods. For each period
there are three columns of results. The left column reports the coefficient estimates including
the market-wide control variables. The middle column adds the weekly fixed effects. The
right column adds the bank-specific coefficients on the controls and bank fixed effects.
Our main focus is on testing for the sovereign-bank feedback, so we examine the post-
bailout results first. Column (7) shows that β is positive, as expected, and highly statistically
significant. The magnitude is also economically important, implying that an independent
increase in sovereign CDS of 10% translates into a 1.63% increase in bank CDS. The control
coefficients are both statistically significant and the signs are as expected; an increase in
aggregate CDS levels or volatility is associated with a rise in banks’ CDS. As would be
expected, the coefficient on CDS market is large. Altogether, the variables explain 31.6% of
the variation in weekly bank CDS.
Column (8) adds the weekly fixed effects. The coefficient on sovereign CDS decreases but
remains highly statistically significant. The decrease is not surprising, as time fixed effects
represent a very rich set of controls. Note that the weekly fixed effects are collinear with the
market-wide control variables. Therefore, we do not estimate coefficients on the market-wide
18
Collin-Dufresne, Goldstein, and Martin (2001) find that a substantial part of the variation in corporate
credit spread changes is driven by a single factor that is independent of changes in risk factors or measures
of liquidity. They therefore conclude that this variation represents ‘local supply/demand shocks’ in the
corporate bond market.
30
controls. There is an increase in the R-squared ofabout 7% over column (7), indicating that
most of the variation was already captured by the market-wide controls.
Column (9) shows that the coefficient on sovereign CDS is essentially unchanged and
remains highly statistically significant after adding bank-specific coefficients on the control
variables . Given the flexibility of this specification we interpret the survival of the coefficient
on sovereign CDS as robust evidence in favor of direct sovereign-bank feedback.
Comparing these results with those for the around-bailout period in columns (4)-(6)
shows interesting differences. For the around-bailout period, the coefficient on sovereign
CDS is negative. In other words, in the around-bailout period an independent increase in
sovereign CDS is associated with a decrease in bank CDS. This is very much consistent with
the evidence presented above that the sovereigns took onto themselves credit risk from their
financial sectors during this phase. More precisely, it is evidence that in fact sovereigns that
took on more credit risk, and hence saw a greater increase in their CDS, decreased by a
greater amount their banks’ CDS. Perhaps due to the short time-series, the coefficients are
only marginally significant. When the panel is estimated at the daily frequency they do in
fact come up significant (unreported).
Columns (1)-(3) show the results for the pre-bailout period. They show a small coefficient
on sovereign CDS that is indistinguishable from 0. Hence, in the pre-bailout period there is
no evidence for sovereign-bank feedback. In contrast, the CDS market control coefficient is
significant and has a large magnitude, similar to the results for the other periods.
4.3.1 Controlling for Bank Fundamentals
The results above establish that there is a strong sovereign-bank feedback. However, there
may remain a concern that our strategy to this point does not control for country-specific
macroeconomic shocks that bank-level fundamentals. For example, a negative macroeco-
nomic shock will decrease the value of banks’ assets or future earnings power, while at the
same time reducing national output. This will raise bank and sovereign CDS even in the
absence of a direct feedback between them. This shock may not be fully captured by our
market-wide controls if it has a heterogenous impact across countries.
We address this concern using a strategy that utilizes a particularly useful feature of
government ‘guarantees’–they are targeted specifically at bank debt holders. This implies
that sovereign-specific shocks should have a disproportionate impact on debt holders because,
in addition to changing the value of bank assets, such shocks change the value of government
‘guarantees’ (implicit or explicit). Therefore, to establish whether there is a direct channel
31
we can test if sovereign CDS is still a determinant of bank CDS after we control for the
impact of shocks to bank fundamentals.
Our strategy for dealing with this concern is motivated by the model. Proposition 4 shows
that bank equity returns are sufficient for determining changes in bank CDS in the absence
of government ‘guarantees’.
19
This implies that once we control for bank equity returns
we should not find that changes in sovereign CDS have any further explanatory power for
changes in bank CDS. Indeed, a-priori it seems that even if there are direct sovereign-to-
bank channels, their impact may be subsumed into an equity control, making it difficult
to document their existence. On the other hand, if one finds that sovereign CDS does
have further explanatory power beyond equity returns, then this is strongly supportive of a
sovereign-to-bank feedback channel. Proposition 4 shows that ‘guarantees’ present a source
for such a potential finding because they discriminate precisely in favor of debtholders. In
the presence of ‘guarantees’, a regression that controls for equity returns should still find a
(negative) beta on changes in sovereign CDS.
We therefore augment the regression with banks’ weekly equity returns. The estimates
are shown in Table 4. The structure is similar to Table 3. Columns (7)-(9) show that
in the post-bailout period we find that the coefficient on sovereign CDS survives and is
highly statistically significant. As shown in columns (7) and (8), although the bank stock
return coefficient is highly statistically significant and possesses the expected negative sign,
its inclusion has little impact on the magnitude of the sovereign CDS coefficient. Column
(9) includes bank-specific coefficients on bank stock returns. The results show that the
coefficient on sovereign CDS decreases somewhat but remains highly significant.
For the bailout and pre-bailout periods, the results are quite similar to those in Table 3.
As shown in columns (1) to (6), we again find a negative coefficient on sovereign CDS in the
bailout period and an essentially zero coefficient in the pre-bailout period.
4.3.2 Bank-level Heterogeneity
To analyze further sovereign-bank feedback we also examine whether heterogeneity in bank
characteristics affects banks’ sensitivity to changes in sovereign CDS. To this end, we estimate
the coefficient on an interaction term of changes in sovereign CDS and a bank’s Tier 1 capital
ratio. The Tier 1 capital ratio is commonly used in the banking industry as a proxy for a
19
This result is in fact quite general. It holds in models of defaultable bond pricing that build on the
canonical model of Merton (1974), where stock returns contain all information about changes in bank asset
values and therefore can (locally) capture all variation in the price of debt.
32
bank’s probability of solvency. This specification is motivated by the model. Equation (12)
suggests that we should find that the magnitude of the coefficient on changes in sovereign
CDS is decreasing in banks’ Tier 1 ratios. Intuitively, the impact of changes in the value of
government guarantees stronger for less well-capitalized banks.
Table 5 reports the results. The coefficient of interest is the interaction term. For the
post-bailout period, the negative estimates for the interaction term indicate that banks with
higher Tier 1 ratios were less sensitive to variation in sovereign CDS rates. Although the
estimates are negative in columns (7) to (9), only column (9), which includes the full set of
controls is statistically significant.
The results in the bailout period are surprising. Similar to the post-bailout period, the
coefficient on the interaction term is negative. However, this implies that well-capitalized
banks experienced a larger CDS decrease relative to poorly capitalized banks during the
bailout period. In the pre-bailout period, the interaction term is positive but small and
statistically insignificant.
4.3.3 The Impact on Equity Value
For the purposes of establishing the existence of a two-way feedback we have mainly focused
on changes in bank CDS. It is also interesting to look at the impact of bailouts on bank
stock returns. From the viewpoint of the model, bank stock returns should reflect changes
in sovereign credit risk due to their impact on the value of continuing bailout payments and
banks’ holdings of government bonds. To that end, we estimate the following regression:
Bank Stock Return
ijt
= α
i

t
+β∆log(Sovereign CDS
jt
) +γ∆X
ijt

ijt
where Bank Stock Return
ijt
is the stock return of bank i headquartered in country j from
time t to time t −1. We use the same control variables as in Table 4.
Table 6 presents the results. Columns (7) to (9) show that in the post-bailout period an
increase in sovereign CDS is associated with a decrease in bank stock returns. This result
is robust to the inclusion of the full set of controls. Note also that the coefficients on the
controls have the expected signs and are significant.
Perhaps somewhat surprisingly, the coefficient on sovereign CDS is not distinguishable
from zero during the bailout period. The point estimate is positive in columns (5) and (6),
but is not statistically significant. Still, across the board, the point estimates are clearly
higher than for the post-bailout period, which is consistent with the idea that equity holders
33
benefitted from bank bailouts. One potential offsetting effect on this result is that in some
countries the benefits of bailouts to equity holders were offset by charges (or expectations of
charges) levied by the government. Columns (1) to (3) show that as before, the coefficient
on sovereign CDS in the pre-bailout period is essentially zero.
4.4 European Bank Stress Test Analysis
As a final part of our analysis, we use data on the sovereign bond holdings of European
banks. These data were released as part of the European bank stress tests, which were
conducted in the first half of 2010. The data provide a view of a bank’s bond holdings of
both its own government bonds and those of other European countries.
We first describe the data and provide summary statistics. We collect the stress test
data from the websites of national bank regulators in Europe. The data consists of bank
characteristics and holdings of European sovereign bonds. A total of 91 banks participated
in the bank stress tests. These banks represent about 70 percent of bank assets in Europe.
For all banks, we search for CDS prices in the database Datastream. Using bank names,
we match 51 banks to CDS prices. Unmatched banks are mostly smaller banks from Spain
and Eastern Europe that do not have publicly quoted CDS prices. For each bank we match
sovereign holdings to sovereign CDS and compute a measure of exposure to sovereign credit
risk.
Table 7 presents summary statistics for all banks that participated in the European bank
stress tests. As of March 2010, the average bank had risk-weighted assets of 126 billion
euros and a Tier 1 capital ratio of 10.2%. The average holdings of gross and net European
sovereign bonds are 20.6 billion euros and 19.7 billion euros, respectively. Hence, the average
bank holds about one sixth of risk-weighted assets in sovereign bonds. Banks have a strong
home bias in their sovereign holdings: about 69.4% of bonds are issued by the country in
which a bank is headquartered. This is supportive of the model’s assumption that banks are
exposed to home-country sovereign risk through their holdings of government bonds.
We use these data to conduct an alternative test of the impact of sovereign credit risk
on bank credit risk. Our test focuses uses information on changes in the value of foreign-
sovereign holdings rather than own-country sovereign holdings. The benefit of this approach
is that it circumvents the usual concerns about omitted country-specific macro shocks.
To implement this test, we construct a bank-specific variable measuring variation in
the value of banks’ foreign sovereign bond holdings. Let SovBond
ik
be the share of for-
34
eign sovereign holdings of country k by bank i. We calculate the foreign holdings variable,
ForeignBondCDS
it
as the following weighted average:
ForeignBondCDS
it
=

i=j
SovBond
ik
∗ SovereignCDS
kt
.
where SovereignCDS
kt
is the sovereign CDS of country k on day t. Note that for each bank
the foreign holdings variable excludes home-sovereign bonds. We then estimate the following:
∆log(Bank CDS
it
) = δ
t
+γ∆log(ForeignBondCDS
it
) +ε
it
. (13)
where δ
t
are time fixed effects. The coefficient of interest is γ, which captures the effect of
changes in the value of foreign bond holdings on the bank’s CDS.
We estimate the regressions using the period one month before and one month after
the reporting date for the sovereign bond holdings. By estimating this regression, we are
implicitly assuming that the marginal CDS investor either knows or at least has some idea
of the bank holdings.
Table 8 shows the results. Column (1) shows a positive and statistically significant asso-
ciation between changes in banks’ CDS and their foreign sovereign holdings. This coefficient
suggests that a one-standard deviation increase in the change in the foreign sovereign hold-
ings variable leads to an increase of about half of a one-standard deviation in the change of
the bank CDS. Column (2) shows that the coefficient remains unchanged when bank fixed
effects are included. Column (3) controls for week fixed effects. The coefficient of interest
decreases from 0.325 to 0.261. This suggests that common shocks affect both the change
in bank CDS and the change in the foreign holdings variable. Column (4) controls for day
fixed effects. In this case, the coefficient is identified only off the cross-sectional variation in
the value of foreign sovereign holdings. The coefficient decreases to 0.141 but remains sta-
tistically significant at the 1%-level. This result suggests that variation in foreign sovereign
holdings contains economically important information about variation in bank credit risk.
To further check for robustness, column (5) adds back bank fixed effects. This does
not have any effect on the coefficient. Column (6) estimates the same regression as in
Column (5) but excludes the holdings of German bonds from the construction of the foreign-
holdings variable. We do this to address a potential concern about reverse causality due
to the possibility that Germany may provide bailouts to other countries, or banks in other
countries. The column shows that this has no effect on the coefficient of interest.
35
5 Related literature
Our paper is related to three different strands of literature: (i) the theoretical literature on
bank bailouts; (ii) the literature on costs of sovereign default; and, (iii) the recent empirical
literature on effects of bank bailouts on sovereigns.
The theoretical literature on bank bailouts has mainly focused on how to structure bank
bailouts efficiently. While the question of how necessarily involves an optimization with
some frictions, the usual friction assumed is the inability to resolve failed bank’s distress
entirely due to agency problems. This could be due to under-investment problem as in our
setup (e.g., Philippon and Schnabl, 2009), adverse selection (e.g., Gorton and Huang, 2004),
risk-shifting or asset substitution (e.g., Acharya, Shin and Yorulmazer, 2008, Diamond and
Rajan, 2009), or tradeoff between illiquidity and insolvency problems (e.g., Diamond and
Rajan, 2005). Some other papers (Philippon and Schnabl, 2010, Bhattacharya and Nyborg,
2010, among others) focus on specific claims through which bank bailouts can be structured
to limit these frictions.
A large body of existing literature in banking considers that bank bailouts are inherently
a problem of time consistency and induce moral hazard at individual-bank level (Mailath
and Mester, 1994) and at collective level through herding (Penati and Protopapadakis, 1988,
Acharya and Yorulmazer, 2007). Aghion, Bolton and Fries (1999) consider the cost that
bank debt restructuring can in some cases delay the recognition of loan losses. Brown and
Dinc (2009) show empirically that the governments are more likely to rescue a failing bank
when the banking system, as a whole, is weak.
A small part of this literature, however, does consider ex-post costs of bailouts. Notably,
Diamond and Rajan (2005, 2006) study how bank bailouts can take away a part of the
aggregate pool of liquidity from safe banks and endanger them too. Acharya and Yorulmazer
(2007, 2008) model, in a reduced-form manner, a cost of bank bailouts to the government
or regulatory budget that is increasing in the quantity of bailout funds. They provide
taxation-related fiscal costs as a possible motivation. Panageas (2010a,b) considers the
optimal taxation to fund bailouts in a continuous-time dynamic setting, also highlighting
when banks might be too big to save.
In the theoretical literature on sovereign defaults, Bulow and Rogoff (1989a, 1989b) ini-
tiated a body of work that focused on ex-post costs to sovereigns of defaulting on external
debt, e.g., due to reputational hit in future borrowing, imposition of international trade
sanctions and conditionality in support from multi-national agencies. Broner and Ventura
36
(2005), Broner, Martin and Ventura (2007), Acharya and Rajan (2010) and Gennaioli, Mar-
tin and Rossi (2010), among others, consider a collateral damage to the financial institutions
and markets when a sovereign defaults. They employ this as a possible commitment device
that gives the sovereign “willingness to pay” its creditors. Our model considers both of these
effects, an ex-post deadweight cost of sovereign default in external markets as well as an
internal cost to the financial sector through bank holdings of government bonds.
One strand of recent empirical work focuses on the distortionary design of bank bailout
packages. Acharya and Sundaram (2009) document how the loan guarantee program of the
Federal Deposit Insurance Corporation in the Fall of 2008 was charged in a manner that
favored weaker banks at the expense of safer ones, producing a downward revision in CDS
spreads of the former. Veronesi and Zingales (2009) conduct an event study and specifically
investigate the U.S. government intervention in October 2008 through TARP and calculate
the benefits to banks and costs to taxpayers. They find that the government intervention
increased the value of banks by over $100 billion, primarily of bank creditors, but also
estimate a tax payer cost between $25 to $47 billion. Panetta et al. (2009) and King (2009)
assess the Euro zone bailouts and reach the conclusion that while bank equity was wiped out
in most cases, bank creditors were backstopped reflecting a waiting game on part of bank
regulators and governments.
Another strand of recent empirical work relating financial sector and sovereign credit
risk during the ongoing crisis shares some similarity to the very recent papers on this theme.
Sgherri and Zoli (2009) and Attinasi, Checherita and Nickel (2009) focus on the effect of
bank bailout announcements on sovereign credit risk measured using CDS spreads. Some
of their evidence mirrors our descriptive evidence. Dieckmann and Plank (2009) analyze
sovereign CDS of developed economies around the crisis and document a significant rise
in co-movement following the collapse of Lehman Brothers. Demirguc-Kunt and Huizinga
(2010) do an international study of equity prices and CDS spreads around bank bailouts and
show that some large banks may be too big to save rather than too big to fail. Our analysis
corroborates and complements some of this work. In particular, our empirical investigation
of banking sector holdings of government debt and how this introduces a linkage between
bank CDS and sovereign CDS is novel.
Finally, Reinhart and Rogoff (2009a, b) and Reinhart and Reinhart (2010) document that
economic activity remains in deep slump “after the fall” (that is, after a financial crisis), and
private debt shrinks significantly while sovereign debt rises, especially beyond a threshold of
90% debt to GDP ratio of the sovereign. These effects are potentially all consistent with our
37
model of how financial sector bailouts affect sovereign credit risk and economic growth.
6 Conclusion
This paper examines the intimate and intricate link between bank bailouts and sovereign
credit risk. We develop a model in which the government faces an important trade-off:
bank bailouts ameliorate the under-investment problem in the financial sector but reduce
investment incentives in the non-financial sector due to costly future taxation. In the short-
run, bailouts are funded through the issuance of government bonds, which dilutes the value
of existing government bonds and creates a two-way feedback mechanism because financial
firm hold government bonds for liquidity purposes. We also provide supporting evidence for
our model using data from the financial crisis of 2007-10. In particular, we document that
developed country governments transferred credit risk from the financial sector to taxpayers
during the height of the crisis in October 2008. Using credit ratings data and data on
sovereign bond holdings from the European bank stress test in May 2010, we find that
sovereign credit risk in turn affected bank credit risk.
Overall, we consider the emergence of meaningful sovereign credit risk as an important
potential cost of bank bailouts. This cost is a reflection of the future taxation (or more
generally, even inflation) risk imposed on corporate and household sectors of the economy.
Such an ex-post cost of bailouts has received little theoretical attention and has also not been
analyzed much empirically. Taking cognizance of this ultimate cost of bailouts has important
consequences for the future resolution of financial crises, the design of fiscal policy, and the
nexus between the two.
References
Acharya, Viral V. and Raghuram G. Rajan, 2010, “Financial Sector and Sovereign Credit
Risk”, Work in progess, New York University Stern School of Business
Acharya, Viral, Hyun Song Shin and Tanju Yorulmazer, 2008, “Crisis Resolution and Bank
Liquidity”, Review of Financial Studies, forthcoming.
Acharya, Viral V. and Rangarajan Sundaram, 2009, “The Financial Sector ‘Bailout’: Sow-
ing the Seeds of the Next Crisis”, Chapter 15 in Acharya, Viral V. and Matthew
38
Richardson (editors), 2009. “Restoring Financial Stability: How to Repair a Failed
System”, New York University Stern School of Business, John Wiley & Sons.
Acharya, Viral V. and Tanju Yorulmazer, 2007 “Too Many to Fail - An Analysis of Time-
inconsistency in Bank Closure Policies,” Journal of Financial Intermediation, 16(1),
1–31.
Acharya, Viral V. and Tanju Yorulmazer, 2008, “Cash-in-the-Market Pricing and Optimal
Resolution of Bank Failures,” Review of Financial Studies, 21, 2705–2742.
Aghion, Philipp, Patrick Bolton and Steven Fries, 1999, “Optimal Design of Bank Bailouts:
The Case of Transition Economies”, Journal of Institutional and Theoretical Eco-
nomics, 155, 51–70.
Attinasi, Maria-Grazia, Cristina Checherita and Christiane Nickel, 2009, “What Explains
the Surge in Euro Area Sovereign Spreads during the Financial Crisis of 2007-09?”,
European Central Bank Working Paper No. 1131.
Bhattacharya, Sudipto and Kjell G. Nyborg, 2010, “Bank Bailout Menus”, Swiss Finance
Institute Research Paper Series No. 10–24.
Broner, Martin and Jaume Ventura, 2005, “Globalization and Risk Sharing”, Working
Paper, Universitat Pompeu Fabra.
Broner, Martin, Alberto Martin and Jaume Ventura, 2007, “Enforcement Problems and
Secondary Markets”, Working Paper, Universitat Pompeu Fabra.
Brown, C. and Dinc, S., 2009, “Too Many to fail? Evidence of regulatory forbearance When
the banking sector is weak”, Review of financial Studies, forthcoming.
Bulow, J., and K. Rogoff, 1989a, “Sovereign debt: Is to forgive to forget?”, American
Economic Review, 79, 43-50.
Bulow, J., and K. Rogoff, 1989b, “A constant recontracting model of sovereign debt”,
Journal of Political Economy, 97, 155-178.
Demirguc-Kunt, Asli and Harry Huizinga, 2010, “Are Banks Too Big to Fail or Too Big
to Save? International Evidence from Equity Prices and CDS Spreads”, World Bank
Policy Research Working Paper 5360.
39
Diamond, Douglas and Raghuram G. Rajan, 2005, “Liquidity Shortages and Banking
Crises”, Journal of Finance, 60(2), 615–647.
Diamond, Douglas and Raghuram G. Rajan, 2006, “Bank Bailouts and Aggregate Liquid-
ity”, American Economic Review Papers and Proceedings, 92(2), 38–41.
Diamond, Douglas and Raghuram G. Rajan, 2009, “Fear of Fire Sales, Illiquidity Seeking
and Credit Freeze”, Quarterly Journal of Economics, forthcoming.
Dieckmann, Stephan and Thomas Plank, 2009, “Default Risk of Advanced Economies:
An Empirical Analysis of Credit Default Swaps during the Financial Crisis”, Working
Paper, Wharton School of Business, University of Pennsylvania.
Gennaioli, Nicola, Alberto Martin and Stefano Rossi, 2010, “Sovereign Default, Domestic
Banks and Financial Institutions,” Working Paper, Universitat Pompeu Fabra.
Gorton, Gary and Lixin Huang. 2004. Liquidity, Efficiency, and Bank Bailouts. American
Economic Review, 94(3): 455-483.
King, Michael R, 2009, “Time to Buy or Just Buying Time? The Market Reaction to Bank
Rescue Packages”, BIS Papers No 288 (September).
Mailath, George and Loretta Mester, 1994, “A Positive Analysis of Bank Closure,” Journal
of Financial Intermediation, 3(3), 272–299.
Myers, Stewart C., 1977, “The Determinants of Corporate Borrowing,” Journal of Financial
Economics, 5(2), 147–175.
Panageas, Stavros, 2010a, “Optimal taxation in the presence of bailouts”, Journal of Mon-
etary Economics, 57, 101?-116.
Panageas, Stavros, 2010b, “Bailouts,the incentive to manage risk,and financial crises”, Jour-
nal of Financial Economics, 95, 296?-311.
Panetta, Faib, Thomas Faeh, Giuseppe Grande, Corrinne Ho, Michael King, Aviram Levy,
Federico M Signoretti, Marco Taboga, and Andrea Zaghini, 2009, “An Assessment of
Financial Sector Rescue Programmes”, BIS Papers No 48 (July).
40
Penati, A. and Protopapadakis, A., 1988, “The effect of implicit deposit insurance on
banks? portfolio choices, with an application to international overexposure”, Journal
of Monetary Economics, 21(1), 107–126.
Veronesi, P. and Zingales, L., 2009, “Paulson’s Gift”, Journal of Financial Economics,
forthcoming.
Philippon, Thomas and Philipp Schnabl, 2009, “Efficient Recapitalization”, Working Paper,
New York University Stern School of Business.
Reinhart, Carmen M. and Vincent Reinhart, 2010, “After the Fall”, presentation at the
Federal Reserve Bank of Kansas City Economic Symposium at Jackson Hole, Wyoming,
August 27, 2010.
Reinhart, Carmen M. and Kenneth S. Rogoff, 2009a, This Time Is Different: Eight Cen-
turies of Financial Folly, Princeton University Press.
Reinhart, Carmen M. and Kenneth S. Rogoff, 2009b, “Growth in a Time of Debt”, American
Economic Review Papers and Proceedings, forthcoming.
Sgherri, Silvia and Edda Zoli, 2009, “Euro Area Sovereign Risk During the Crisis”, Working
paper, International Monetary Fund, No. 09–222.
41
Figure 1: Sovereign CDS and bank CDS of Ireland
 
0
2
0
0
4
0
0
6
0
0
8
0
0
01jan2007 01jan2008 01jan2009 01jan2010 01jan2011
date
(mean) cds (mean) countrycds
Figure 1 plots the sovereign CDS and bank CDS for Ireland in the period from 3/1/2007
to 8/31/2010. The bank CDS is computed as the unweighted average of bank CDS for
banks headquartered in Ireland (Allied Irish Bank, Anglo Irish Bank, Bank of Ireland,
and Irish Life and Permanent). The data are from Datastream.
Figure 2: Marginal Gain and Loss of Raising T (Certainty Case)
0.05 0.1 0.15 0.2 0.25 0.3
τ
d
G
/
d
T
0

a
n
d

d
L
/
d
τ
dG vs. dL
Value of Government’s Objective Function
0.05 0.1 0.15 0.2 0.25 0.3
τ
v
a
l
.

g
o
v
.

o
b
j
Gov. Objective Value
The top panel of Figure 2 plots the marginal gain (dG/dT ) of raising tax revenues
(solid line and dash-dot line) and the marginal loss (dL/dT , dashed line) as functions
of T for the certainty model of Section 3.3. The dash-dot line corresponds to a higher
level of existing government debt, N
D
, than the solid line. The bottom panel of the
Figure shows the resulting value of the government’s objective function (equation (3)),
with the the solid and dash-dot line corresponding to their counterparts in the top
panel. The plots correspond to a parameterization of the model where
˜
A
1
∼ U[0, 1],
L
1
= 0.5, α = 1, ϑ = 0.3, γ = 0.2, β = 0.5, m = 1.3, D = 0.02, k = 0, and N
D
= 0.25
(solid line).
Figure 3: The Default Boundary (Certainty Case)
Default and No−Default Regions
L
1
N
D
Default
No Default
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
0
0.05
0.1
0.15
0.2
0.25
0.3
Figure 3 shows the Default and No-Default Regions in the space of L
1
×N
D
(financial
sector leverage/debt overhang × pre-existing sovereign debt) for the certainty model
parameterized as in Figure 2. The black curve separating these two regions gives the
Default Boundary.
Figure 4: Marginal Gain and Loss of Increasing H
0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6
H
d
G
/
d
H

a
n
d

d
L
/
d
H
dG vs. dL
Value of Government’s Objective Function
0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6
H
value of gov. objective
The top panel of Figure 4 plots the marginal gain of increasing H while holding constant
T , dG/dH (solid line and dash-dot line) and the resulting marginal increase in expected
dead-weight default cost D
dpdef
dH
(dashed line). Uncertainty over growth/tax revenues,
˜
R
V
, is assumed to have a uniform distribution. The dash-dot line corresponds to a
higher level of L
1
than for the solid line. The bottom panel of the Figure shows the
resulting value of the government’s objective function, with the the solid and dash-dot
line corresponding to their counterparts in the top panel. The plots correspond to a
parameterization of the model where
˜
R
V
∼ U[0.6, 1.4],
˜
A
1
∼ U[0, 1], L
1
= 0.5 (solid
line), α = 1, ϑ = 0.3, γ = 0.2, β = 0.5, m = 1.3, D = 0.06, k = 0, and N
D
= 0.25.
Figure 5: Comparative Statics for L
1
0.4 0.6 0.8
L
1
τ
0.4 0.6 0.8
L
1
H
0.4 0.6 0.8
L
1
T
0
0.4 0.6 0.8
L
1
P
0
Comparative Statics for N
D
0.1 0.2 0.3 0.4
N
D
τ
0.1 0.2 0.3 0.4
N
D
H
0.1 0.2 0.3 0.4
N
D
T
0
0.1 0.2 0.3 0.4
N
D
P
0
Figure 5 plots the equilibrium values of T (expected tax revenue), H, T
0
(the transfer),
and P
0
(price of sovereign bond) as L
1
(top panel) and N
D
(bottom panel) are varied.
The dotted line in the plots represents the point at which total default (H → ∞)
is optimal, resulting in a discontinuity in the plot. The parameters of the model
correspond to those in Figure 4.
F
i
g
u
r
e
6
:
C
h
a
n
g
e
i
n
S
o
v
e
r
e
i
g
n
a
n
d
B
a
n
k
C
D
S
b
e
f
o
r
e
B
a
n
k
B
a
i
l
o
u
t
s

5
0 0
5
0
1
0
0
1
5
0
2
0
0
2
5
0
3
0
0
3
5
0
4
0
0
4
5
0
b a s i s   p o i n t s
S
o
v
e
r
e
i
g
n
 
C
D
S
B
a
n
k
 
C
D
S
F
i
g
u
r
e
6
p
l
o
t
s
t
h
e
s
o
v
e
r
e
i
g
n
C
D
S
a
n
d
b
a
n
k
C
D
S
f
o
r
I
r
e
l
a
n
d
i
n
t
h
e
p
e
r
i
o
d
f
r
o
m
3
/
1
/
2
0
0
7
t
o
8
/
3
1
/
2
0
1
0
.
T
h
e
b
a
n
k
C
D
S
i
s
c
o
m
p
u
t
e
d
a
s
t
h
e
u
n
w
e
i
g
h
t
e
d
a
v
e
r
a
g
e
o
f
b
a
n
k
C
D
S
f
o
r
b
a
n
k
s
h
e
a
d
q
u
a
r
t
e
r
e
d
i
n
I
r
e
l
a
n
d
(
A
l
l
i
e
d
I
r
i
s
h
B
a
n
k
,
A
n
g
l
o
I
r
i
s
h
B
a
n
k
,
B
a
n
k
o
f
I
r
e
l
a
n
d
,
a
n
d
I
r
i
s
h
L
i
f
e
a
n
d
P
e
r
m
a
n
e
n
t
)
.
T
h
e
d
a
t
a
a
r
e
f
r
o
m
D
a
t
a
s
t
r
e
a
m
.
F
i
g
u
r
e
7
:
C
h
a
n
g
e
i
n
S
o
v
e
r
e
i
g
n
a
n
d
B
a
n
k
C
D
S
d
u
r
i
n
g
t
h
e
P
e
r
i
o
d
o
f
B
a
n
k
B
a
i
l
o
u
t
s

2
5
0

2
0
0

1
5
0

1
0
0

5
0 0
5
0
1
0
0
A
T
A
U
B
E
C
H
D
E
D
K
E
S
F
R
G
B
G
R
I
E
I
T
N
L
N
O
P
T
S
E
b a s i s   p o i n t s
S
o
v
e
r
e
i
g
n
 
C
D
S
B
a
n
k
 
C
D
S
F
i
g
u
r
e
7
p
l
o
t
s
t
h
e
c
h
a
n
g
e
i
n
a
v
e
r
a
g
e
b
a
n
k
C
D
S
a
n
d
s
o
v
e
r
e
i
g
n
C
D
S
f
o
r
W
e
s
t
e
r
n
E
u
r
o
p
e
a
n
c
o
u
n
t
r
i
e
s
i
n
t
h
e
p
e
r
i
o
d
f
r
o
m
9
/
2
6
/
2
0
0
8
t
o
1
0
/
2
1
/
2
0
0
8
.
T
h
e
b
a
n
k
C
D
S
i
s
c
o
m
p
u
t
e
d
a
s
t
h
e
u
n
w
e
i
g
h
t
e
d
a
v
e
r
a
g
e
o
f
b
a
n
k
C
D
S
f
o
r
b
a
n
k
s
h
e
a
d
q
u
a
r
t
e
r
e
d
i
n
t
h
a
t
c
o
u
n
t
r
y
.
T
h
e
d
a
t
a
a
r
e
f
r
o
m
D
a
t
a
s
t
r
e
a
m
(
n
o
d
a
t
a
a
v
a
i
l
a
b
l
e
f
o
r
S
w
i
t
z
e
r
l
a
n
d
C
o
u
n
t
r
y
C
D
S
a
n
d
G
r
e
e
k
b
a
n
k
s
C
D
S
d
u
r
i
n
g
t
h
i
s
p
e
r
i
o
d
)
.
F
i
g
u
r
e
8
:
C
h
a
n
g
e
i
n
S
o
v
e
r
e
i
g
n
a
n
d
B
a
n
k
C
D
S
a
f
t
e
r
B
a
n
k
B
a
i
l
o
u
t
s

2
0
0 0
2
0
0
4
0
0
6
0
0
8
0
0
1
0
0
0
1
2
0
0
S
o
v
e
r
e
i
g
n
 
C
D
S
B
a
n
k
 
C
D
S
F
i
g
u
r
e
8
p
l
o
t
s
t
h
e
c
h
a
n
g
e
i
n
a
v
e
r
a
g
e
b
a
n
k
C
D
S
a
n
d
s
o
v
e
r
e
i
g
n
C
D
S
f
o
r
W
e
s
t
e
r
n
E
u
r
o
p
e
a
n
c
o
u
n
t
r
i
e
s
i
n
t
h
e
p
e
r
i
o
d
f
r
o
m
1
0
/
2
2
/
2
0
0
8
t
o
6
/
3
0
/
2
0
1
0
.
T
h
e
b
a
n
k
C
D
S
i
s
c
o
m
p
u
t
e
d
a
s
t
h
e
u
n
w
e
i
g
h
t
e
d
a
v
e
r
a
g
e
o
f
b
a
n
k
C
D
S
f
o
r
b
a
n
k
s
h
e
a
d
q
u
a
r
t
e
r
e
d
i
n
t
h
a
t
c
o
u
n
t
r
y
.
T
h
e
d
a
t
a
a
r
e
f
r
o
m
D
a
t
a
s
t
r
e
a
m
.
Figure 9: Correlation between Sovereign CDS and Public Debt before and after bank bailouts


Figure 9 shows the correlation between sovereign CDS and public liabilities (as a per-
centage of GDP) for Western European countries before and after the bank bailouts.
The top figure shows no correlation before the bailouts (as of 3/1/2007). The bottom
figure shows a strong correlation after the bank bailouts (as of 3/1/2010). The data
are from Datastream and the OECD Economic database.
Table 1: Summary Statistics
The sample show cross-sectional and time-series summary statistics for the bank risk sample. The sample
includes all European, U.S., and Australian banks with available data on bank CDS and share prices. The
data are the weekly level. Panel A shows summary statistics for the week of July 1
st
, 2007. Panel B
shows summary statistics for the period from January 1
st
, 2007 to August 31
st
, 2010.

Panel A: Cross-Section (7/1/2007)
# Mean Std.Dev
50
th

Percentile
5
th

Percentile
95
th

Percentile
Assets ($ billion)
81 589.3
594.9 362.6 77.9 1,896.9
Equity ($ billion)
81 26.8
29.3 20.2 3.5 112.4
Equity Ratio (%) 81 5.1 2.7 4.8 1.9 10.0
Tier 1 Ratio (%)
66 8.5
1.9 8.2 6.5 12.4
Bank CDS (bp)
75 21.8
13.4 10.5 6.5 41.0
Sovereign CDS (bp)
56 6.6 11.5 2.0 1.5 52.1
Panel B: Time-Series
Pre-Bailout Period (1/1/2007-8/31/2008)
# Mean Std.Dev
50
th

Percentile
5
th

Percentile
95
th

Percentile
Bank CDS (bp)
3,633 98.2 159.4 65.5 6.6 293.3
Sovereign CDS (bp)
3,633 12.1 10.9 8.8 1.6 37.2
CDS Volatility
3,633 0.001 0.028 0.001 -0.038 0.042
Bank Stock Return (%)
2,859 -0.008 0.079 -0.004 -0.102 0.082
∆ Bank CDS (%)
3,630 0.027 0.249 0.017 -0.281 0.324
∆ Sovereign CDS (%)
3,610 0.016 0.341 0.018 -0.302 0.333
∆ CDS Market Index (%)
3,633 0.015 0.138 0.000 -0.199 0.217
Bailout Period (9/1/2008-10/31/2008)
Bank CDS (bp)
606 301.3 615.1 138.7 65.9 727.5
Sovereign CDS (bp)
606 33.6 21.9 28.2 9.5 81.8
CDS Volatility
606 0.067 0.103 0.058 -0.079 0.278
Bank Stock Return (%)
455 -0.074 0.246 -0.033 -0.364 0.140
∆ Bank CDS (%)
605 0.041 0.378 0.032 -0.515 0.518
∆ Sovereign CDS (%)
606 0.137 0.177 0.101 -0.095 0.519
∆ CDS Market Index (%)
606 0.055 0.168 0.043 -0.267 0.373
Post-Bailout (31/10/2008 - 31/8/2010)
Bank CDS (bp)
6,496 194.1 178.9 131.5 65.5 567.8
Sovereign CDS (bp)
6,496 90.8 100.0 58.7 24.7 244.7
CDS Volatility
6,496 -0.002 0.034 -0.002 -0.045 0.043
Bank Stock Return (%)
4,814 0.003 0.100 0.003 -0.122 0.136
∆ Bank CDS (%)
6,495 0.001 0.129 -0.003 -0.174 0.174
∆ Sovereign CDS (%)
6,495 0.000 0.122 -0.002 -0.207 0.208
∆ CDS Market Index (%)
6,496 -0.005 0.089 -0.009 -0.124 0.130
Table 2: Emergence of Sovereign Credit Risk
This table shows the relation between sovereign credit risk, public debt, and bank quality. The sample includes all Eurozone countries and
Australia, Denmark, Great Britain, Sweden, and Switzerland with publicly available data on sovereign and bank CDS. The data are at the country-
level. The independent variable in Columns (1) and (2) is the sovereign CDS before the bank bailouts (as of 1/1/2008). Columns (1) and (2)
control for Public Debt measured as General Government Gross Financial Liabilities as percentage of GDP (collected from the OECD Economic
Outlook). Column (2) controls for average bank quality measured as the average banks CDS before the bank bailouts (as of 9/22/2008). The
independent variable in Columns (3) and (4) is the sovereign CDS after the bank bailouts (as of 3/31/2010). The dependent variables are the same
as in Columns (1) and (2), respectively. The independent variable in column (5) is the change in public debt from June 2008 to June 2010. The
dependent variable is the average bank quality. The independent variable in column (6) is public debt in June 2010. The dependent variables are
the public debt in June 2008 and average bank quality. We report robust standard errors. ** 1% significant, * 5% significant, and + 10%
significant

Log (Sovereign CDS) % Public Debt
Pre-Bailout Post-Bailout Around Bailout Post-Bailout
1/1/2008 3/31/2010 ∆ 2010-2008 3/31/2010

% Public Debt (June 2008) 0.006 0.005 0.015* 0.013+ 1.107**
(0.004) (0.005) (0.006) (0.007) (0.144)
Log (Average Bank CDS Sep 2008) 0.311 0.965* 20.118+ 21.726+
(0.208) (0.357) (10.168) (11.555)
Constant 2.137** 0.601 3.112** -1.593 -86.920 -101.548
(0.320) (1.154) (0.401) (2.019) (49.456) (60.923)
Observations 15 14 17 15 15 15
R-squared 0.134 0.171 0.261 0.488 0.364 0.843



Table 3: Change in Bank and Sovereign Credit Risk
This table shows the effect of sovereign credit risk on bank credit risk during the financial crisis. The sample includes all European, U.S., and
Australian banks with available data on bank CDS and share prices. The data are at the weekly level. Columns (1) to (3) cover the pre-bailout
period (1/1/2007-31/8/2010), Columns (4) to (6) cover the bailout period (9/1/2008-10/31/2008), and Columns (7) to (9) cover the post-bailout
period (November 2008 to August 2010). The dependent variable is the weekly change in the natural logarithm of bank CDS. The main
independent variable is the weekly change in the sovereign CDS. The sovereign CDS is assigned based on the country where the bank is
headquartered. The control variables are the change in CDS market index, volatility, and bank stock return. Columns (2), (3), (5), (6), (8), and (9)
include week fixed effects. Column (3), (6), (9) include bank fixed effects and interactions of bank fixed effects with volatility, bank stock return,
and the CDS Market index. The standard errors are clustered at the bank-level. ** 1% significant, * 5% significant, and + 10% significant

∆ Log(Bank CDS)
Period Pre-Bailout (Jan 07-Aug 08) Around Bailout (Sep-Oct 08) Post-Bailout (Nov 08-Sep 10)
(1) (2) (3) (4) (5) (6) (7) (8) (9)
∆ Log(Sovereign CDS)
0.023* 0.015 0.019 0.026 -0.403+ -0.430 0.163** 0.079** 0.080**
(0.010) (0.014) (0.015) (0.082) (0.232) (0.287) (0.019) (0.030) (0.027)
∆ Log(CDS Market Index) 0.860** 0.932** 0.689**
(0.041) (0.094) (0.027)
∆ Volatility Index
0.214 -0.539** 0.122*
(0.155) (0.081) (0.050)





Week FE N Y Y N Y Y N Y Y
Interactions N N Y N N Y N N Y
Observations
3,508 3,508 3,508 577 577 577 7,086 7,086 7,086
Banks
84 84 84 71 71 71 84 84 84
R-squared
0.171 0.253 0.387 0.134 0.308 0.504 0.316 0.384 0.441

Table 4: Change in Bank and Sovereign Credit Risk
This table shows the effect of sovereign credit risk on bank credit risk during the financial crisis. The sample includes all European, U.S., and
Australian banks with available data on bank CDS and share prices. The data are at the weekly level. Columns (1) to (3) cover the pre-bailout
period (1/1/2007-31/8/2010), Columns (4) to (6) cover the bailout period (9/1/2008-10/31/2008), and Columns (7) to (9) cover the post-bailout
period (November 2008 to August 2010). The dependent variable is the weekly change in the natural logarithm of bank CDS. The main
independent variable is the weekly change in the sovereign CDS. The sovereign CDS is assigned based on the country where the bank is
headquartered. The control variables are the change in CDS market index, volatility, and bank stock return. Columns (2), (3), (5), (6), (8), and (9)
include week fixed effects. Column (3), (6), (9) include bank fixed effects and interactions of bank fixed effects with volatility, bank stock return,
and the CDS Market index. The standard errors are clustered at the bank-level. ** 1% significant, * 5% significant, and + 10% significant

∆ Log(Bank CDS)
Period Pre-Bailout (Jan 07-Aug 08) Around Bailout (Sep-Oct 08) Post-Bailout (Nov 08-Sep 10)
(1) (2) (3) (4) (5) (6) (7) (8) (9)
∆ Log(Sovereign CDS)
0.019* 0.008 0.014 -0.020 -0.236 -0.235 0.150** 0.100** 0.105**
(0.009) (0.013) (0.015) (0.096) (0.153) (0.200) (0.025) (0.034) (0.030)
Bank Stock Return
-0.142 -0.062 -0.255+ -0.295* -0.174** -0.154**
(0.118) (0.106) (0.132) (0.147) (0.034) (0.036)
∆ Log(CDS Market Index)
0.929** 0.848** 0.662**
(0.043) (0.123) (0.032)
∆ Volatility Index
0.043 -0.711** 0.030
(0.120) (0.096) (0.051)





Week FE N Y Y N Y Y N Y Y
Interactions N N Y N N Y N N Y
Observations
2,745 2,745 2,745 437 437 437 5,278 5,278 5,278
Banks
63 63 63 53 53 53 60 60 60
R-squared
0.224 0.308 0.481 0.208 0.403 0.728 0.359 0.424 0.491

Table 5: Change in Bank and Sovereign Credit Risk (by Tier 1 Capital)
This table shows the effect of sovereign credit risk on bank credit risk during the financial crisis. The sample includes all European, U.S., and
Australian banks with available data on bank CDS and share prices. The data are at the weekly level. Columns (1) to (3) cover the pre-bailout
period (1/1/2007-31/8/2010), Columns (4) to (6) cover the bailout period (9/1/2008-10/31/2008), and Columns (7) to (9) cover the post-bailout
period (November 2008 to August 2010). The dependent variable is the weekly change in the natural logarithm of bank CDS. The Tier 1 capital
ratio is the regulatory bank capital ratio. All other variables are defined in Tables 3 to 5. The regression includes the Tier 1 capital ratio and an
interaction between the Tier1 capital ratio and the change in sovereign CDS. All other controls are the same as in Table5. The standard errors are
clustered at the bank-level. ** 1% significant, * 5% significant, and + 10% significant

∆ Log(Bank CDS)
Period Pre-Bailout (Jan 07-Aug 08) Around Bailout (Sep-Oct 08) Post-Bailout (Nov 08-Sep 10)
(1) (2) (3) (4) (5) (6) (7) (8) (9)
∆ Log(Sovereign CDS)*Tier 1
0.648 0.340 1.014 -6.265* -6.730+ -4.011 -1.765 -1.610 -2.824*
(0.882) (0.712) (1.035) (2.493) (3.823) (12.976) (1.470) (1.423) (1.360)
∆ Log(Sovereign CDS)
-0.031 -0.019 -0.079 0.575* 0.437 0.032 0.337* 0.317* 0.368**
(0.072) (0.066) (0.096) (0.237) (0.336) (1.182) (0.127) (0.126) (0.118)
Tier 1
-15.118 -13.492 61.937 67.087 -1.589 -2.114
(30.883) (32.400) (108.792) (78.028) (8.224) (8.568)
Stock Return -0.347** -0.335** -0.117 -0.054 -0.165** -0.134**
(0.095) (0.115) (0.139) (0.140) (0.037) (0.043)
Other Controls Y Y Y Y Y Y Y Y Y
Week FE N Y Y N Y Y N Y Y
Interactions N N Y N N Y N N Y
Observations
2,256 2,256 2,256 351 351 351 4,163 4,163 4,163
Bank
48 48 48 41 41 41 47 47 47
R-squared
0.261 0.345 0.510 0.353 0.528 0.804 0.380 0.449 0.517

Table 6: Bank Stock Return and Change in Sovereign Credit Risk
This table shows the effect of sovereign credit risk on bank stock returns during the financial crisis. The sample includes all European, U.S., and
Australian banks with available data on bank CDS and share prices. The data are at the weekly level. Columns (1) to (3) cover the pre-bailout
period (1/1/2007-31/8/2010), Columns (4) to (6) cover the bailout period (9/1/2008-10/31/2008), and Columns (7) to (9) cover the post-bailout
period (November 2008 to August 2010). The dependent variable is the weekly bank stock return. The main independent variable is the weekly
change in the sovereign CDS. The sovereign CDS is assigned based on the country where the bank is headquartered. The control variables are the
change in CDS market index and volatility. Columns (2), (3), (5), (6), (8), and (9) include week fixed effects. Column (3), (6), (9) include bank
fixed effects and interactions of bank fixed effects with volatility and the CDS Market index. The standard errors are clustered at the bank-level.
** 1% significant, * 5% significant, and + 10% significant

Bank Stock Return
Period Pre-Bailout (Jan 07-Aug 08) Around Bailout (Sep-Oct 08) Post-Bailout (Nov 08-Sep 10)
(1) (2) (3) (4) (5) (6) (7) (8) (9)
∆ Log(Sovereign CDS)
-0.011** -0.002 -0.002 -0.040 0.041 0.114 -0.177** -0.054* -0.068**
(0.004) (0.002) (0.002) (0.035) (0.075) (0.114) (0.026) (0.026) (0.026)
∆ Log(CDS Market Index)
-0.106** -0.474** -0.243**


(0.015) (0.078) (0.017)


∆ Volatility Index
-0.368** -0.317** -0.761**

(0.070) (0.082) (0.057)


Week FE N Y Y N Y Y N Y Y
Bank FE N N Y N N Y N N Y
Interactions N N Y N N Y N N Y
Observations
2,895 2,895 2,895 446 446 446 5,324 5,324 5,324
Banks
65 65 65 54 54 54 60 60 60
R-squared
0.070 0.240 0.311 0.118 0.212 0.564 0.285 0.488 0.533
Table 7: Summary Statistics of European Bank Stress Test Sample
The table shows summary statistics for all banks that participated in the EU Bank Stress Tests from July
2010. The data was collected from the website of the Committee of European Banking Regulators and
nation websites of the respective bank regulators. The sovereign holdings are computed as the total value
of sovereign holdings relative to risk-weighted assets. We report both the gross and net exposure as
reported to bank regulators. The share of trading book and banking book are the share of sovereign
holdings held in the respective book. The shares are computed based on gross exposure (net exposure
was not reported).

Sovereign Holdings
Euro Bank Stress Tests Sample, March 31, 2010
N Mean Std.Dev
50th
Percentile
5th
Percentile
95th
Percentile
(1) (2) (3) (4) (5) (6)

Bank Characteristics

Risk-weighted Assets (EUR
million) 91 126,337 179,130 63,448 3,269 493,307
Tier 1 Capital Ratio (%) 91 10.2 2.4 9.8 7.2 14.4
Sovereign Exposure

Sovereign Holdings (gross,
EUR million) 91 20,668 27,948 7,930 105 81,765

Sovereign Holdings (net,
EUR million) 91 19,719 27,329 6,960 105 78,959

Home Sovereign Holdings
(gross, EUR million) 91 11,493 14,422 5,774 182 42,800

Home Sovereign Holdings
(net, EUR million) 91 11,023 13,956 5,348 117 42,800
Home Share (%) 91 69.4 30.0 81.6 18.9 100
Greek Sovereign Holdings 91 669 2,844 0 0 5,601
Share Banking Book (%) 91 84.9 19.9 92.2 35.4 100.0



Table 8: Summary Statistics of European Bank Stress Test Sample

The table shows regression of change in bank CDS on change in exposure to sovereign bank holdings.
The sovereign bond holdings data were collected from the website of the Committee of European
Banking Regulators and nation websites of the respective bank regulators. We construct the exposure
variable as the weighted average of country CDS with sovereign holdings as weights. Changes are
computed as log changes. The data covers the period from 3/1/2010 to 4/30/2010. Columns (2), (5) and
(6) include bank fixed effects. Column (3) includes week fixed effects. Column (4) to (6) include day
fixed effect. The exposure variable in Column (6) excludes German bonds. The standard errors are
clustered at the bank-level (51 banks). ** 1% significant, * 5% significant, and +10% significant

Change in Bank CDS
Sample All All All All All
Excluding
Germany
(1) (2) (3) (4) (5) (6)

Change in Sovereign
Exposure 0.325** 0.326** 0.261** 0.141** 0.135** 0.137**
(0.027) (0.028) (0.027) (0.049) (0.046) (0.046)
Bank FE N Y N N Y Y
Week FE N N Y N N N
Day FE N N N Y Y Y
Observations 2,317 2,317 2,317 2,317 2,317 2,317
Banks 51 51 51 51 51 0.357
R-squared 0.173 0.188 0.228 0.342 0.357 0.357
Adjusted R-Squared 0.173 0.170 0.224 0.329 0.329 0.329

A Pyrrhic Victory? – Bank Bailouts and Sovereign Credit Risk

Abstract We show that financial sector bailouts and sovereign credit risk are intimately linked. A bailout benefits the economy by ameliorating the under-investment problem of the financial sector. However, increasing taxation of the corporate sector to fund the bailout may be inefficient since it weakens its incentive to invest, decreasing growth. Instead, the sovereign may choose to fund the bailout by diluting existing government bondholders, resulting in a deterioration of the sovereign’s creditworthiness. This deterioration feeds back onto the financial sector, reducing the value of its guarantees and existing bond holdings and increasing its sensitivity to future sovereign shocks. We provide empirical evidence for this two-way feedback between financial and sovereign credit risk using data on the credit default swaps (CDS) of the Eurozone countries for 2007-10. We show that the announcement of financial sector bailouts was associated with an immediate, unprecedented widening of sovereign CDS spreads and narrowing of bank CDS spreads; however, post-bailouts there emerged a significant co-movement between bank CDS and sovereign CDS, even after controlling for banks’ equity performance, the latter being consistent with an effect of the quality of sovereign guarantees on bank credit risk. J.E.L. Classification: G21, G28, G38, E58, D62. Keywords: financial crises, forbearance, deleveraging, sovereign debt, growth, credit default swaps

1

1

Introduction

Just two and a half years ago, there was essentially no sign of sovereign credit risk in the developed economies and a prevailing view was that this was unlikely to be a concern for them in the near future. Recently, however, sovereign credit risk has become a significant problem for a number of developed countries. In this paper, we are motivated by three closely related questions surrounding this development. First, were the financial sector bailouts an integral factor in igniting the rise of sovereign credit risk in the developed economies? We show that they were. Second, what was the exact mechanism that caused the transmission of risks between the financial sector and the sovereign? We propose a theoretical explanation wherein the government can finance a bailout through both increased taxation and via dilution of existing government debt-holders. The bailout is beneficial; it alleviates a distortion in the provision of financial services. However, both financing channels are costly. When the optimal bailout is large, dilution can become a relatively attractive option, leading to deterioration in the sovereign’s creditworthiness. Finally, we ask whether there is also feedback going in the other direction– does sovereign credit risk feedback on to the financial sector? We explain – and verify empirically – that such a feedback is indeed present, due to the financial sector’s implicit and explicit holdings of sovereign bonds. Our results call into question the usually implicit assumption that government resources are vastly deep and that the main problem posed by bailouts is primarily that of moral hazard – that is, the distortion of future financial sector incentives. While the moral hazard cost is certainly pertinent, our conclusion is that bailout costs are not just in the future. They are tangible right around the timing of bailouts and are priced into the sovereign’s credit risk and cost of borrowing. Thus, aggressive bailout packages that stabilize financial sectors in the short run but ignore the ultimate taxpayer cost might end up being a Pyrrhic victory. Motivation: The case of Irish bailout. On September 30, 2008 the government of Ireland announced that it had guaranteed all deposits of the six of its biggest banks. The immediate reaction that grabbed newspaper headlines the next day was whether such a policy of a full savings guarantee was anti-competitive in the Euro area. However, there was something deeper manifesting itself in the credit default swap (CDS) markets for purchasing protection against the sovereign credit risk of Ireland and that of its banks. Figure 1 shows that while the cost of purchasing such protection on Irish banks – their CDS fee – fell overnight from around 400 basis points to 150 basis points, the CDS fee for the Government 2

of Ireland’s credit risk rose sharply. Over the next month, this rate more than quadrupled to over 100 basis points and within six months reached 400 basis points, the starting level of its financial firms’ CDS. While there was a general deterioration of global economic health over this period, the event-study response in Figure 1 suggests that the risk of the financial sector had been substantially transferred to the government balance sheet, a cost that Irish taxpayers must eventually bear. Viewed in the Fall of 2010, this cost rose to dizzying heights prompting economists to wonder if the precise manner in which bank bailouts were awarded have rendered the financial sector rescue exorbitantly expensive. Just one of the Irish banks, Anglo Irish, has cost the government up to Euro 25 bln (USD 32 bln), amounting to 11.26% of Ireland’s Gross Domestic Product (GDP). Ireland’s finance minister Brian Lenihan justified the propping up of the bank “to ensure that the resolution of debts does not damage Ireland’s international credit-worthiness and end up costing us even more than we must now pay.” Nevertheless, rating agencies and credit markets revised Ireland’s ability to pay future debts significantly downward. The original bailout cost estimate of Euro 90 bln was re-estimated to be 50% higher and the Irish 10-year bond spread over German bund widened significantly, ultimately leading to a bailout of Irish government by the stronger Eurozone countries.1 This episode is not isolated to Ireland though it is perhaps the most striking case. In fact, a number of Western economies that bailed out their banking sectors in the Fall of 2008 have experienced, in varying magnitudes, similar risk transfer between their financial sector and government balance-sheets. Our paper develops a theoretical model and provides empirical evidence that help understand this interesting phenomenon. Model. Our theoretical model consists of two sectors of the economy – “financial” and “corporate” (more broadly this includes also the household and other non-financial parts of the economy), and a government. The two sectors contribute jointly to produce aggregate output: the corporate sector makes productive investments and the financial sector invests in intermediation “effort” (e.g., information gathering and capital allocation) that enhance the return on corporate investments. Both sectors, however, face a potential under-investment problem. The financial sector is leveraged (in a crisis, it may in fact be insolvent) and underinvests in its contributions due to the well-known debt overhang problem (Myers, 1977). For simplicity, the corporate sector is un-levered. However, if the government undertakes a
See “Ireland’s banking mess: Money pit – Austerity is not enough to avoid scrutiny by the markets”, the Economist, Aug 19th 2010; “S&P downgrades Ireland” by Colin Barr, CNNMoney.com, Aug 24th 2010; and, “Ireland stung by S&P downgrade”, Reuters, Aug 25th, 2010.
1

3

“bailout” of the financial sector, in other words, makes a transfer from the rest of the economy that results in a net reduction of the financial sector debt, then the transfer must be funded in the future (at least in part) through taxation of corporate profits. Such taxation, assumed to be proportional to corporate sector output, induces the corporate sector to under-invest. A government that is fully aligned with maximizing the economy’s current and future output determines the optimal size of the bailout. We show that tax proceeds that can be used to fund the bailout have, in general, a Laffer curve property, so that the optimal bailout size and tax rate are interior. The optimal tax rate that the government is willing to undertake for the bailout is greater when the financial sector’s debt overhang is higher and its relative contribution (or size) in output of the economy is larger. In practice, governments fund bailouts in the short run by borrowing or issuing bonds, which are repaid by future taxation. There are two interesting constraints on the bailout size that emerge from this observation. One, the greater is the legacy debt of the government, the lower is its ability to undertake a bailout. This is because the Laffer curve of tax proceeds leaves less room for the government to increase tax rates for repaying its bailout-related debt. Second, the announcement of the bailout lowers the price of government debt due to the anticipated dilution from newly issued debt. Now, if the financial sector of the economy has assets in place that are in the form of government bonds, the bailout is in fact associated with some “collateral damage” for the financial sector itself. Illustrating the possibility of such a two-way feedback is a novel contribution of our model. To get around these constraints, the government can potentially undertake a strategic default. Assuming that there are some deadweight costs of such default, for example, due to international sanctions or from being unable to borrow in debt markets for some time, we derive the optimal boundary for sovereign default as a function of its legacy debt and financial sector liabilities. This boundary explains that a heavily-indebted sovereign faced with a heavily-insolvent financial sector will be forced to “sacrifice its credit rating” to save the financial sector and at the same time sustain economic growth. We then extend the model to allow for uncertainty about the realized output growth of the corporate sector.This introduces a possibility of solvency-based default on government debt. Interestingly, given the collateral damage channel, an increase in uncertainty about the sovereign’s economic output not only lowers its own debt values but also increases the financial sector’s risk of default. This is because the financial sector’s government bond holdings fall in value, and in an extension of the model, so do the value of the government guarantees accorded to the financial sector as a form of bailout. In turn, these channels 4

we examine sovereign and bank CDS in the period from 2007 to 2010 and find three distinct periods. This evidence is consistent with a significant increase in the default risk of the banking sector with little effect on sovereigns in the prebailout period. The second period covers the bank bailouts starting with the announcement of a bailout in Ireland in late September 2008 and ending with a bailout in Sweden in late October 2008. We document that on average Eurozone banks stress-tested in 2010 held Eurozone government bond holdings that were as large as onesixth of their risk-weighted assets. we see a large. sustained rise in bank CDS as the financial crisis develops. we find a significant decline in bank CDS across all countries and a corresponding increase in sovereign CDS. This evidence suggests that the banks and sovereigns share the default risk after the announcement of banks bailouts and that the risk is increasing in the relative size of countries’ public debt.induce a post-bailout co-movement between the financial sector’s credit risk and that of the sovereign. The third period covers the period after the bank bailouts and until 2010. We find that both sovereign and bank CDS increased during this period and that the increase was larger for countries with significant public debt ratios. The first period covers the start of the financial crisis in January 2007 until the bankruptcy of Lehman Brothers. We confirm all of the above results also in our regression analysis linking levels and changes in financial sector CDS to levels and changes. Finally. However. we show that bank CDS co-move with sovereign CDS on banks’ holdings 5 . This evidence suggests that bank bailouts produced a transfer of default risk from the banking sector to the sovereign. we collect bank-level data on holdings of different sovereign government bonds released as part of the bank stress tests conducted for European banks in 2010. In our non-parametric analysis. Our empirical work analyzes this two-way feedback between the financial sector and sovereign credit risk. Consistent with the economic importance of the collateral damage channel. We also confirm model’s implications that banks with higher leverage experience a stronger relationship between sovereign and bank credit risks after the bailouts. respectively. Empirics. During this one-month period. in support of the collateral damage channel as being potentially relevant for the co-movement between financial sector and sovereign CDS. of sovereign CDS in the three periods. sovereign CDS spreads remains very low. Our analysis focuses mainly on the Western European economies during the financial crisis of 2007-10. Across all Western economies. even though the immediate effect of the bailout is to lower the financial sector’s credit risk and raise that of the sovereign.

L1 denotes the liabilities of the financial sector. Freddie Mac) and perhaps also the value of explicit and implicit government guarantees or support. In addition. while A1 represents the payoff of the other assets held by the financial sector.g. the variable T0 represents the value of the time 0 transfer made by the government to the financial sector and is discussed further below. All agents are risk-neutral. To produce s0 units. denoted A1 and AG . Financial sector: The operator of the financial sector solves the following problem. 2 Model There are three time periods in the model: t = 0. 2 While we refer to government claims principally as government bonds. The payoff and value of government bonds is discussed below. It receives the payoff from its efforts only if the value of assets exceeds liabilities at t = 1. with ws determined in equilibrium. as a continuously valued random variable that is realized at t = 1 and takes values in [0. The productive economy consists of two parts. AG is the value of ˜ the financial sector’s holdings of a fraction kA of outstanding government bonds. and 2. which are due (mature) at t = 1. We assume that c (s0 ) > 0 and c (s0 ) > 0. a financial sector and a non-financial sector. Section 2 presents our theoretical analysis. Section 6 concludes. Finally. All proofs not in the main text are in the online Appendix. ∞). Section 5 discusses the related literature. the amount of financial services to supply at t = 1 in order to maximize his expected payoff at t = 1. which is to choose. at t = 0.. which is risky. net of the effort cost required to produce these services: max E0 s s0 ˜ ws ss − L1 + A1 + AG + T0 × 1{−L1 +A1 +AG +T0 >0} − c(ss ) . Fannie Mae. The remainder of the paper is organized as follows. a broader interpretation can include claims on quasi-governmental agencies (e. Section 4 provides empirical evidence. The 0 financial sector earns revenues at the rate of ws per unit of financial service supplied. There ˜ are two types of assets held by the financial sector. The financial sector has both liabilities and assets on its books. This solvency condition ˜ is given in equation (1) by the indicator function for the expression −L1 + A1 + AG + T0 > 0. 1.of government bonds. 6 . there is a government and a representative consumer. the operator of the financial sector expends c(s0 ) units of effort. ˜ 0 0 (1) where ss is the amount of financial services supplied by the financial sector at t = 1.2 We model the payoff ˜ A1 .

whose payoff is realized at t = 2. the government issues bonds. both new and outstanding. debtholders receive ownership of all financial sector assets and wage revenue. which induces it to supply more financial services. We let ND denote the number of bonds that the government has issued in the past – its outstanding stock of debt. at an incremental cost of (K1 − K0 ). The expectation at t = 1 of ˜ this payoff is V (K1 ) = E1 [V (K1 )] and. that it then transfers to the balance sheet of the financial sector. In particular. Government: The government’s objective is to maximize the total output of the economy and hence the welfare of the consumer. A proportion θ0 of the payoff of the continuation project is taxed by the government to pay its debt. sd ) − ws sd + (1 − θ0 )V (K1 ) − (K1 − K0 ) 0 0 (2) sd . and the capital stock.In case of insolvency. sd . 3 7 . These bonds are repaid with taxes levied on the non-financial sector at a tax-rate of θ0 . At t = 1. which would provide an additional channel for wages to feed back into the probability of solvency. This project represents the future or continuation value of the non-financial sector and is in general subject to uncertainty. as we explain next. It does this by reducing the debt-overhang problem of the financial sector. which occur at t = 1 and t = 2: ˜ max E0 f (K0 . ND . the non-financial sector is faced with a decision of ˜ how much capital K1 to invest. We assume that V (K1 ) > 0 and V (K1 ) < 0.3 Non-financial sector: The non-financial sector comes into t = 0 with an existing capital stock K0 . the tax rate θ0 is set by the government at t = 0 ˜ and is levied at t = 2 upon realization of the payoff V (K1 ). so the face value of outstanding debt equals the number of bonds. K 1 0 The function f is the production function of the non-financial sector. which takes as inputs at t = 0 the financial services it demands. to produce consumption 0 goods at t = 1. For simplicity. we assume that f is increasing in both arguments and concave. we choose to abstract from this to avoid the additional complexity. The government issues NT new Note that we could include the wage revenues in the solvency indicator function. Although such a channel would reinforce the mechanism at work in the model. in a project V . K0 . is a function of the investment K1 . Its objective is to maximize the sum of the expected values of its net payoffs. thereby increasing output. To achieve this. We assume that the government credibly commits to this tax rate. so that the expected payoff is increasing but concave in investment. bonds have a face value of one. as indicated. Moreover. The output of f is deterministic.

In that case.bonds to accomplish the transfer to the financial sector. respectively. it pays out all the tax revenue raised to bondholders. On the other hand. the consumer chooses optimal portfolio allocations. NT where s0 is the equilibrium provision of financial services. {ni }. government bond holdings. it will have less to lose from default. just rebates them to the consumer. let P0 denotes the price of government bonds. Since the representative consumer is assumed to be risk-neutral. If a country’s reputation is already weak. At t = 0. which is determined in equilibrium. degradation of the legal system and so forth. Let P (i) and P (i) denote the price and payoff of asset i. He solves a simple consumption and portfolio choice problem by allocating his wealth W between consumption. Hence. s0 ) + V (K1 ) − c(s0 ) − (K1 − K0 ) − 1def D + A1 (3) θ0 . 4 8 . that solve We can consider more generally a policy whereby the government pays only a fraction m of tax revenue ˜ and gives back any remaining tax funds to the consumer. In case of default. the sovereign incurs a fixed deadweight loss of D. This implies that the value of the financial sector’s holding of government bonds is AG = kA P0 ND . ˜ 5 Although D here is obviously reduced-form. if tax revenues fall short of NT + ND . who consumes the combined output of the financial and non-financial sector.4 We further assume that default incurs a deadweight loss.5 Finally. Hence. then the government defaults on its debt. We assume that the government credibly commits to this payout policy. and equity in the financial and non-financial sectors. the government faces the following problem: ˜ ˜ max E0 f (K0 . Consumer: The representative consumer consumes the output of the economy. asset prices equal the ex˜ pected values of asset payouts. We assume that if there are still tax revenues left over (a surplus). or equivalently. Since it will actually be optimal for the government to commit to paying bondholders all the tax revenue raised. Hence. one can think of the deadweight cost in terms of loss of government reputation internationally. Note that 1def is an indicator function that equals 1 if the government ˜ defaults (if θ0 V (K1 ) < NT + ND ) and 0 otherwise. loss of domestic government credibility. default is costly and there is an incentive to avoid it. The government’s objective is to maximize the expected utility of the representative consumer. This maximization is subject to the budget constraint: T0 = P0 NT and subject to the choices made by the financial and non-financial sectors. we restrict ourselves to the case m = 1. at t = 2 the government ˜ receives realized taxes equal to θ0 V (K1 ) and then uses them to pay bondholders NT + ND . the government spends them on programs for the representative consumer.

the first order condition of the financial sector can be written as: ws psolv − c (ss ) = 0 . 9 . 1 We assume that at the optimal ss the first-order condition is satisfied. 0 (5) ∞ ˜ ˜ where psolv ≡ A p(A1 )dA. From ˆ0 ˆ0 here on. we drop the superscripts on s0 and denote the equilibrium quantity of financial services simply by s0 . 3 Equilibrium Outcomes ˜ We begin by examining the maximization problem of the financial sector.sd ) (6) 0 0 Since f is concave in its arguments. 0 Consider now the problem of the non-financial sector at t = 0. P (i) = E0 [P (i)]. ∂ 2 sd 0 Henceforth. Its demand for financial services. 0 In equilibrium the demand and supply of financial services are the same: sd = ss .the following problem: ˜ max E0 Σi ni P (i) + (W − Σi ni P (i)) ni (4) The consumer’s first order condition gives the standard result that equilibrium price of an ˜ asset equals the expected value of its payoff. is the probability that the financial sector is solvent at t = 1. Then. sd . let A1 be the minimum realization of A1 for which the financial sector does not default: A1 = L1 − AG − T0 . sd ) 0 = ws . The parametric choice we will use 0 1 below for c(s0 ) is c(s0 ) = β m sm where m > 1. s0 ) = αK0 sϑ . Let p(A) denote ˜ ˜ the probability density of A. the second order condition is satisfied: < 0. d ∂s0 ∂ 2 f (K . Furthermore. The second-order ˆ0 condition of the financial sector’s problem is −c (ss ) < 0. is determined by its first-order condition: ˆ0 ∂f (K0 . we will parameterize f as Cobb-Douglas with the factor share of financial services 1−ϑ given by ϑ: f (K0 .

an increase in the tax rate increases the tax revenue at a rate equal to V (K1 ). which reduces the marginal revenue of a tax increase. K1 . Raising taxes has two effects. while the marginal cost. be denoted by T . θ0 . the tax distortion eliminates the incentive for investment and tax revenues are reduced to zero. It can be shown that since the production function V (K1 ) is concave. Hence.2 Tax Revenues: A Laffer Curve Next. At the extreme.3. the first-order conditions of the financial sector (5) and non-financial sector (6) show how debt-overhang impacts the provision of financial services by the financial sector. consider the first-order condition 10 . thereby leading to reduced investment. an increase in the tax rate θ0 captures a larger proportion of the future value of the non-financial sector.1 Transfer Reduces Underprovision of Financial Services Taken together. the impact on tax revenue of an increase in the tax rate is given by: dK1 ∂T = V (K1 ) + θ0 V (K1 ) ∂θ0 dθ0 Note that at θ0 = 0. The marginal benefit of an extra unit of financial services to the economy is given by ws . Let the expected tax revenue. This is given by the second term on the right-hand side of the expression. we first look at how expected tax revenue responds to the tax rate. Hence. An increase in the transfer T0 leads to an increase in the provision of financial services since this raises the probability psolv that the financial sector is solvent at t = 1. c (s0 ). when θ0 = 1. the future value of the non-financial sector. this reduces the incentive of the non-financial sector to invest in its future. The reason is that the possibility of liquidation psolv < 1 drives a wedge between the social and private marginal benefit of an increase in the provision of financial services. θ0 V (K1 ). To see this. This implies that the equilibrium allocation is sub-optimal. we obtain that Lemma 1. thereby raising tax revenues. On the one hand. The result is that as long as psolv < 1. On the other hand. tax revenues are non-monotonic in the tax rate and maximized by a tax rate strictly less than 1. as taxes are increased the incentive to invest is decreased by the tax rate. to understand the government’s problem. Formally. is less than ws if there is a positive probability of insolvency. 3. there is an under-provision of financial services relative to the first-best case (psolv = 1).

6 As γ γ Appendix A. To summarize.for investment of the non-financial sector at t = 1: (1 − θ0 )V (K1 ) − 1 = 0 (7) Since V (K1 ) < 0.3 shows. γ Henceforth. is increasing (dT /dθ0 > 0) max max 2 and concave (d2 T /dθ0 < 0) on [0. as it increases from zero (no taxes). NT + ND = T . In fact. T . (7) implies that T = θt+1 γ 1−γ (1 − θt+1 ) 1−γ . θ0 . θ0 ). increase in the tax rate. This functional form is a natural choice for an increasing and concave function of K1 . it can only issue a number of bonds NT that it can pay off in full. tax revenues satisfy the Laffer curve property as a function of the tax rate: Lemma 2. (2) we force the government to remain solvent. since every bond has a sure payoff of 1. If the government must remain solvent. 0 < γ < 1. By assumption (1). θ0 V (K1 ). Appendix A.2 provides a more structural motivation for this choice based on the calculation of a continuation value under our choice of production function. it must be the case that as the tax rate is increased. and hence by assumption (2). given its tax revenue. We make two simplifying assumptions: (1) we set to zero the variance of the ˜ realized future value of the non-financial sector. 1). since we know that at θ0 = 1 the tax revenue is zero. The tax revenues. the marginal tax revenue decreases until it eventually becomes negative. 3. Taking the derivative with respect to θ0 by using the Implicit Function theorem gives: dK1 V (K1 ) = <0 dθ0 (1 − θ0 )V (K1 ) which shows that as the tax rate is increased. and decreasing (dT /dθ0 < 0) on (θ0 . we know that the bond price is P0 = 1. 6 11 . the tax revenue is known exactly (it is equal to T ).3 Optimal Transfer Under Certainty and No Default We analyze next the government’s decision starting first with a simplified version of the general setup. we parameterize V with the functional form V (K1 ) = K1 . so that V (K1 ) = V (K1 ). Moreover. It can then be shown that: max Lemma 3. This calculation suggests that the continuation value implied by a multiperiod model should take a similar functional form. and then eventually decline. The tax revenue. is maximized at θ0 = 1 − γ. the second-order condition holds. the non-financial sector reduces investment. In subsequent sections we remove these assumptions.

7 Under-Investment Loss Due to Taxes: The term dL/dT in (8) is the marginal underinvestment loss to the economy due to a marginal increase in expected tax revenue. which distorts the non-financial sector’s incentive to invest. s0 ) ds0 = (1 − psolv ) dT ∂s0 dT0 dL dK1 =θ0 V (K1 ) dT dT which expresses the first-order condition in terms of the choice of transfer size and expected tax revenue. For illustration. when the financial sector is at high risk of insolvency and debt-overhang is significant. 8 The curve shows that −L is convex as raising additional tax revenues incurs an increasingly large marginal (8) 12 . the marginal gain from increasing tax revenue (and hence the transfer) will be large when psolv is low. rather than in terms of the tax rate. then dL/dT < 0. Appendix A. we have that the transfer to the financial sector is T0 = θ0 V (K1 ) − ND and there is no probability of default. and as the proof to Proposition 1 shows.4 shows that the first-order condition for the government can be written as: dG dL + =0 dT dT where dG ∂f (K0 . For illustration. Since dK1 /dT < 0. ds0 /dT0 > 0. and we can consider the problem directly at t = 0. Equation (7) shows that it reflects the size of the tax-induced distortion. the government’s information regarding expected tax revenue is the same at t = 0 as at t = 1. Equations (5) and (6) show that it reflects the wedge between the social and private benefits of an increase in financial services. E[1def ] = 0. the graph of dG/dT is given by the solid curve in the top panel of Figure 2. which is greater than zero as long as the tax rate is positive. Since.Under the two simplifying assumptions. Gain From Increased Provision of Financial Services: The term dG/dT in (8) is the marginal gain to the economy of increasing expected tax revenue. Since there is no change in the non-financial sector’s investment opportunities between t = 0 and t = 1. Hence. as shown above. a graph of −dL/dT is shown as the upward-sloping dashed curve in the top panel of Figure 2. G is concave in T since the marginal gain from increasing tax revenues (to increase the transfer) is decreasing. that is. As we explain below.8 7 As the graph indicates. this condition is intuitive since it equates the marginal gain and marginal loss of increasing tax revenue. the only choice variable for the government in this case is the tax rate.

this ˆ is at the intersection point of the two curves. 2. which is strictly less than θ0 . are increasing in the factor share of the financial sector. The optimal tax rate and revenue are increasing in L1 . If also m ≤ 2. 2 (no default) and m ≥ 2ϑ. max The optimal tax rate is less than θ0 due to the Laffer-curve property of tax revenues. revenue. Then newly issued sovereign debt has face value ˆ NT = T − ND and a price of P0 = 1. then the optimal tax rate will be strictly greater than zero. 1.e. and newly issued sovereign debt. Therefore. max ˆ Proposition 1. since a lower probability of solvency increases the distortion in the provision of financial services.5. For any level of transfer. in the top panel of Figure 2. the marginal gain available is greater the more severe is the debt-overhang. corresponding to the intersection point in the top panel. this marginal loss due to underinvestment worsens as T is increased. The face value of newly issued sovereign debt (the transfer) is increasing in the financial sector liabilities L1 . then the optimal tax rate. psolv < 1).e.. by increasing tax revenue and outstanding debt. There is a unique optimal tax rate. the outstanding government debt.6. if there is any debt-overhang (i. since at a zero tax rate there is a marginal benefit to having a transfer but no marginal cost. and in ND . Furthermore. The bottom panel of Figure 2 graphs the value of the government’s objective function. 13 . As the graph illustrates. Next. Moreover. is proven in Appendix A. the x-coordinate of which represents T . but decreasing in the amount of existing government debt ND . The optimal government action is to increase the transfer. T0 + kA ND . d2 L/dT 2 < 0. the financial sector liabilities. is also decreasing in ND . Moreover.. consider the three parts of Proposition 1. whose slope is given by (8). θ0 . This represents an upward shift in the marginal gain curve. underinvestment loss. Let ˆ T represent the associated tax revenues. which describes the solution to the government’s problem under assumptions 1 (certainty).The Optimal Tax Rate and Issuance of Debt: The following proposition. the gross transfer. until the marginal gain from the transfer no longer exceeds the associated marginal loss due to underinvestment. In addition. 3. For illustration. the objective is concave in T and the unique ˆ optimum occurs at T . i. whereby the marginal underinvestment loss induced by raising revenue becomes infinite as max the tax rate rises to θ0 . as shown in Appendix A.

but to hold total output constant while doing so. If the level of pre-existing government debt (ND ) is increased. Intuitively. and therefore the probability of solvency is lower. this pushes the government to increase the optimal tax rate. Hence. 9 10 NT = T − ND and 1def = 0 (No Default) Later. the financial sector liabilities.9 Finally. however. Recall that the transfer T0 equals P0 NT . if the financial sector’s output is a more important input into production.4 Default Under Certainty Now we allow the government to deviate from the no-default choice of setting NT = T − ND . The cost is that when NT > T − ND . As is clear from the new intersection point in the top panel.as (1) and (2) of Proposition 1 state.10 3. and as (1) of Proposition 1 states. more tax revenue. under the no-default and certainty assumptions. However. an increase in L1 . a larger factor share of the financial sector in aggregate production implies that the government will issue a greater amount of new debt and a larger transfer. In that case. we may think about comparing the ratio to total output of our variable of interest while varying the factor share. then there will be a greater marginal gain from an increase in the provision of financial services due to the transfer. we show that the possibility of default or the introduction of uncertainty can alter this result. ceteris paribus. an increase in pre-existing government debt corresponds to a decrease in newly issued sovereign debt and a smaller transfer T0 . splits the parameter space into two regions: 1. NT . triggering the dead-weight loss of D. the government’s decision on how many new bonds to issue. Equivalently. the effective transfer (T0 ) is smaller. Hence. Note. that the comparative static is not simply to vary ϑ. 14 . The reason for this decrease is that the underinvestment cost of raising additional tax revenues is increasing. the government will not be able to fully cover its obligations. The factor share is given by ϑ. as (2) of Proposition 1 shows. The benefit is that this can increase the transfer to the financial sector. and greater issuance of new sovereign debt to fund a larger transfer. tax revenue. P0 < 1 and the government will default. T /(NT + ND )) is the price of the government bond. Proposition 1 shows that. leads to a higher tax rate. The reason is that for any level of tax revenue. Increasing NT above this threshold has both an associated cost and benefit. and overall amount of sovereign debt. the rate of increase in total sovereign debt is less than the increase in ND . there is again an upward shift in the marginal gain curve. as shown by the dash-dot curve in the top panel of Figure 2. where P0 = max(1.

or shift the default-boundary itself. This implies that Wdef = Wno def N =0 − D. Formally. the amount of existing government debt ND . Wdef ). if the choice to default is made. and W = max(Wno def . ˆ ˆ Appendix A. We next consider the factors that push the sovereign towards default. Therefore. Ceteris paribus. let Wno def denote the maximum value of the government’s objective function conditional on no-default.4. and in the factor share of the financial sector 2. and (3) equilibrium provision of financial services is ˆ is bigger. s0 def > s0 no def . The following lemma characterizes the optimal government action and resulting equilibrium: Lemma 4. then it is optimal for the government to issue an infinite amount of new debt in order to fully dilute existing debt (P0 becomes 0) and hence capture all tax revenues towards the transfer.7. T0 higher. Wdef denote the maximum value conditional on default. Conditional on default. the benefit to defaulting is: 1. increasing in the financial sector liabilities L1 (severity of debt-overhang).2. NT > T − ND and 1def = 1 (Default) As shown in Appendix A. The following proposition characterizes how a number of factors move the ‘location’ of the sovereign relative to the default-boundary. decreasing in the dead-weight default cost D.7 proves the lemma. Moreover. Proposition 2.1 Default Boundary Figure 3 displays the optimal default boundary in L1 × ND space along with the No-Default and Default regions. the gross transfer ˆ def > T0 no def + kA ND . if default is undertaken then (1) the D ˆdef ˆno ˆ optimal tax rate is lower. 3. to determine whether defaulting is optimal. The resulting situation is the same as if pre-existing debt ND had been set to zero. and in the fraction of existing government debt held by the financial sector kA 15 . θ0 < θ0 def (2) provided that kA ND < T def . it is optimal to set NT → ∞ (and hence P0 → 0). the government evaluates whether its objective function for given ND and no default exceeds by at least D (the deadweight default cost) its objective function with ND set to zero.

decreasing the distance to the default-boundary. since the marginal gain of further transfer is higher at every level of transfer. causes collateral damage to the financial sector balance sheet. 3. will have a financial sector with a worse solvency situation.Appendix A. an increase in L1 ) leads to a large expansion in new debt (NT ) by the sovereign. as it acts to mitigate the under-provision of financial services. Both the extra transfer and decreased underinvestment represent benefits to defaulting. As the marginal cost of raising the tax revenue (dL/dT ) to fund this debt expansion is increasing.g. Consider a worsening of the financial sector’s health.11 It is clear that an increase in the deadweight loss raises the threshold for default. A similar kind of result holds if the factor share of financial services is increased. From the vantage point of Figure 3. Going in the other direction. both an increase in D and k cause an outwards shift in the default boundary. increases the gain to the sovereign from defaulting. a severe deterioration in the financial sector’s probability of solvency (e. a financial sector crisis pushes the sovereign towards distress.. This is represented by a move towards the right in Figure 3. This increases the marginal gain from further government transfer. since the marginal loss from funding extra debt is increasing. Hence. by Proposition 1.4.2 Two-way Feedback Propositions 1 and 2 indicate that there is a two-way feedback between the solvency situation of the financial sector and of the sovereign. An increase in existing debt implies a bigger spread between the optimal transfer and optimal tax revenue with and without default. 16 . Hence. This is represented by a move upwards in Figure 3. as well as is its maximum debt capacity (lemma 3). and importantly. and.8 provides the proof.. this benefit is convex in ND . Finally. by Proposition 1. which is aimed at freeing up resources towards the transfer. e. the sovereign is pushed closer to the decision to default (Proposition 2). one with high existing debt (ND ). a distressed sovereign. This is because it is very costly for such a sovereign to fund increased debt to make the transfer to the financial sector. in turn. leading to a decreased provision of financial services. If the sovereign has a lot to lose from defaulting (think a sovereign with strong domestic credibility or international reputation) then the net benefit to default will be relatively lower.g. an increase in the fraction of existing sovereign debt held by the financial sector also raises the threshold for default since the act of defaulting. First. a more distressed sovereign will tend to correspond to a more 11 Moreover. again decreasing the distance to the default-boundary.

the expected tax revenue. Hence. and Pricing We now introduce uncertainty about future output (i.distressed financial sector (lower post-transfer psolv ). leading to a ‘spillover’ of the financial sector crisis onto the solvency of the sovereign. Instead of a binary default vs.5 Uncertainty. In this case. are the variables the government directly chooses. However.e. The first variable is T .. Default. it will turn out to be more enlightening to look at two other variables that map to them in a one-to-one fashion. the mapping from H to NT is invertible. these alternative control variables map uniquely to the original ones on the region of interest. if raising taxes further incurs a large under-investment loss. no-default decision. limiting the benefit from this option (Proposition 2). the government can choose to increase debt issuance while holding the tax rate constant. large holdings of sovereign debt (k) by the financial sector means that taking this avenue simultaneously causes collateral damage to the balance sheet of the financial sector. respectively. The government’s problem (3) then is equivalent to optimally choosing T and H. The trade-off is an increase in the government’s probability of default and expected dead-weight default loss. In this case. This dilutes the claim of existing bondholders to tax revenues. Strategically defaulting is an avenue for a distressed sovereign to free debt capacity for additional transfer. where RV ≥ 0 represents the ˜ ˜ ˜ shock to V (K1 ). We also assume that the distribution of RV max Formally. In this case. a distressed sovereign is further incapacitated in its ability to strengthen the solvency of its financial sector. 12 17 . the mapping from θ0 to T is invertible on [0. 3. By construction. thereby generating a larger transfer without inducing further underinvestment. Although θ0 and NT .12 Note that the no-default and total-default cases under certainty correspond to setting H = 1 and H → ∞. The second variable is: H= NT + ND T In words. It measures the sovereign’s ability to cover its total debt at face value. E[RV ] = 1. H is the ratio of outstanding debt to expected tax revenue. θ0 ] (as before. which again equals θ0 V (K1 ). the government now implicitly chooses a continuous probability of default when it sets the tax rate and new debt-issuance. the sovereign effectively ‘sacrifices’ its own creditworthiness to improve the solvency of the financial sector. we can limit our concern to this region) and given T . ˜ ˜ ˜ To represent uncertainty we write V (K1 ) = V (K1 )RV . growth) by allowing the variance ˜ of V (K1 ) to be nonzero.

and as proven in Appendix A. as H increases. this cost is a flat function of H until the upper end of the support of the distribution. increasing H while holding T constant increases the transfer. This captures a greater faction of tax revenues towards the transfer but raises the probability of default. respectively. The top panel of Figure 4 illustrates the marginal gain (solid line) and loss (dashed line) incurred by increasing H for a fixed level of T . ˜ θ0 V (K1 ) NT + ND = E0 min 1. As (9)–(11) show. The first-order condition for T involves the same transfer-underinvestment trade-off as under certainty. For ˜ RV uniformly distributed. Figure 4 indicates that (with T held constant) there are two potential candidates for the 18 . To generate the plots. Raising H beyond this point represents sure default (pdef = 1). and NT .is independent of the variables K1 . raising H increases the transfer by diluting existing bondholders–it raises outstanding debt but without increasing expected tax revenue. pdef . we need ˜ to assume a specific distribution for uncertainty. but also increases the probability of sovereign default and decreases the sovereign bond price. The dash-dot line represents the marginal gain curve at a higher level of L1 than for the solid line. falling to zero beyond this upper point. Varying H involves a new trade-off. as NT = (T − ND /H)H. and probability of default. Pricing. = 1 ˜ RV H (9) (10) ˜ = prob θ0 V (K1 ) < NT + ND ˜ prob RV < H Note that these quantities depend only on H and do not directly change with T . This is shown by the dashed green line in Figure 4. Default Probability and the Transfer: Using H we can easily express the sovereign’s bond price. Next. RV H (11) The Optimal Probability of Default: Appendix A. As the figure shows. P0 . For simplicity. the marginal gain curve is downwards sloping. The marginal cost of an increase in H is the rise in expected dead-weight default cost.10 derive the first-order conditions for T and H.9 and A.10. θ0 . we express the transfer in terms of T and H: T0 = NT P0 = (T − ND ˜ )E0 min H. we let RV have a uniform distribution.This is because. the marginal increase in the transfer due to further dilution decreases. as follows: P0 pdef = E0 min 1. adjusted to account for H. Intuitively.

it is optimal for the government to fully dilute existing bondholders to obtain the largest possible transfer. the objective function again rises in H. the government will optimally begin to ‘sacrifice’ its own creditworthiness to generate additional transfer. H) is an interior solution to the government’s problem on a region ˆ of the parameter space. Th reason for this is that. and decreasing in ˆ D. T is also increasing in L1 . a high L1 (i. the government’s optimal choices of H and T in equilibrium: ˆ ˆ Proposition 3. If (T . a financial sector crises) will be associated with both a high T and a high level of H. The second is to let H → ∞. it pushes down the curve because it lowers overall welfare. it corresponds to an optimal non-zero probability of default. Therefore. 3. In the bottom panel. once debt issuance is large enough that default is certain. This comes at the cost to the sovereign of a further increase in the probability of default. The bottom panel of Figure 4 plots the corresponding value of the government’s objective as a function of H. in ND and in ϑ.5. Furthermore. 19 . The plot shows that for the configuration displayed. As this optimal H is beyond the left end of the support of RV (which is the origin in the figure).1 Comparative Statics Under Uncertainty The government jointly chooses T and H in an optimal way by comparing the relative marginal cost and benefit of adjusting each quantity. it would become optimal for a sufficiently high level of L1 .optimal choice of H. The first is the value of H at which the gain and loss curves intersect. the dash-dot curves in Figure 4 correspond to an increase in L1 (more severe debt-overhang) relative to the solid lines.. As is apparent in the top panel. When the marginal benefit of additional transfer is large. up to the point where total default becomes optimal. representing a total default and full dilution of existing bondholders. As indicated. The ˆ ˆ following proposition characterizes how different factors impact H and T . but the marginal underinvestment loss due to taxation is already high. From the bottom panel it is apparent that while total default is still suboptimal. which is at the intersection of the gain and loss curves ˜ in the top panel.e. pushing up the marginal gain curve in the top panel. then H is increasing in L1 . this worsening of financial sector solvency increases the marginal gain from an increase in H. Finally. the optimal response of the sovereign is to increase the optimal H (by issuing more debt) in order to increase the transfer. a relatively small value of H achieves the optimum. Note ˜ that beyond the upper end of the support of RV .

For simplicity. freeing extra capacity for the sovereign to generate the transfer. We then plot in Figure 5 comparative statics of the equilibrium (optimal) values of T . Correspondingly. the underinvestment costs of increasing tax revenue becomes very high and again the sovereign begins to increase H to fund the transfer. the optimal government response can be a total default. P0 remains fully valued at 1. the dilution of existing bondholders is an effective channel for increasing the transfer. which begins to decrease once H begins to rise. at high enough ND . so the probability of sovereign default remains 0. the plot shows that in this range. P0 drops to 0. The plots also show that. the government chooses to increase H. The top panel of Figure 5 is for L1 . For low levels of L1 . The outcome of a total default is illustrated in the plots at the point of the dotted line. H is held constant at a low value. This occurs because higher H means a a greater proportion of expected tax revenue is captured towards the transfer 13 20 . Once ND is sufficiently high. At the same time. As indicated in the plots. Moreover.13 . total default becomes optimal. Note that the transfer increases more rapidly in L1 once H begins to increase. similar to the case of certainty. so there is no probability of default and P0 remains at 1. As in the certainty case. It is apparent that for low levels of debt the sovereign keeps H constant at the low end of its support. financial crisis). up to the point where the sovereign chooses total default. ˜ we again let RV have a uniform distribution. as shown in the plot. the probability of default rises and P0 begins to decrease. The reason for this is that for large ND .˜ To obtain more precise results we choose a specific distribution for RV . T0 . as indicated by the dotted lines. while the damage to the sovereign’s creditworthiness is apparent in the plot for P0 . This value ˜ corresponds to the lower end of the support of RV . The discontinuities that appears in the plots. H. As discussed above. Consequently. as the plots show. The bottom panel of Figure 5 shows the comparative statics for ND . For these values of ND . The increase in the transfer is apparent in the subplot for T0 . total default fully dilutes existing bondholders. in this range the combination of increased H and T imply that the transfer is actually increasing in ND . Note that in this range the transfer is decreasing in ND . represent the point at which total default becomes optimal. when the financial sector’s situation is severe enough (L1 is large)..g. and P0 as L1 and ND are varied. Interestingly. For sufficiently high L1 (e. The corresponding plot for H tells a different story. this leads to a jump up in T0 and a jump down in T . the government funds the transfer exclusively through increases in tax revenues. it ‘sacrifices’ its own creditworthiness in order to achieve a larger transfer. It shows that T increases monotonically in L1 .

That is.6. In the interest of simplicity. Second. and since debt-overhang alleviation is the central feature of bailouts in the model. we add to the model a simple notion of a government guarantee of financial sector debt. We model debtholders as potentially liquidating (or inducing a run on) the financial sector if they are required to incur losses in case of financial sector default. the government ‘guarantees’ their debt. it has been common for sovereigns to step in to prevent the liquidation of banks by guaranteeing their debt. the ‘guarantee’ is just equivalent to a claim that issues L1 − A1 − T0 new government bonds to debtholders in case of insolvency. To prevent debtholders from liquidating. ˜ In fact. This The fallout from the failure of Lehman brothers and the apparent desire to prevent a repeat of this experience has strongly reinforced this view 14 21 . This unique feature is important in helping us identify empirically a main implication of our model.3. We do this for two reasons. which is necessary pre-condition for the sovereign to act to alleviate debt-overhang and increase the provision of financial services. the ‘guarantee’ has the same credit risk as other claims on the sovereign. 3.6 Government ‘Guarantees’ Government guarantees of financial sector debt have been an explicit part of a number of countries’ financial sector bailouts. that there is direct feedback between sovereign and financial sector credit risk. Instead. In this section. we do not explore the feedback of the guarantees on the transfer and taxation decisions analyzed above.1 Avoiding Liquidation A precursor to the government’s actions to increase the provision of financial services is to prevent liquidation of the financial sector. This ‘guarantee’ is pari-passu with other claims on tax revenue. their benefit is targeted at debtholders and not equity holders. Hence. which strongly suggests that there is an implicit ‘safety net’. we simply set the stage for the implications of the guarantees for our empirical strategy. it pledges to bondholders ˜ L1 − A1 − T0 from tax revenues in case of insolvency. notably Ireland. by construction. Moreover. First. guarantees are a measure that serves to prevent liquidation of the financial sector by debtholders. 14 . Note that this claim accrues exclusively to debtholders and not to equityholders. guarantees are rather unique in that.

Importantly. The following proposition gives a formal statement of this. Let D be the value of debt. to maintain simplicity we will not incorporate the impact of the guarantee on the government’s optimal choice of tax revenue or debt issuance.15 3. In the presence of a guarantee. derived under a ˜ uniform distribution for A1 (same as used above to generate the figures). the return on equity is sufficient for knowing the return on debt. This corresponds to the canonical model of debt. E the value of equity. 22 . such as the asset paying A1 or the transfer. the return on debt is a bivariate function of both the return on equity and the return on the sovereign bond price.6. This implies that.2 Two-way Feedback As illustrated by Proposition 3 and the discussion of Figure 5. is concentrated primarily with debt. following Merton (1974). where changes in the total value of the firm are reflected in both debt and equity values. (1 − psolv )(1 − P0 ) E dE (1 − psolv )2 L1 P0 dP0 dD ≈ + D psolv D E 2 D P0 (12) The term involving the equity return captures the impact on the debt value of any changes in the value of the general pool of assets of the firm. in the presence of a guarantee. the change in equity value will not be sufficient for determining the change in debt value. Instead. including expected future profits of the firm. This is because the guarantee represents additional (state-contingent) debt issuance by the sovereign. a severe financial sector crisis ‘spills over’ onto the sovereign via an increase in H and deterioration of the sovereign’s 15 Note that there is theoretically also an indirect feedback of sovereign credit risk on financial sector credit risk. In the absence of a guarantee. T0 . there is the additional second component. This bivariate dependence is approximated by the following relation. while a change in the value of the guarantee changes the value of debt but not the value of equity. It is therefore not captured adequately by the return on equity and requires the second term. which picks up the change in the value of debt coming from changes in the value of the government guarantee. if there are guarantees. In contrast. which is derived in the Appendix. which raises expected default costs of the sovereign and potentially reduces the size of the transfer to the financial sector. ˜ Proposition 4. The change in value of the guarantee. which reflects variation in the credit risk of the sovereign. As mentioned above. a change in the value of general assets of the firm changes the value of equity and debt in a certain proportion.˜ differentiates it from general assets of the financial sector. and A1 be distributed uniformly. we will mainly use the above result to motivate some of our empirical work.

creditworthiness. and a significant loss in the market value of their equity. furthermore. as the bottom panel of Figure 5 shows. which culminated in Lehman Brother’s bankruptcy. The pre-bailout period.g. an additional important consequence of increased H that can be seen from the model with uncertainty. In addition. there is an upper value where total default becomes optimal. With the introduction of uncertainty about realized tax revenues. saw a severe deterioration in banks’ balance sheets. post-transfer. growth). Moreover. the increase in H varies continuously with the severity of the financial sector crisis (L1 ) and is reflected continuously in the price of the sovereign’s bonds (equivalently. around. We interpret this as setting the stage for the initial time 23 . ceteris paribus. as we have shown. Once H is increased. then any resulting decrease in the sovereign’s bond price due to increasing H causes collateral damage to the financial sector’s balance sheet. This decreases the net transfer to the financial sector. in CDS levels). not only does the probability of default increase. a weaker post-transfer financial sector (lower post-transfer psolv ). there will be increased co-movement between the likelihood of sovereign and financial sector solvency. This negative shock generated substantial debt overhang in the financial sector and significantly increased the likelihood of failure of. or runs on. There is. financial institutions. but so does the sensitivity of the bond’s final payoff to realized growth (and hence tax revenue) shocks. higher pre-existing sovereign debt is associated with lower post-transfer sovereign bond prices. the increase in H will make the fi˜ nancial sector more sensitive to shocks to RV (e. leaving it less well-off post transfer. Moreover. Since the financial sector’s solvency post-transfer is dependent on this payoff due to its holdings of transfer and pre-existing government bonds. Going in the other direction. and after. if the financial sector has substantial holdings of existing sovereign bonds (substantial kA ). 4 Empirics In this section we present empirical evidence in favor of the main arguments made by this paper: (1) the bailouts reduced financial sector credit risk but were a key factor in triggering the rise in sovereign credit risk of the developed countries and (2) there is a two-way feedback between the creditworthiness of the sovereign and the financial sector. This implies that. The setting for our empirical analysis is the financial crisis of 2007-10. higher pre-existing debt corresponds to a smaller transfer and therefore. We divide the crisis into three separate periods relative to the bailouts: pre.. a substantial rise in the credit risk of financial firms.

and Australia with more than $50 billion in assets as of the end of fiscal year 2006. the United States. A significant challenge in demonstrating direct sovereign-bank feedback is the concern that another (unobserved) factor directly affects both bank and sovereign credit risk. we search for investment grade credit ratings using S&P Ratings Express. We choose this sample because smaller banks and banks outside these countries usually do not have traded CDS. We find CDS prices for 99 banks and match CDS prices to bank characteristics from Bankscope. giving rise to co-movement between them even in the absence of any direct feedback. 4. This is followed by the two sections of detailed results and the evidence based on the European bank stress test data. per the model. We address these concerns by utilizing a particularly useful feature of government ‘guarantees’–that they are targeted specifically at bank debt holders–to control for changes in bank fundamentals. The second part of our analysis focuses on point (2) by testing for the sovereign-bank two-way feedback. We also gather and exploit data on the sovereign bond holdings of European banks that was released after the stress tests conducted in the first half of 2010. This relationship is predicted by the model and is supportive of argument that the bailouts induced increases in sovereign credit risk. The first part focuses on point (1). We find credit 24 . We presents evidence that the bailouts transmitted risk from the banks to the sovereigns and triggered a rise in sovereign credit risk across a broad cross-section of developed countries. We also analyze the ability of the pre-bailout credit risk of the financial sector and prebailout government debt-to-gdp ratio to predict post-bailout credit risk.period in the model. and the bank bailouts as the sovereign’s response. Next. We then confirm a prediction of the model by documenting the emergence post-bailout of a positive relationship between sovereign credit risk and government debt-to-gdp ratios. We then search for CDS prices in the database Datastream. Using this data we show that banks’ holdings of foreign sovereign bonds has information about how foreign sovereign credit risk affects a bank’s credit risk. We next describe the data construction and provide some summary statistics.1 Data and Summary Statistics We use Bankscope to identify all banks headquartered in Western Europe. We make extensive use of a broad panel of bank and sovereign CDS data to carry out tests that establish this channel and show that it is quantitatively important. We present our empirical results in two parts.

1 and 90. 4. 2008. As of July 2007. we match these data to sovereign CDS of bank headquarters and OECD Economic data on public debt.5%. the increase in bank credit risk was triggered by the Lehman bankruptcy. some of the financial sector risk was transferred to sovereigns. respectively. We note that bank equity values declined sharply during this period with a negative weekly return of 7. 2008 in Ireland.4%. This evidence suggests the emergence of significant sovereign credit risk after the bank bailouts.6 basis points.1% and the average Tier 1 ratio was 8. average bank and sovereign CDS were 194.3 billion and equity of $26.2 The Sovereign Risk Trigger The first bank bailout announcement in Western Europe was on September 30. Panel B of Table 1 present summary statistics of weekly changes in bank CDS and sovereign CDS for the main bailout periods. the average bank had assets of $589. We start it in March 2007. Before the bank bailouts. This low level of sovereign credit risk suggests that financial market participants did not anticipate large-scale bank bailouts prior to September 2008. This level of bank credit risk reflected primarily banks’ exposure to subprime mortgages and related assets. In the bailout period.8 basis points and the average sovereign CDS (if available as of July 2007) was 6. As a result of the bailouts. All results presented below are robust to including these observations. After Lehman. We also find significant variation in sovereign CDS with a standard deviation of weekly changes of 12. These CDS levels are suggestive of a significant transfer of financial sector credit risk on sovereign balance sheets.1 basis points. we see a significant rise in both bank and credit risk with average bank and sovereign CDS of 301. respectively.2 basis points. Finally. The average bank CDS was 21.8 basis points. As we discuss below.8 billion. The average sovereign CDS was 12. In the post-bailout period. which increased average sovereign credit risk. most governments announced bank bailouts aimed at reducing bank credit risk. prior to the start of the financial crisis.9%.3 and 33. Note that this period 25 . The average equity ratio was 5. Therefore we define the pre-bailout period as ending on September 29. Panel A of Table 1 presents summary statistics for all banks with CDS prices and investment grade credit ratings. We drop all observations with zero changes in bank CDS or sovereign CDS to avoid stale data. the average bank CDS was 98.ratings for 86 banks and match these data to CDS prices and bank characteristics.6 basis points.

the pre-bailout period captures both the prolonged increase in bank credit risk during 2007-2008 and the post-Lehman spike that occurs before the bank bailouts. they can explain the small rise in sovereign CDS that occurs late in the pre-bailout period. We compute the change in sovereign CDS and bank CDS during this period for all countries in our data set. Hence. the fact that the impact in this period is so small quantitatively suggest that the bank bailouts were a surprise to the majority of investors. so that it includes the immediate effect of Lehman’s bankruptcy on other banks prior to the Ireland announcement. 2008. the average bank CDS in Ireland increases by 300 basis points over this period. Figure 6 summarizes the results for the pre-bailout period. there is almost no change in sovereign CDS. However. 2010. The figure shows that there is a large increase in banks CDS prior to the bank bailouts. Most bank support programs consisted of asset purchase programs. the figure shows that the credit risk of the financial sector was greatly increased over the pre-bailout period but that there was little impact on sovereign credit risk. Several countries made more than one announcement during this period. However. 2008. the first column depicts the change in sovereign CDS and the second column depicts the change in bank CDS over the pre-bailout period. As a result. the bank bailout announcements were not truly independent since sovereigns partly reacted to other sovereigns’ announcements. For each country we compute the change in bank CDS as the unweighted average of all the banks with CDS prices. bank CDS significantly decreased over this one-month period. debt guarantees. Almost every Western European country announced a bank support program in October 2008. 2008. in part to offset outflows from their own financial sectors to newly secured financial sectors. the first country to make a formal announcement was Ireland on September 30. For each country. Figure 7 plots the average change in bank CDS and sovereign CDS over the bailouts period. As noted above. We therefore define the around-bailouts period as the period in which the bank bailout announcements occurred.includes the bankruptcy of Lehman Brothers on September 15. To the extent that investors held such expectations prior to September 30. As shown in the figure. and equity injections or some combination therefore. but also the period immediately afterwards. 26 . Many other countries soon followed Ireland’s example. Such an expectation would reduce the observed increase in bank CDS and shift forward in time the rise in sovereign CDS. We note that some investors may have expected bank bailouts even before the first official announcement on September 30. For example. We omit countries for which either sovereign CDS or banks CDS are not available. Overall.

it is clear that bank CDS and sovereign CDS move together after the bank bailouts. It shows that the sovereigns responded to the distress in the financial sector with the bailouts. However.For example. The results are robust to using other cut-off dates in 2010. At the same time. Indeed the figures clearly show that these sharp increases were aligned tightly with the bailout announcement period. Figure 8 plots the change in bank CDS and sovereign CDS over this post-bailout period. Table 2 presents the result of our analysis. There is a clear. For example. It shows that both sovereign CDS and bank CDS increased across all countries. the sovereign CDS of Ireland increased by about 50 basis points. We measure the government debt-to-gdp ratio as the government gross liabilities as a percentage of gdp. 2010. the reporting date for the European bank stress tests. the average bank CDS in Ireland decreased by about 200 basis points. immediate increase in the sovereigns’ credit risk. This suggests that they are tied together and feedback on each other as our model suggests. Moreover. especially ones that had a large increase over the pre-bailout period. most other countries had a significant decrease in bank CDS. For the postbailout date we choose March 31. 2008. and government debt-to-gdp ratios. in accordance with our model. there is a significant increase in sovereign CDS. The appearance of this striking pattern across a broad cross-section of countries is directly in line with the predictions of our model. We choose March 2010 because this is the date for which the European bank stress data results were released. Similarly. Moreover. We measure pre-bailout financial sector distress at the country level by averaging bank CDS on September 22. Column (3) shows the result from regressing post-bailout log sovereign CDS on post-bailout debt-to-gdp. this caused a contemporaneous. positive 27 . Our model predicts that the bailouts should lead to an increase in sovereign credit risk and that the post-bailout level of credit risk should depend on pre-bailout debt and the pre-bailout level of financial sector distress. Most other countries exhibit a similar pattern with decreasing bank CDS and increasing sovereign CDS. the model suggests that there should emerge after the bailouts a positive relationship between measures of government debt-to-gdp even if such relationship appears before. We define the post-bailout period as beginning after the end of the bank bailouts and ending in March 2010. achieving a substantial reduction in banks’ credit risk. financial sector distress. We choose this date midway between Lehman’s bankruptcy and the first bailout announcement. We test these predictions using data on sovereign CDS.

From the point of model. Note. Column (3) shows that the coefficient on financial sector distress in the pre-bailout period is also statistically insignificant. Columns (5) examines the ability of pre-bailout financial sector distress to predict the change in government debt-to-gdp from the pre-bailout to the post-bailout period. The coefficient on debt-to-gdp decreases slightly but remains marginally significant. whereas debt-to-gdp ratios are lagging. In contrast. that the emergence of this relationship coincides with an overall rise in sovereign debt ratios. We can address this to some extent by using leading debt-to-gdp. This is displayed in the upper panel of Figure 9. the model predicts that H determines the level of sovereign CDS. debtto-gdp ratios are an imperfect proxy for θ0 H because H takes into account any future issuance of debt to pay for current obligations related to the bailouts. The R-squared of the regression is close to 50%. Column (4) shows the result when the log of the pre-bailout financial distress variable is also added as a dependent variable. First. From these results we can see that there emerged a relationship between debt-to-gdp and sovereign credit risk that was not present beforehand. Consistent with the model. The coefficient is small and is statistically insignificant. we find that financial sector distress is positively related to the increase in debt-to-gdp.relationship and the coefficient is statistically significant. Both coefficients are statistically significant and together the two variables explain 84% of the variation in post-bailout debt-to-gdp. the debt-to-gdp ratio corresponds to θ0 H in the model rather than simply H. As Column (4) shows the coefficient on pre-bailout financial distress is large and highly statistically significant. 4. post-bailout debt-to-gdp is predicted by pre-bailout debt-to-gdp and pre-bailout financial sector distress.965%. the prediction of the model carries over to θ0 H since θ0 is increasing in financial sector distress.16 This relationship is shown in the bottom panel of Figure 9. Column (1) of Table 2 shows that in the pre-bailout period there is only a very weak relationship between debt-to-gdp and sovereign CDS. the sovereigns have increased debt-ratios into the region where dilution occurs and there is a negative relationship between debt-ratio and price (see Figure 5). the price of its debt We make note of two points. Nevertheless. The coefficient is positive and marginally statistically significant. However. 16 28 .3 The Sovereign-Bank Feedback In this section we analyze the two-way feedback between sovereign and bank sector credit risk. The coefficient shows that a 1% increase in pre-bailout financial sector distress increases postbailout sovereign CDS by 0. Second. Once the sovereign opens itself up to credit risk due to bailouts. Column (6) shows that consistent with the model’s predictions. moreover.

Our market-wide controls are a CDS-market index and a measure of aggregate volatility. Such a factor would explain their co-movement without there necessarily being an underlying direct channel between them. 17 29 . we interpret changes in sovereign credit risk as changes in expectations about macroeconomic fundamentals. the estimate of β is 0. Hence. At the weekly frequency. Our model indicates that subsequent changes in the sovereign’s credit risk should impact the financial sector’s credit risk through three channels: (i) ongoing bailout payments and subsidies. from the sovereign to the financial sector. We take a number of steps to address this concern. In the first step. the sovereign’s creditworthiness is important in determining the potential for and magnitude of future bailout steps. and productivity. This result shows that sovereign and bank CDS exhibit a strong comovement and is consistent with direct sovereign-bank feedback. This may occur when early efforts are deemed to be insufficient. More specifically. which is comprised of 125 of the most liquid While we model the bailout as happening in one stage. we add controls that capture market-wide changes that affect both bank and sovereign risk directly. Our CDS market index is the iTraxx Europe index. This leads a second direction of feedback. an obvious concern is that there is another (unobserved) factor that affects both bank and sovereign credit risk. This presents another. such as employment. (iii) explicit and implicit government guarantees. The feedback channels imply that we should find that changes in sovereign and bank credit risk are positively correlated.47 and is highly statistically signficant. growth. We start by estimating the following relationship in the post-bailout period: ∆ log(Bank CDSijt ) = α + β∆ log(Sovereign CDSjt ) + εijt where ∆ log(Bank CDSijt ) is the change in the log CDS of bank i headquartered in country j from time t to time t − 1 and ∆ log(Sovereign CDSjt ) is the change in the log Sovereign CDS of country j from time t to time t − 1.becomes sensitive to macroeconomic shocks. This means that a 10% increase in sovereign CDS is associated with a 4. changes in macroeconomic conditions may generate a correlation between sovereign and bank credit risk even in the absence of direct feedback.17 (ii) direct holdings of government debt. dynamic channel through which changes in sovereign creditworthiness impact both financial sector debt and equity values. As with the one-stage bailout in our model.7% increase in bank CDS. However. establishing that there is indeed direct feedback between sovereign and financial sector credit risk is a significant challenge. in practice bailouts may occur in mutliple steps. Therefore. These fundamentals also have a direct effect on the value of bank assets such as mortgages or bank loans.

This captures changes in aggregate volatility which is an important factor in the pricing of credit risk. Column (8) adds the weekly fixed effects. and CDS-market specific shocks. the fixed effect captures any variation that is common across all banks. as expected. They therefore conclude that this variation represents ‘local supply/demand shocks’ in the corporate bond market. Goldstein. implying that an independent increase in sovereign CDS of 10% translates into a 1. The coefficient on sovereign CDS decreases but remains highly statistically significant.63% increase in bank CDS. so we examine the postbailout results first. the VDAX. Note that the weekly fixed effects are collinear with the market-wide control variables. and post-bailout periods. In the third step. which is the DAX-counterpart to the VIX index for the S&P 500.18 For the volatility index we follow the empirical literature and use a VIX-like index. The magnitude is also economically important. The CDS market index captures market-wide variation in CDS rates caused by changes in fundamental credit-risk. Therefore. the coefficient on CDS market is large. liquidity. δt are the weekly fixed effects. The right column adds the bank-specific coefficients on the controls and bank fixed effects. The left column reports the coefficient estimates including the market-wide control variables. The decrease is not surprising. as time fixed effects represent a very rich set of controls. In the second step.CDS names referencing European investment grade credits. around. The middle column adds the weekly fixed effects. We implement this approach by estimating the following regression: ∆ log(Bank CDSijt ) = αi + δt + β∆ log(Sovereign CDSjt ) + γ∆Xijt + εijt where ∆Xijt are the changes in the control variables from time t to time t − 1. Altogether. and Martin (2001) find that a substantial part of the variation in corporate credit spread changes is driven by a single factor that is independent of changes in risk factors or measures of liquidity. and highly statistically significant. As would be expected. This accommodates potential non-linearities in relationships. For each week. The control coefficients are both statistically significant and the signs are as expected. and the αi are bank fixed-effects Table 3 shows the results for the pre. an increase in aggregate CDS levels or volatility is associated with a rise in banks’ CDS. For each period there are three columns of results. Column (7) shows that β is positive. we include bank-specific coefficients on all the control variables and bank fixed effects. we do not estimate coefficients on the market-wide Collin-Dufresne.6% of the variation in weekly bank CDS. Our main focus is on testing for the sovereign-bank feedback. the variables explain 31. 18 30 . we include weekly fixed effects.

Column (9) shows that the coefficient on sovereign CDS is essentially unchanged and remains highly statistically significant after adding bank-specific coefficients on the control variables .1 Controlling for Bank Fundamentals The results above establish that there is a strong sovereign-bank feedback. the CDS market control coefficient is significant and has a large magnitude. 4.controls. a negative macroeconomic shock will decrease the value of banks’ assets or future earnings power. Therefore. in the pre-bailout period there is no evidence for sovereign-bank feedback. it is evidence that in fact sovereigns that took on more credit risk.3. Hence. This will raise bank and sovereign CDS even in the absence of a direct feedback between them. We address this concern using a strategy that utilizes a particularly useful feature of government ‘guarantees’–they are targeted specifically at bank debt holders. there may remain a concern that our strategy to this point does not control for country-specific macroeconomic shocks that bank-level fundamentals. Columns (1)-(3) show the results for the pre-bailout period. and hence saw a greater increase in their CDS. while at the same time reducing national output. in the around-bailout period an independent increase in sovereign CDS is associated with a decrease in bank CDS. Given the flexibility of this specification we interpret the survival of the coefficient on sovereign CDS as robust evidence in favor of direct sovereign-bank feedback. Perhaps due to the short time-series. This implies that sovereign-specific shocks should have a disproportionate impact on debt holders because. However. For example. Comparing these results with those for the around-bailout period in columns (4)-(6) shows interesting differences. More precisely. They show a small coefficient on sovereign CDS that is indistinguishable from 0. When the panel is estimated at the daily frequency they do in fact come up significant (unreported). similar to the results for the other periods. decreased by a greater amount their banks’ CDS. the coefficient on sovereign CDS is negative. In other words. to establish whether there is a direct channel 31 . indicating that most of the variation was already captured by the market-wide controls. For the around-bailout period. In contrast. such shocks change the value of government ‘guarantees’ (implicit or explicit). the coefficients are only marginally significant. This shock may not be fully captured by our market-wide controls if it has a heterogenous impact across countries. This is very much consistent with the evidence presented above that the sovereigns took onto themselves credit risk from their financial sectors during this phase. in addition to changing the value of bank assets. There is an increase in the R-squared ofabout 7% over column (7).

Proposition 4 shows that bank equity returns are sufficient for determining changes in bank CDS in the absence of government ‘guarantees’. Our strategy for dealing with this concern is motivated by the model. To this end. Columns (7)-(9) show that in the post-bailout period we find that the coefficient on sovereign CDS survives and is highly statistically significant. making it difficult to document their existence. It holds in models of defaultable bond pricing that build on the canonical model of Merton (1974). where stock returns contain all information about changes in bank asset values and therefore can (locally) capture all variation in the price of debt. The structure is similar to Table 3. The results show that the coefficient on sovereign CDS decreases somewhat but remains highly significant.2 Bank-level Heterogeneity To analyze further sovereign-bank feedback we also examine whether heterogeneity in bank characteristics affects banks’ sensitivity to changes in sovereign CDS. we estimate the coefficient on an interaction term of changes in sovereign CDS and a bank’s Tier 1 capital ratio. Indeed. As shown in columns (7) and (8). In the presence of ‘guarantees’. As shown in columns (1) to (6). On the other hand. a-priori it seems that even if there are direct sovereign-tobank channels. Column (9) includes bank-specific coefficients on bank stock returns. a regression that controls for equity returns should still find a (negative) beta on changes in sovereign CDS. 19 32 . The Tier 1 capital ratio is commonly used in the banking industry as a proxy for a This result is in fact quite general. We therefore augment the regression with banks’ weekly equity returns.we can test if sovereign CDS is still a determinant of bank CDS after we control for the impact of shocks to bank fundamentals. For the bailout and pre-bailout periods. Proposition 4 shows that ‘guarantees’ present a source for such a potential finding because they discriminate precisely in favor of debtholders. although the bank stock return coefficient is highly statistically significant and possesses the expected negative sign.19 This implies that once we control for bank equity returns we should not find that changes in sovereign CDS have any further explanatory power for changes in bank CDS.3. we again find a negative coefficient on sovereign CDS in the bailout period and an essentially zero coefficient in the pre-bailout period. 4. The estimates are shown in Table 4. the results are quite similar to those in Table 3. its inclusion has little impact on the magnitude of the sovereign CDS coefficient. then this is strongly supportive of a sovereign-to-bank feedback channel. their impact may be subsumed into an equity control. if one finds that sovereign CDS does have further explanatory power beyond equity returns.

Equation (12) suggests that we should find that the magnitude of the coefficient on changes in sovereign CDS is decreasing in banks’ Tier 1 ratios. Intuitively. we estimate the following regression: Bank Stock Returnijt = αi + δt + β∆ log(Sovereign CDSjt ) + γ∆Xijt + εijt where Bank Stock Returnijt is the stock return of bank i headquartered in country j from time t to time t − 1. Table 6 presents the results. Table 5 reports the results. the point estimates are clearly higher than for the post-bailout period. Columns (7) to (9) show that in the post-bailout period an increase in sovereign CDS is associated with a decrease in bank stock returns. the negative estimates for the interaction term indicate that banks with higher Tier 1 ratios were less sensitive to variation in sovereign CDS rates. 4. across the board. Similar to the post-bailout period. Note also that the coefficients on the controls have the expected signs and are significant. To that end. This result is robust to the inclusion of the full set of controls. For the post-bailout period. However. Although the estimates are negative in columns (7) to (9).3 The Impact on Equity Value For the purposes of establishing the existence of a two-way feedback we have mainly focused on changes in bank CDS. bank stock returns should reflect changes in sovereign credit risk due to their impact on the value of continuing bailout payments and banks’ holdings of government bonds. It is also interesting to look at the impact of bailouts on bank stock returns. this implies that well-capitalized banks experienced a larger CDS decrease relative to poorly capitalized banks during the bailout period.bank’s probability of solvency.3. the impact of changes in the value of government guarantees stronger for less well-capitalized banks. which includes the full set of controls is statistically significant. the coefficient on sovereign CDS is not distinguishable from zero during the bailout period. The point estimate is positive in columns (5) and (6). This specification is motivated by the model. From the viewpoint of the model. The coefficient of interest is the interaction term. In the pre-bailout period. which is consistent with the idea that equity holders 33 . the interaction term is positive but small and statistically insignificant. Perhaps somewhat surprisingly. Still. We use the same control variables as in Table 4. The results in the bailout period are surprising. but is not statistically significant. only column (9). the coefficient on the interaction term is negative.

To implement this test. we search for CDS prices in the database Datastream. The benefit of this approach is that it circumvents the usual concerns about omitted country-specific macro shocks. Columns (1) to (3) show that as before.4% of bonds are issued by the country in which a bank is headquartered. These data were released as part of the European bank stress tests. the average bank had risk-weighted assets of 126 billion euros and a Tier 1 capital ratio of 10. Let SovBondik be the share of for- 34 . A total of 91 banks participated in the bank stress tests. This is supportive of the model’s assumption that banks are exposed to home-country sovereign risk through their holdings of government bonds. respectively. we match 51 banks to CDS prices. we use data on the sovereign bond holdings of European banks.benefitted from bank bailouts. These banks represent about 70 percent of bank assets in Europe. Our test focuses uses information on changes in the value of foreignsovereign holdings rather than own-country sovereign holdings. We collect the stress test data from the websites of national bank regulators in Europe. We first describe the data and provide summary statistics. Using bank names. For all banks. Banks have a strong home bias in their sovereign holdings: about 69. The average holdings of gross and net European sovereign bonds are 20. 4. the average bank holds about one sixth of risk-weighted assets in sovereign bonds.4 European Bank Stress Test Analysis As a final part of our analysis.6 billion euros and 19. The data consists of bank characteristics and holdings of European sovereign bonds. the coefficient on sovereign CDS in the pre-bailout period is essentially zero. we construct a bank-specific variable measuring variation in the value of banks’ foreign sovereign bond holdings. The data provide a view of a bank’s bond holdings of both its own government bonds and those of other European countries. We use these data to conduct an alternative test of the impact of sovereign credit risk on bank credit risk. Hence. For each bank we match sovereign holdings to sovereign CDS and compute a measure of exposure to sovereign credit risk.7 billion euros. Unmatched banks are mostly smaller banks from Spain and Eastern Europe that do not have publicly quoted CDS prices.2%. Table 7 presents summary statistics for all banks that participated in the European bank stress tests. One potential offsetting effect on this result is that in some countries the benefits of bailouts to equity holders were offset by charges (or expectations of charges) levied by the government. which were conducted in the first half of 2010. As of March 2010.

35 . This does not have any effect on the coefficient. Column (3) controls for week fixed effects. We calculate the foreign holdings variable. Column (6) estimates the same regression as in Column (5) but excludes the holdings of German bonds from the construction of the foreignholdings variable. To further check for robustness. column (5) adds back bank fixed effects. The coefficient of interest is γ. we are implicitly assuming that the marginal CDS investor either knows or at least has some idea of the bank holdings. the coefficient is identified only off the cross-sectional variation in the value of foreign sovereign holdings. This coefficient suggests that a one-standard deviation increase in the change in the foreign sovereign holdings variable leads to an increase of about half of a one-standard deviation in the change of the bank CDS. which captures the effect of changes in the value of foreign bond holdings on the bank’s CDS. The column shows that this has no effect on the coefficient of interest. (13) where δt are time fixed effects. Column (1) shows a positive and statistically significant association between changes in banks’ CDS and their foreign sovereign holdings. The coefficient of interest decreases from 0. ForeignBondCDSit as the following weighted average: ForeignBondCDSit = i=j SovBondik ∗ SovereignCDSkt .141 but remains statistically significant at the 1%-level. We then estimate the following: ∆ log(Bank CDSit ) = δt + γ∆ log(ForeignBondCDSit ) + εit . In this case. The coefficient decreases to 0. This result suggests that variation in foreign sovereign holdings contains economically important information about variation in bank credit risk.eign sovereign holdings of country k by bank i. where SovereignCDSkt is the sovereign CDS of country k on day t. Table 8 shows the results. We do this to address a potential concern about reverse causality due to the possibility that Germany may provide bailouts to other countries. or banks in other countries.325 to 0. Column (2) shows that the coefficient remains unchanged when bank fixed effects are included. This suggests that common shocks affect both the change in bank CDS and the change in the foreign holdings variable. Column (4) controls for day fixed effects. Note that for each bank the foreign holdings variable excludes home-sovereign bonds.261. We estimate the regressions using the period one month before and one month after the reporting date for the sovereign bond holdings. By estimating this regression.

a cost of bank bailouts to the government or regulatory budget that is increasing in the quantity of bailout funds. 2010.b) considers the optimal taxation to fund bailouts in a continuous-time dynamic setting. Diamond and Rajan. 1989b) initiated a body of work that focused on ex-post costs to sovereigns of defaulting on external debt. Acharya. Panageas (2010a.. Bhattacharya and Nyborg. risk-shifting or asset substitution (e. imposition of international trade sanctions and conditionality in support from multi-national agencies. in a reduced-form manner. 2004). 2010.g. Philippon and Schnabl. Diamond and Rajan (2005. 2007). 2005).g. While the question of how necessarily involves an optimization with some frictions.g. Aghion. 2009). Bulow and Rogoff (1989a. is weak. (iii) the recent empirical literature on effects of bank bailouts on sovereigns.g. A large body of existing literature in banking considers that bank bailouts are inherently a problem of time consistency and induce moral hazard at individual-bank level (Mailath and Mester. the usual friction assumed is the inability to resolve failed bank’s distress entirely due to agency problems.. does consider ex-post costs of bailouts. as a whole. Shin and Yorulmazer. however.g. Diamond and Rajan. e. (ii) the literature on costs of sovereign default. They provide taxation-related fiscal costs as a possible motivation. 1994) and at collective level through herding (Penati and Protopapadakis. 2008.. among others) focus on specific claims through which bank bailouts can be structured to limit these frictions. In the theoretical literature on sovereign defaults. 1988. due to reputational hit in future borrowing. Brown and Dinc (2009) show empirically that the governments are more likely to rescue a failing bank when the banking system. 2008) model. Broner and Ventura 36 . adverse selection (e. Gorton and Huang. 2006) study how bank bailouts can take away a part of the aggregate pool of liquidity from safe banks and endanger them too.5 Related literature Our paper is related to three different strands of literature: (i) the theoretical literature on bank bailouts. Some other papers (Philippon and Schnabl. or tradeoff between illiquidity and insolvency problems (e. 2009). also highlighting when banks might be too big to save. and. The theoretical literature on bank bailouts has mainly focused on how to structure bank bailouts efficiently. Notably. Acharya and Yorulmazer... Acharya and Yorulmazer (2007. This could be due to under-investment problem as in our setup (e. Bolton and Fries (1999) consider the cost that bank debt restructuring can in some cases delay the recognition of loan losses. A small part of this literature.

primarily of bank creditors. Checherita and Nickel (2009) focus on the effect of bank bailout announcements on sovereign credit risk measured using CDS spreads. consider a collateral damage to the financial institutions and markets when a sovereign defaults. Martin and Ventura (2007). These effects are potentially all consistent with our 37 . Another strand of recent empirical work relating financial sector and sovereign credit risk during the ongoing crisis shares some similarity to the very recent papers on this theme. Finally. bank creditors were backstopped reflecting a waiting game on part of bank regulators and governments. especially beyond a threshold of 90% debt to GDP ratio of the sovereign. and private debt shrinks significantly while sovereign debt rises. Veronesi and Zingales (2009) conduct an event study and specifically investigate the U. among others. Acharya and Rajan (2010) and Gennaioli. producing a downward revision in CDS spreads of the former. One strand of recent empirical work focuses on the distortionary design of bank bailout packages. Reinhart and Rogoff (2009a. They find that the government intervention increased the value of banks by over $100 billion.(2005). Demirguc-Kunt and Huizinga (2010) do an international study of equity prices and CDS spreads around bank bailouts and show that some large banks may be too big to save rather than too big to fail. In particular. Some of their evidence mirrors our descriptive evidence. Sgherri and Zoli (2009) and Attinasi. Our analysis corroborates and complements some of this work. Acharya and Sundaram (2009) document how the loan guarantee program of the Federal Deposit Insurance Corporation in the Fall of 2008 was charged in a manner that favored weaker banks at the expense of safer ones. an ex-post deadweight cost of sovereign default in external markets as well as an internal cost to the financial sector through bank holdings of government bonds. after a financial crisis). Dieckmann and Plank (2009) analyze sovereign CDS of developed economies around the crisis and document a significant rise in co-movement following the collapse of Lehman Brothers. Broner. b) and Reinhart and Reinhart (2010) document that economic activity remains in deep slump “after the fall” (that is. our empirical investigation of banking sector holdings of government debt and how this introduces a linkage between bank CDS and sovereign CDS is novel. Martin and Rossi (2010). Our model considers both of these effects. They employ this as a possible commitment device that gives the sovereign “willingness to pay” its creditors.S. but also estimate a tax payer cost between $25 to $47 billion. government intervention in October 2008 through TARP and calculate the benefits to banks and costs to taxpayers. Panetta et al. (2009) and King (2009) assess the Euro zone bailouts and reach the conclusion that while bank equity was wiped out in most cases.

Using credit ratings data and data on sovereign bond holdings from the European bank stress test in May 2010. In the shortrun. Hyun Song Shin and Tanju Yorulmazer. “Crisis Resolution and Bank Liquidity”. even inflation) risk imposed on corporate and household sectors of the economy. Chapter 15 in Acharya. Viral V. which dilutes the value of existing government bonds and creates a two-way feedback mechanism because financial firm hold government bonds for liquidity purposes. New York University Stern School of Business Acharya. Taking cognizance of this ultimate cost of bailouts has important consequences for the future resolution of financial crises. References Acharya. bailouts are funded through the issuance of government bonds. “The Financial Sector ‘Bailout’: Sowing the Seeds of the Next Crisis”. and Matthew 38 . 6 Conclusion This paper examines the intimate and intricate link between bank bailouts and sovereign credit risk. Overall. This cost is a reflection of the future taxation (or more generally. Work in progess.model of how financial sector bailouts affect sovereign credit risk and economic growth. 2009. Review of Financial Studies. the design of fiscal policy. Such an ex-post cost of bailouts has received little theoretical attention and has also not been analyzed much empirically. Viral. In particular. Viral V. and Raghuram G. we document that developed country governments transferred credit risk from the financial sector to taxpayers during the height of the crisis in October 2008. we consider the emergence of meaningful sovereign credit risk as an important potential cost of bank bailouts. We also provide supporting evidence for our model using data from the financial crisis of 2007-10. and the nexus between the two. we find that sovereign credit risk in turn affected bank credit risk. 2010. forthcoming. Viral V. We develop a model in which the government faces an important trade-off: bank bailouts ameliorate the under-investment problem in the financial sector but reduce investment incentives in the non-financial sector due to costly future taxation. 2008. and Rangarajan Sundaram. Acharya. “Financial Sector and Sovereign Credit Risk”. Rajan.

“Cash-in-the-Market Pricing and Optimal Resolution of Bank Failures. Martin. European Central Bank Working Paper No. “Too Many to fail? Evidence of regulatory forbearance When the banking sector is weak”. “What Explains the Surge in Euro Area Sovereign Spreads during the Financial Crisis of 2007-09?”. forthcoming. 2005. J. Brown. 2009. “Globalization and Risk Sharing”. Universitat Pompeu Fabra. 2010.. Bulow. Viral V. American Economic Review. Acharya. Journal of Institutional and Theoretical Economics. 1131. Aghion. Review of financial Studies. 1–31.Richardson (editors). and Dinc. New York University Stern School of Business. 2007 “Too Many to Fail . “Restoring Financial Stability: How to Repair a Failed System”. Cristina Checherita and Christiane Nickel. Demirguc-Kunt. 39 . John Wiley & Sons. 2009. J. and Tanju Yorulmazer. and Tanju Yorulmazer. 10–24. 1989a. 155-178. Journal of Political Economy. Rogoff. 2705–2742. Philipp. 43-50. and K. and K. Working Paper. “A constant recontracting model of sovereign debt”. Broner. “Optimal Design of Bank Bailouts: The Case of Transition Economies”.An Analysis of Timeinconsistency in Bank Closure Policies. Universitat Pompeu Fabra. Patrick Bolton and Steven Fries. Broner. 51–70. Nyborg. Bulow. World Bank Policy Research Working Paper 5360.. C. Martin and Jaume Ventura. 97. Sudipto and Kjell G. “Bank Bailout Menus”. 2008.” Review of Financial Studies. Rogoff. 16(1). “Enforcement Problems and Secondary Markets”. Working Paper. 1999. Maria-Grazia. Viral V. Bhattacharya. Asli and Harry Huizinga. 2009. 1989b. Attinasi.. “Are Banks Too Big to Fail or Too Big to Save? International Evidence from Equity Prices and CDS Spreads”. S. 2007. Alberto Martin and Jaume Ventura. 79.” Journal of Financial Intermediation. Swiss Finance Institute Research Paper Series No. 21. 2010. “Sovereign debt: Is to forgive to forget?”. Acharya. 155.

Rajan. “Liquidity Shortages and Banking Crises”. 2009. George and Loretta Mester. and Andrea Zaghini. Marco Taboga. 101?-116. American Economic Review. “A Positive Analysis of Bank Closure. 3(3). 2005. 57. 1977. and Bank Bailouts. Myers. Domestic Banks and Financial Institutions. 2010. 296?-311. Stavros. Illiquidity Seeking and Credit Freeze”. 94(3): 455-483. 2009. “An Assessment of Financial Sector Rescue Programmes”. Dieckmann. 2006. Panetta. 2004. “Bailouts. Universitat Pompeu Fabra. Journal of Financial Economics.Diamond. Diamond. 38–41. Nicola. Quarterly Journal of Economics.” Journal of Financial Economics. Rajan. 5(2). 2009. Giuseppe Grande. 1994. Faib. Douglas and Raghuram G. Corrinne Ho. Efficiency. Gorton. Journal of Finance.and financial crises”. “The Determinants of Corporate Borrowing. Panageas.. 272–299. 2009. American Economic Review Papers and Proceedings.the incentive to manage risk. Thomas Faeh.” Journal of Financial Intermediation. “Time to Buy or Just Buying Time? The Market Reaction to Bank Rescue Packages”. Journal of Monetary Economics. 40 . Gennaioli. “Fear of Fire Sales. Mailath. Aviram Levy. Working Paper. Liquidity. Diamond. Panageas. “Optimal taxation in the presence of bailouts”. BIS Papers No 48 (July). BIS Papers No 288 (September). 60(2). Stavros. Michael R. King. 2010a. Wharton School of Business. “Sovereign Default. Gary and Lixin Huang. 2010b. 615–647. Rajan. University of Pennsylvania. 92(2). 95. “Bank Bailouts and Aggregate Liquidity”. Douglas and Raghuram G. “Default Risk of Advanced Economies: An Empirical Analysis of Credit Default Swaps during the Financial Crisis”. 147–175. Michael King. Alberto Martin and Stefano Rossi. Douglas and Raghuram G.” Working Paper. Stephan and Thomas Plank. Federico M Signoretti. forthcoming. Stewart C.

Princeton University Press. Silvia and Edda Zoli. No. Thomas and Philipp Schnabl. Veronesi. 2009b. with an application to international overexposure”. 41 . 2010. August 27. 2009. and Zingales. and Kenneth S. International Monetary Fund. Philippon. 2010. “Paulson’s Gift”. and Protopapadakis. Journal of Monetary Economics. Working Paper.. Sgherri. A. and Kenneth S. Working paper. 1988. 2009a. Wyoming. 107–126. Reinhart. Journal of Financial Economics. “The effect of implicit deposit insurance on banks? portfolio choices. Rogoff. forthcoming. 2009. Carmen M. A. Reinhart. P. Rogoff. “Efficient Recapitalization”.. New York University Stern School of Business. Carmen M. L. and Vincent Reinhart. forthcoming. 09–222. 21(1). “After the Fall”.Penati. “Euro Area Sovereign Risk During the Crisis”. American Economic Review Papers and Proceedings. presentation at the Federal Reserve Bank of Kansas City Economic Symposium at Jackson Hole. This Time Is Different: Eight Centuries of Financial Folly. Reinhart. “Growth in a Time of Debt”. 2009. Carmen M.

Anglo Irish Bank. . Bank of Ireland. The bank CDS is computed as the unweighted average of bank CDS for banks headquartered in Ireland (Allied Irish Bank. The data are from Datastream.Figure 1: Sovereign CDS and bank CDS of Ireland 01jan2007 0 200 400 600 800 01jan2008 (mean) cds 01jan2009 date 01jan2010 (mean) countrycds 01jan2011   Figure 1 plots the sovereign CDS and bank CDS for Ireland in the period from 3/1/2007 to 8/31/2010. and Irish Life and Permanent).

1]. α = 1.3 Value of Government’s Objective Function Gov. gov.15 τ 0.1 0. dL dG/dT0 and dL/dτ 0.5. γ = 0. dashed line) as functions of T for the certainty model of Section 3. The dash-dot line corresponds to a higher level of existing government debt. D = 0. β = 0.5.2. m = 1.25 (solid line). Objective Value val. The plots correspond to a parameterization of the model where A1 ∼ U [0.1 0.3 The top panel of Figure 2 plots the marginal gain (dG/dT ) of raising tax revenues (solid line and dash-dot line) and the marginal loss (dL/dT .3.15 τ 0.3. ϑ = 0.05 0. than the solid line. ND . L1 = 0.2 0.3.25 0.Figure 2: Marginal Gain and Loss of Raising T (Certainty Case) dG vs.05 0. with the the solid and dash-dot line corresponding to their counterparts in the top ˜ panel. and ND = 0. The bottom panel of the Figure shows the resulting value of the government’s objective function (equation (3)). k = 0.25 0.2 0. obj 0. .02.

2 Default ND 0.3 0.7 0.8 0.05 0 0.Figure 3: The Default Boundary (Certainty Case) Default and No−Default Regions 0.6 L1 0.1 No Default 0.3 0.4 0.9 Figure 3 shows the Default and No-Default Regions in the space of L1 × ND (financial sector leverage/debt overhang × pre-existing sovereign debt) for the certainty model parameterized as in Figure 2.15 0.2 0.25 0. . The black curve separating these two regions gives the Default Boundary.5 0.

25.4].5 (solid line).06.4 1. objective 0.8 1 1.6 Value of Government’s Objective Function value of gov.5. γ = 0. Uncertainty over growth/tax revenues. D = 0. dL dG/dH and dL/dH 0.2 2. m = 1.Figure 4: Marginal Gain and Loss of Increasing H dG vs.4 1.4 2. dG/dH (solid line and dash-dot line) and the resulting marginal increase in expected dead-weight default cost D dpdef (dashed line).3.2. The plots correspond to a ˜ ˜ parameterization of the model where RV ∼ U [0.6 The top panel of Figure 4 plots the marginal gain of increasing H while holding constant T . ϑ = 0. and ND = 0. .6 H 1.3. k = 0.4 2.6. with the the solid and dash-dot line corresponding to their counterparts in the top panel. The dash-dot line corresponds to a higher level of L1 than for the solid line. A1 ∼ U [0. 1].8 1 1. α = 1. dH ˜ RV . The bottom panel of the Figure shows the resulting value of the government’s objective function.8 2 2. is assumed to have a uniform distribution.2 1.2 1. L1 = 0. β = 0.2 2.6 H 1. 1.8 2 2.

2 ND 0.1 0.2 ND 0.1 0.2 ND 0.Figure 5: Comparative Statics for L1 0.3 0.4 0.6 L1 0.8 P0 T0 0.8 H τ 0.8 0.1 0.4 0.3 0.4 P0 T0 0. The dotted line in the plots represents the point at which total default (H → ∞) is optimal.4 0.4 Figure 5 plots the equilibrium values of T (expected tax revenue).4 H τ 0.6 L1 0. . H.6 L1 0.4 0. and P0 (price of sovereign bond) as L1 (top panel) and ND (bottom panel) are varied. T0 (the transfer). The parameters of the model correspond to those in Figure 4.3 0.4 0.8 Comparative Statics for ND 0. resulting in a discontinuity in the plot.6 L1 0.3 0.1 0.2 ND 0.

Bank of Ireland. Anglo Irish Bank. The bank CDS is computed as the unweighted average of bank CDS for banks headquartered in Ireland (Allied Irish Bank.Figure 6: Change in Sovereign and Bank CDS before Bank Bailouts 450 400 350 300 250 basis points 200 Sovereign CDS Bank CDS 150 100 50 0 ‐50 Figure 6 plots the sovereign CDS and bank CDS for Ireland in the period from 3/1/2007 to 8/31/2010. and Irish Life and Permanent). The data are from Datastream. .

.Figure 7: Change in Sovereign and Bank CDS during the Period of Bank Bailouts 100 50 0 AT AU BE CH DE DK ES FR GB GR IE IT NL NO PT SE ‐50 Sovereign CDS Bank CDS ‐100 basis points ‐150 ‐200 ‐250 Figure 7 plots the change in average bank CDS and sovereign CDS for Western European countries in the period from 9/26/2008 to 10/21/2008. The data are from Datastream (no data available for Switzerland Country CDS and Greek banks CDS during this period). The bank CDS is computed as the unweighted average of bank CDS for banks headquartered in that country.

. The bank CDS is computed as the unweighted average of bank CDS for banks headquartered in that country. The data are from Datastream.Figure 8: Change in Sovereign and Bank CDS after Bank Bailouts 1200 1000 800 600 Sovereign CDS Bank CDS 400 200 0 ‐200 Figure 8 plots the change in average bank CDS and sovereign CDS for Western European countries in the period from 10/22/2008 to 6/30/2010.

The top figure shows no correlation before the bailouts (as of 3/1/2007). . The bottom figure shows a strong correlation after the bank bailouts (as of 3/1/2010). The data are from Datastream and the OECD Economic database.Figure 9: Correlation between Sovereign CDS and Public Debt before and after bank bailouts Figure 9 shows the correlation between sovereign CDS and public liabilities (as a percentage of GDP) for Western European countries before and after the bank bailouts.

055 0.246 -0.8 1.496 6.0 12.267 65.079 -0.082 0.8 -0.034 0.9 9.4 41.341 0.6 3. The sample includes all European.101 606 0.9 20.5 10.174 0. 2010.896.045 -0.067 0.1 90.2 606 0.074 0.122 -0.3 26.0 1.027 0.000 -0.633 0.177 0.5 5th 50th Percentile Percentile 362.005 178.630 0.5 2.003 -0.017 -0.7 0.002 -0.8 1.100 0.249 0.102 3.041 0.058 455 -0.032 606 0.5 6.1 10.5 4.1 138.2 159.859 -0.496 194. 2007.5 81.042 0.519 0.495 6.018 -0.043 Post-Bailout (31/10/2008 .8 244.8 6.089 131.003 -0.138 0.324 0.495 6.2 6.103 0.Table 1: Summary Statistics The sample show cross-sectional and time-series summary statistics for the bank risk sample.3 615.122 0.333 0.136 0.7 1.5 58.130 Bank CDS (bp) Sovereign CDS (bp) CDS Volatility Bank Stock Return (%) ∆ Bank CDS (%) ∆ Sovereign CDS (%) ∆ CDS Market Index (%) Bank CDS (bp) Sovereign CDS (bp) CDS Volatility Bank Stock Return (%) ∆ Bank CDS (%) ∆ Sovereign CDS (%) ∆ CDS Market Index (%) Bank CDS (bp) Sovereign CDS (bp) CDS Volatility Bank Stock Return (%) ∆ Bank CDS (%) ∆ Sovereign CDS (%) ∆ CDS Market Index (%) .S.278 0.378 0.4 65.016 0.002 0.199 Bailout Period (9/1/2008-10/31/2008) # Mean Std.129 0.9 112.095 -0.Dev 606 301.9 100.6 Std.137 0.6 77.610 0.207 -0.208 0.281 3.633 12.9 28.633 0.4 10.5 6.302 3.4 11.001 -0.9 13.079 -0.043 0.217 727.8 0.496 6.002 0.2 3.633 98.001 0.496 4.5 24.140 0.518 0.038 2.8 5. U.9 29.033 605 0.000 -0.7 -0.008 0.001 0.3 37. 2007 to August 31st.5 -0.515 -0.814 6.6 3..Dev 594. Panel A shows summary statistics for the week of July 1st.168 0.0 52.028 0. The data are the weekly level.174 -0.003 0. Panel B shows summary statistics for the period from January 1st. Panel A: Cross-Section (7/1/2007) # Assets ($ billion) Equity ($ billion) Equity Ratio (%) Tier 1 Ratio (%) Bank CDS (bp) Sovereign CDS (bp) 81 81 81 66 75 56 Mean 589.5 95th Percentile 1.7 -0.1 8.3 2.009 65.015 0.124 95th Percentile 293.6 21.364 -0.0 0.1 Panel B: Time-Series Pre-Bailout Period (1/1/2007-8/31/2008) 5th 50th Percentile Percentile 3.2 0.7 606 33.5 21.004 -0.373 567.9 8.9 8. and Australian banks with available data on bank CDS and share prices.31/8/2010) 6.

137** (0. We report robust standard errors. and bank quality.Table 2: Emergence of Sovereign Credit Risk This table shows the relation between sovereign credit risk. Sweden. Columns (1) and (2) control for Public Debt measured as General Government Gross Financial Liabilities as percentage of GDP (collected from the OECD Economic Outlook).843 3.112** (0.171 0. The dependent variables are the public debt in June 2008 and average bank quality.118+ (10.726+ (11. respectively.401) 17 0. Denmark.154) 14 0. The dependent variable is the average bank quality. Column (2) controls for average bank quality measured as the average banks CDS before the bank bailouts (as of 9/22/2008).601 (1. The independent variable in column (6) is public debt in June 2010.357) -1. and Switzerland with publicly available data on sovereign and bank CDS. The independent variable in column (5) is the change in public debt from June 2008 to June 2010.923) 15 0.015* (0.005 (0. and + 10% significant Log (Sovereign CDS) Pre-Bailout Post-Bailout 1/1/2008 3/31/2010 % Public Debt (June 2008) Log (Average Bank CDS Sep 2008) Constant 2.208) 0.168) -86. public debt.965* (0.004) 0.013+ (0.006) 0.007) 0. The independent variable in Columns (3) and (4) is the sovereign CDS after the bank bailouts (as of 3/31/2010).548 (60.488 % Public Debt Around Bailout Post-Bailout ∆ 2010-2008 3/31/2010 1.005) 0.311 (0.006 (0.593 (2. The data are at the countrylevel.364 Observations R-squared . The independent variable in Columns (1) and (2) is the sovereign CDS before the bank bailouts (as of 1/1/2008). Great Britain. * 5% significant.555) -101.261 20. The sample includes all Eurozone countries and Australia.320) 15 0.456) 15 0.920 (49.107** (0.019) 15 0.144) 21. The dependent variables are the same as in Columns (1) and (2).134 0. ** 1% significant.

The data are at the weekly level.316 Y N 7.508 84 0.023* (0.384 Y Y 7. The control variables are the change in CDS market index. Columns (4) to (6) cover the bailout period (9/1/2008-10/31/2008).079** (0.015) Around Bailout (Sep-Oct 08) (4) (5) (6) 0. U.026 (0.860** (0.027) Week FE Interactions Observations Banks R-squared .S.134 Y N 577 71 0.027) 0.163** (0.689** (0.086 84 0. (3).403+ (0.232) -0.253 Y Y 3.508 84 0.504 -0. and Australian banks with available data on bank CDS and share prices.430 (0. (5).387 0. Column (3).308 Y Y 577 71 0.019 (0.082) 0. The main independent variable is the weekly change in the sovereign CDS.Table 3: Change in Bank and Sovereign Credit Risk This table shows the effect of sovereign credit risk on bank credit risk during the financial crisis. (6).441 0. and the CDS Market index.122* (0.015 (0.094) -0.081) N N 577 71 0.080** (0. volatility.019) 0. (9) include bank fixed effects and interactions of bank fixed effects with volatility.287) Post-Bailout (Nov 08-Sep 10) (7) (8) (9) 0. The standard errors are clustered at the bank-level.014) 0. and bank stock return.155) N N 3. Columns (1) to (3) cover the pre-bailout period (1/1/2007-31/8/2010). The sovereign CDS is assigned based on the country where the bank is headquartered.086 84 0.171 Y N 3.539** (0.030) 0. Columns (2).932** (0.508 84 0.086 84 0. * 5% significant.041) 0. and Columns (7) to (9) cover the post-bailout period (November 2008 to August 2010). ** 1% significant.050) N N 7. (8). The dependent variable is the weekly change in the natural logarithm of bank CDS. and + 10% significant ∆ Log(Bank CDS) Period ∆ Log(Sovereign CDS) ∆ Log(CDS Market Index) ∆ Volatility Index Pre-Bailout (Jan 07-Aug 08) (1) (2) (3) 0.010) 0.. bank stock return. The sample includes all European.214 (0. and (9) include week fixed effects. (6).

308 Y Y 2. The main independent variable is the weekly change in the sovereign CDS. volatility. and bank stock return.235 (0.105** (0.062 (0.153) -0.043) 0.020 (0.200) Post-Bailout (Nov 08-Sep 10) (7) (8) (9) 0.034) 0.359 Y N 5.106) 0.030 (0.848** (0.711** (0.096) -0. Columns (4) to (6) cover the bailout period (9/1/2008-10/31/2008). and + 10% significant ∆ Log(Bank CDS) Period ∆ Log(Sovereign CDS) Bank Stock Return ∆ Log(CDS Market Index) ∆ Volatility Index Pre-Bailout (Jan 07-Aug 08) (1) (2) (3) 0.Table 4: Change in Bank and Sovereign Credit Risk This table shows the effect of sovereign credit risk on bank credit risk during the financial crisis.745 63 0.491 0.032) 0.030) Week FE Interactions Observations Banks R-squared .278 60 0.174** (0. (5). (9) include bank fixed effects and interactions of bank fixed effects with volatility.100** (0. and Australian banks with available data on bank CDS and share prices.034) -0.118) 0.015) Around Bailout (Sep-Oct 08) (4) (5) (6) -0.255+ (0.154** (0.S.008 (0.929** (0.403 Y Y 437 53 0.278 60 0.043 (0. ** 1% significant. (6).. and (9) include week fixed effects.424 Y Y 5. The control variables are the change in CDS market index. Columns (2). U. and Columns (7) to (9) cover the post-bailout period (November 2008 to August 2010). (8).120) N N 2.481 0.278 60 0.014 (0.295* (0.150** (0.096) N N 437 53 0. (6).728 -0.236 (0. The data are at the weekly level. Column (3). bank stock return. * 5% significant. Columns (1) to (3) cover the pre-bailout period (1/1/2007-31/8/2010).224 Y N 2. The sample includes all European.009) -0.745 63 0.142 (0.132) 0. and the CDS Market index. The dependent variable is the weekly change in the natural logarithm of bank CDS. The sovereign CDS is assigned based on the country where the bank is headquartered.013) -0. The standard errors are clustered at the bank-level.123) -0.662** (0.208 Y N 437 53 0.036) 0. (3).051) N N 5.019* (0.147) -0.745 63 0.025) -0.

648 (0.568) -0. U. Columns (1) to (3) cover the pre-bailout period (1/1/2007-31/8/2010). All other controls are the same as in Table5.011 (12.400) -0.163 47 0..380 -1.S. and Columns (7) to (9) cover the post-bailout period (November 2008 to August 2010).335** (0.114 (8.096) Post-Bailout (Nov 08-Sep 10) (7) (8) (9) -1.256 48 0.347** (0.237) 61.126) -2.165** (0.224) -0.028) -0. All other variables are defined in Tables 3 to 5.360) 0. The Tier 1 capital ratio is the regulatory bank capital ratio.115) Y Y N 2.261 0. The sample includes all European.792) -0.019 (0.163 47 0.883) -0.510 Y Y Y 351 41 0.340 (0.095) Y N N 2. and + 10% significant ∆ Log(Bank CDS) Around Bailout (Sep-Oct 08) (4) (5) (6) -6.470) 0. The regression includes the Tier 1 capital ratio and an interaction between the Tier1 capital ratio and the change in sovereign CDS.589 (8.976) 0.054 (0.437 (0.317* (0.035) -0.037) Y N N 4.493) 0.072) -15.517 .256 48 0.492 (32.265* (2.139) Y N N 351 41 0.730+ (3.Table 5: Change in Bank and Sovereign Credit Risk (by Tier 1 Capital) This table shows the effect of sovereign credit risk on bank credit risk during the financial crisis.118) Other Controls Week FE Interactions Observations Bank R-squared Y Y Y 2.368** (0.134** (0. The dependent variable is the weekly change in the natural logarithm of bank CDS.937 (108.804 Y Y Y 4.066) -13. * 5% significant. ** 1% significant.712) -0.449 -2. and Australian banks with available data on bank CDS and share prices.118 (30. Columns (4) to (6) cover the bailout period (9/1/2008-10/31/2008).117 (0. The data are at the weekly level.079 (0. The standard errors are clustered at the bank-level.087 (78.423) 0.345 1.182) Period ∆ Log(Sovereign CDS)*Tier 1 ∆ Log(Sovereign CDS) Tier 1 Stock Return Pre-Bailout (Jan 07-Aug 08) (1) (2) (3) 0.140) Y Y N 351 41 0.014 (1.032 (1.528 -4.882) -0.575* (0.353 -6.163 47 0.337* (0.824* (1.031 (0.610 (1.256 48 0.765 (1.336) 67.127) -1.043) Y Y N 4.823) 0.

(6).054* (0. * 5% significant. Columns (4) to (6) cover the bailout period (9/1/2008-10/31/2008).011** (0.026) -0. Columns (2).895 65 0.078) -0.564 0.041 (0. and (9) include week fixed effects. Columns (1) to (3) cover the pre-bailout period (1/1/2007-31/8/2010).324 60 0.015) -0.004) -0.002 (0. (9) include bank fixed effects and interactions of bank fixed effects with volatility and the CDS Market index. and + 10% significant Bank Stock Return Around Bailout (Sep-Oct 08) (4) (5) (6) -0. and Columns (7) to (9) cover the post-bailout period (November 2008 to August 2010).035) -0.070 -0.057) N N N 5.324 60 0. (3). The control variables are the change in CDS market index and volatility.317** (0.106** (0.017) -0.243** (0.002 (0. (6).002) Post-Bailout (Nov 08-Sep 10) (7) (8) (9) -0.002) -0.177** (0.474** (0. ** 1% significant. The sovereign CDS is assigned based on the country where the bank is headquartered.082) Y N N 2.114 (0.026) Week FE Bank FE Interactions Observations Banks R-squared .240 Y Y Y 2.311 N N N 446 54 0. (5). (8).212 Y Y Y 446 54 0.S.285 Y N N 5.895 65 0. The sample includes all European..324 60 0. Column (3). and Australian banks with available data on bank CDS and share prices.040 (0.068** (0. The data are at the weekly level.114) Period ∆ Log(Sovereign CDS) ∆ Log(CDS Market Index) ∆ Volatility Index Pre-Bailout (Jan 07-Aug 08) (1) (2) (3) -0.075) 0.070) N N N 2.895 65 0.761** (0.368** (0.026) -0. The dependent variable is the weekly bank stock return. U. The standard errors are clustered at the bank-level.118 Y N N 446 54 0.Table 6: Bank Stock Return and Change in Sovereign Credit Risk This table shows the effect of sovereign credit risk on bank stock returns during the financial crisis. The main independent variable is the weekly change in the sovereign CDS.533 -0.488 Y Y Y 5.

March 31.930 6.719 11. EUR million) Sovereign Holdings (net.948 27.493 11.601 100.Table 7: Summary Statistics of European Bank Stress Test Sample The table shows summary statistics for all banks that participated in the EU Bank Stress Tests from July 2010. The data was collected from the website of the Committee of European Banking Regulators and nation websites of the respective bank regulators.269 7.2 105 105 182 117 18.023 69.307 14.422 13.2 493.960 5.6 0 92. EUR million) Home Share (%) Greek Sovereign Holdings Share Banking Book (%) 91 91 126.130 2.9 27. The sovereign holdings are computed as the total value of sovereign holdings relative to risk-weighted assets.0 2.8 3. We report both the gross and net exposure as reported to bank regulators.9 7.448 9.668 19.0 .329 14.2 179.956 30. The shares are computed based on gross exposure (net exposure was not reported).4 63.800 42. EUR million) Home Sovereign Holdings (gross.959 42. 2010 N (1) Bank Characteristics Risk-weighted Assets (EUR million) Tier 1 Capital Ratio (%) Sovereign Exposure Sovereign Holdings (gross.4 669 84.4 81.9 0 35.765 78. Sovereign Holdings Euro Bank Stress Tests Sample.800 100 5. EUR million) Home Sovereign Holdings (net.844 19.337 10.348 81. The share of trading book and banking book are the share of sovereign holdings held in the respective book.Dev (3) 50th Percentile (4) 5th Percentile (5) 95th Percentile (6) 91 91 91 91 91 91 91 20.4 Mean (2) Std.774 5.

173 0.317 51 0.317 51 0.046) Y N Y 2.173 0.329 0.317 51 0. Column (3) includes week fixed effects.027) N N N 2.261** (0.317 51 0.326** (0. Changes are computed as log changes. and +10% significant Change in Bank CDS Sample All (1) All (2) All (3) All (4) All (5) Excluding Germany (6) Change in Sovereign Exposure 0.049) N N Y 2. The data covers the period from 3/1/2010 to 4/30/2010.357 0.357 0.028) Y N N 2. (5) and (6) include bank fixed effects. Column (4) to (6) include day fixed effect.317 0. We construct the exposure variable as the weighted average of country CDS with sovereign holdings as weights.224 0.357 0.170 0. The standard errors are clustered at the bank-level (51 banks). ** 1% significant.027) N Y N 2.342 0. The sovereign bond holdings data were collected from the website of the Committee of European Banking Regulators and nation websites of the respective bank regulators.141** (0.317 51 0. * 5% significant.046) Y N Y 2.135** (0.188 0. Columns (2).325** (0.329 Bank FE Week FE Day FE Observations Banks R-squared Adjusted R-Squared . The exposure variable in Column (6) excludes German bonds.228 0.329 0.Table 8: Summary Statistics of European Bank Stress Test Sample The table shows regression of change in bank CDS on change in exposure to sovereign bank holdings.137** (0.

Sign up to vote on this title
UsefulNot useful