Journal of Banking & Finance 36 (2012) 3125–3132

Contents lists available at SciVerse ScienceDirect

Journal of Banking & Finance

journal homepage: www.elsevier.com/locate/jbf

Systemic risk, macroprudential policy frameworks, monitoring financial

systems and the evolution of capital adequacy
Bruce Arnold a, Claudio Borio b, Luci Ellis c, Fariborz Moshirian d,⇑
a
Australian Prudential Regulation Authority, Sydney, Australia
b
Bank for International Settlements, Basel, Switzerland
c
Reserve Bank of Australia, Sydney, Australia
d
Institute of Global Finance, University of New South Wales, Sydney, Australia

a r t i c l e i n f o a b s t r a c t

Article history: This paper analyses various issues that need to be tackled when promoting financial stability, reviewing
Available online 1 August 2012 the progress made in certain key areas and the remaining challenges. It explores the measurement of sys-
temic risk and of individual institutions’ contribution to it. It discusses aspects of macroprudential frame-
JEL classifications: works, including how the countercyclical capital buffer envisaged in Basel III takes into account the
G15 properties of the financial cycle and the strengths and weaknesses of macro-stress tests. It analyses some
G25 of the challenges of how best to monitor financial systems and the broader economy in order to detect
Keywords:
signs of vulnerability that might lead to future bouts of financial instability and of how to set prudential
Systemic risk policy accordingly. And it discusses the evolution of capital adequacy standards and the new emphasis on
Macroprudential polices liquidity standards in international regulation.
Monitoring financial systems Ó 2012 Elsevier B.V. All rights reserved.
Capital adequacy

1. Introduction This paper begins with a discussion of systemic risk. It focuses

The recent financial crisis has highlighted the importance of
promoting financial stability through better regulation and super-
vision of financial institutions. Key aspects of recent regulatory re-
forms include measuring and regulating systemic risk, and
designing macroprudential policies appropriately. This has been a
focus of institutions such as the European Systemic Risk Board
(in the EU) and the Financial Stability Oversight Council (in the
US, as well as at the global level.
This paper comes at a time of significant policy reform, designed
to respond to the increasingly interconnected nature of financial
institutions and the lessons of the Global Financial Crisis (GFC).
In recent years, the collapse of numerous financial institutions
has imposed significant negative externalities on governments
and the economy at large. This has increased the interest in mea-
suring the riskiness of financial institutions and appropriately allo-
cating risks (and costs) across them in order to account for the
negative externalities associated with financial instability. Against
this backdrop, there is a growing awareness of the need for a mac-
roprudential approach to regulation and of a better management of
the financial cycle. Systemic risk is of the essence here. The chal-
lenges in implementing the resulting reforms should not be
underestimated.
⇑ Corresponding author. Tel.: +61 2 93855859; fax: +61 2 93854763.
E-mail address: f.moshirian@unsw.edu.au (F. Moshirian).
on concerns associated with systemically important institutions
which, through their size and influence, might influence the stabil-
ity of the financial system. It discusses recent advances in the mea-
surement of systemic risk, the difficulties in implementing such
measures across multiple, different markets, and the need to
appropriately calibrate policies according to the risks posed by
banks, so-called shadow banks, and other institutions across many
jurisdictions.
The paper next situates financial institutions within the financial
cycle, with an emphasis on macroprudential policies. It discusses
the importance of the financial cycle within macroprudential frame-
works and the usefulness (or otherwise) of macroprudential stress
tests. Macroprudential factors can feed into the riskiness inherent
in systemically important institutions, though of course institu-
tion-specific factors such as risk appetite and business models mat-
ter as well.
The following section discusses barriers that might arise in
monitoring and regulating issues associated with systemic risk
and macroprudential policies. The two main challenges are those
associated with monitoring and analyzing risk, and those
associated with practical policy making. The paper discusses the
challenges in designing regulation appropriate to multiple jurisdic-
tions, especially if these regulations are drawn too tightly in a rule-
based fashion.
The paper goes onto analyze how understanding of adequate
capital requirements has changed over time, thereby highlighting

0378-4266/$ - see front matter Ó 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jbankfin.2012.07.023
3126 B. Arnold et al. / Journal of Banking & Finance 36 (2012) 3125–3132
one of the challenges involved in regulation: adapting to advances
in financial markets and research. The paper focuses on the evolu-
tion from the Basel I to the Basel III framework, the main lesson
being that regulation must continually evolve in light of changes
in our understanding of risk factors and in the risks that financial
institutions face.
The structure of this paper is as follows. Section 2 discusses
some of the recent developments on measuring systemic risk,
including risks in large banks that could contribute to financial
instability. Section 3 discusses the role of the financial cycle in
shaping policy reforms. Section 4 highlights the difficulties in-
volved in formulating policy and monitoring institutions, espe-
cially in an international context. Section 5 reinforces the points
raised in Section 4 by discussing how our understanding of risk
and monitoring has changed over time. Section 6 concludes.
2. Systemic risk
Achieving macroeconomic stability requires the identification
of systemic risk in the financial system and of the factors that
are driving it. In his speech at the 13th conference of the ECB-
CFS research network, Trichet (2010), President of the European
Central Bank, defined systemic risk as financial instability ‘‘so
widespread that it impairs the functioning of a financial system
to the point where economic growth and welfare suffer materi-
ally’’. While there is no universally accepted definition, let alone
an accepted measure to quantify this risk, there is a consensus that
the regulatory and supervisory framework should have effective
mechanisms to detect it and manage it.
2.1. Measuring systemic risk
There are at least three major issues in the field of measuring
systemic risk. The first issue is how to measure the drivers of sys-
temic risk. Some recent studies have focused on individual mea-
sures of systemic risk, which seek to predict how much the
stocks of financial institutions fall in a major market downturn
(the stress event). Acharya and Richardson (2009) and Acharya
et al. (2010), and Acharya et al. (2012) lay the theoretical founda-
tions of such an approach. In a downturn, financial institutions
may fall short of capital, which can lead to a failure (and possible
contagion) unless some other investor steps in. Governments usu-
ally want to minimize the resulting cost to the taxpayer, which
Acharya et al. (2012) show to be a function of size, leverage and
expected equity losses during a crisis. While the first two compo-
nents are easily available, econometric techniques may be neces-
sary to predict the expected equity loss in a financial crisis.
Acharya et al. (2012) propose a simple historical estimator. Brown-
lees and Engle (2010) suggest a bivariate model of returns that uses
asymmetric GARCH for volatility and an asymmetric DCC model for
correlation. A third method eschews modeling the entire return
process and only models the tail (as in De Jonghe (2010)).
These models have worked reasonably well in the United States,
but might require enhancement for use in other markets. Brown-
lees and Engle (2010) and Acharya et al. (2012) have shown that
their signal worked well in predicting which US major banks would
be severely affected in the 2007–2008 crisis. Their work is now
being extended to other banks in Europe and in Asia. One of the
challenges of researchers and institutions is to determine which,
if any, of these signals about systemic risk in the US work in an
international context over multiple crises while controlling for het-
erogeneity in economic development, financial sector structure
and regulations. For example, Griffin et al. (2010) show that stock
prices in emerging markets may reflect less idiosyncratic informa-
tion. Allen et al. (forthcoming) show that the structure of a
financial system (i.e. bank based versus market based) can influ-
ence the time required to recover from an economic downturn,
implying that financial-market-structure influences the nature of
banking risk within a country. Thus, more research needs to be
done to verify whether signals about systemic weakness in finan-
cial firms are reliable from these markets.
Considerable work has been under way seeking to measure
more precisely individual institutions’ contribution to overall sys-
temic risk. Tarashev et al. (2010) propose the ‘Contribution Ap-
proach’ (CA) method of attributing systemic risk to separate
institutions. The model is based on Shapley values, a game-theo-
retic concept initially applied to the measurement of the contribu-
tion of an individual to the output of a group. The idea is to
determine an institution’s incremental contribution to the overall
level of systemic risk, and is general enough to apply to a wide
variety of systemic risk measures (e.g., a system-wide Value at Risk
or Expected Shortfall). Tarashev et al. (2010) highlight that key
drivers of an institution’s contribution include its relative size, its
probability of default, and its exposure to a common risk factor.
They suggest that regulatory tools could be calibrated with respect
to such contributions.
Drehmann and Tarashev (2011) build on the framework in
Tarashev et al. (2010) to propose a ‘Generalized Contribution Ap-
proach’ (GCA) to apportioning systemic risk, taking explicitly into
account interbank networks, captured only implicitly through the
exposures to common factors in the previous piece of work1 Specif-
ically, Drehmann and Tarashev (2011) account for how a bank can
propagate shocks throughout a system by assuming that if losses
are large enough, counterparties will also fail along a kind of domino
chain. Relatedly, Billio et al. (2012) propose ways to measure this
interconnectedness, which could then feed into the analysis of
risk-allocation.
Bisias et al. (2012) survey most of the issues related to systemic
risk and analyze 31 quantitative measures of systemic risk. They
discuss these issues from both a supervisory and a research per-
spective. They also analyze the critical role of data in this process.
2.2. Banking crises and shadow banking
Another area of research that could be promoted is testing
across multiple crises in different economic situations, because it
is not always clear what triggers a financial crisis. In the theory
on banking crises, several channels have been proposed including
interconnectedness (Rochet and Tirole (1996)), liquidity spirals
(Brunnermeir and Pedersen (2009)) and macro-uncertainty (Chari
and Jagannathan (1988)). Borio and Drehmann (2009) show that
it is possible to create indicators, such as the increase in credit
and asset prices that can help to detect the build-up of risk of fu-
ture banking distress originating from a common source of pri-
vate-sector excesses. In the recent crisis, US banks may have
been affected because of large correlated holdings (Acharya and
Richardson (2009) and Acharya et al. (2010)) but banks in the
Euro-zone may have been impacted because of liquidity shortages
(Diamond and Rajan (2005)), a combination of both or other fac-
tors, as analyzed by Moshirian (2011). As research on systemic risk
continues, one has to ensure that eventually the measures of sys-
temic risk are robust to a variety of different channels that may
have caused distress in financial institutions.
Furthermore, modern financial institutions have a complex net-
work of contracts. During the 2007–2008 financial crisis, AIG, one
of the largest insurance companies in the US, had to be rescued
with $182.5 billion in loans. All the classic studies of financial
1
Some other researchers (e.g. Liu and Staum (2010)) have also used interbank
network for this purposes, albeit with complex linear programming techniques.
 B. Arnold et al. / Journal of Banking & Finance 36 (2012) 3125–3132 3127

crises, such as Reinhart and Rogoff (2009b), have only focused on surcharges adopted in Basel III seek to address these issues head-
banks. However, future research should also advance the literature on.
by considering systemic risk associated with major non-bank
financial players: these institutions can also contribute to systemic
3. The financial cycle: two lessons for macroprudential
risk but are often less regulated.
frameworks

Over the past two decades, the academic profession and policy-
2.3. Non-bank activities, bank concentrations, and systemic risk
makers have had a ‘‘crash course’’ in financial crises. But it was the
latest one, which broke out in 2007, that has done more than most
Another focus of financial market analysts and regulators in re-
to shape our understanding of financial instability. Not least, that
cent times has been the increasing reliance on non-interest income
crisis has dispelled the notion held in some quarters that financial
and non-deposit funding in banks. Demigurc-Kunt and Huizinga
crises with serious global financial and macroeconomic conse-
(2010) empirically corroborate this change in the balance sheet
quences could not occur in mature and highly sophisticated finan-
and revenue sources of banks by examining 1334 banks in 101
cial systems. By striking at the very nerve-centers of the Western
countries leading up to the 2008 crisis. Anecdotal evidence in the
economy, it was bound to trigger a major rethink of analytical
recent financial crisis would suggest that banks with large invest-
frameworks and policy.
ment banking units (Citibank) or trading books generally were sig-
The main analytical paradigm shift has been the rediscovery of
nificantly affected in the recent crisis. This is backed up by financial
the financial cycle as the factor that underlies severe financial cri-
theory, which has long shown the potential for moral hazard and
ses; that is, as the systematic factor hidden behind the myriad of
increased likelihood of failure as a bank expands into other lines
idiosyncratic elements that tend to cloud our understanding of
of business (Boyd et al. (1998).2 For example, Gennaioli et al.
the processes involved. The main policy paradigm shift has been
(2012) emphasize the risks that can arise in the presence of financial
the strengthening of the macroprudential, or systemic, orientation
innovation in non-commercial bank activities. This is not to say that
of regulatory and supervisory frameworks; that is, the recognition
all investment banking operations in financial institutions are dam-
that frameworks focused on seeking to ensure that individual insti-
aging, simply that an awareness of the potential risk associated with
tutions are sound on a stand-alone basis, as prevailed in many
them is necessary.
jurisdictions, were flawed. It could miss the wood for the trees.
Another important area of research has focused on whether
This section summarizes what we have learnt about the finan-
competition in the banking sector should be increased, whether
cial cycle in recent years and derives two implications for macro-
the size of large banks or the nature of their activities may have
prudential policy frameworks. In doing so, it draws largely on BIS
to be modified, and whether funding sources and activities should
research.
be regulated and corporate governance mechanisms altered. There
Three takeaway messages are worth highlighting. First, when
is a need for more research to analyze how these factors influence
the focus in on banking crises, the financial cycle is best character-
systemic risk. There are conflicting views in the academic literature
ized in terms of the behavior of private sector credit and property
about how competition in the banking industry affects fragility.
prices. Moreover, it is a medium-term phenomenon, with a dura-
Boyd and De Nicolo (2005) argue that a more concentrated banking
tion that extends well beyond traditional business cycles. Second,
sector allows banks to charge higher interest rates, which raises
the countercyclical capital buffer envisaged in Basel III takes fully
the risk profile of borrowers and consequently their vulnerability
into account the properties of the financial cycle. Finally, contrary
to default. By contrast, Hellman et al. (2000) posit that a more con-
to what is often believed, macro-stress tests – a popular tool in the
centrated banking sector offers more stability: bank owners reduce
emerging macroprudential frameworks – are ill-suited as early
risk because they want to earn monopoly profits and retain fran-
warning devices, i.e. as tools to identify vulnerabilities during
chise value. Engle et al. (2012) investigate the relationship be-
seemingly tranquil times. By contrast, if properly designed, they
tween business activities chosen by large banks and the
can be quite effective in crisis management and resolution.
concentration of banks within a country. They find that banks in
countries with low levels of concentration have higher levels of
non-interest income. The non-interest income generating activities 3.1. Characterizing the financial cycle3
of these banks improved risk adjusted profitability before the
2007–2008 financial crisis, but caused a sharp decrease in profit- In what follows, the term ‘‘financial cycle’’ denotes those self-
ability after the crisis. They find that non-interest income failed reinforcing fluctuations in perceptions and attitudes towards risk,
to provide diversification benefits for these banks during a finan- financing constraints and asset prices that tend to amplify business
cial crisis because it was more exposed to systemic risk in this fluctuations and that may lead to widespread financial distress and
episode. macroeconomic dislocations. These self-reinforcing fluctuations
A closely related issue is that of ‘Systemically Important Banks’ have also come to be known as the ‘‘procyclicality’’ of the financial
(SIBs) and ‘Global Systemically Important Banks’ (G-SIBs). G-SIBs system.
are banks that contribute a lot to systemic risk by virtue of their Empirical work suggests that the financial cycle, so defined, has
size, interconnectedness and/or other characteristics. A major con- five key properties.
cern is that the failure of a G-SIB can force a government to ‘bail First, the most parsimonious description of the financial cycle is
out’ the institution, thereby imposing significant negative external- in terms of the behavior of private-sector credit and property
ities on the public. One approach is to draw on the research, such prices. Equity prices can be a distraction: they exhibit shorter cy-
as that by Tarashev et al. (2010) and Drehmann and Tarashev cles and tend to be more closely related to short-term fluctuations
(2011), to calibrate regulatory tools with respect to an institution’s in GDP, which may leave the financial sector largely unscathed.
marginal contribution to systemic risk. However, such an approach This is most concretely illustrated by the experience of several
must necessarily work at a global level, requiring greater countries in the 1987-early 1990s and in 2001–2008. In each case,
coordination amongst the institutions’ home countries. The capital the crash in equity prices and the accompanying slowdown in eco-
nomic activity at the beginning of the period hardly made a dent in
2
By contrast, Barth et al. (2004) find that countries that restrict the activities of
3
banks are more prone to financial crisis. This section draws, in particular, on Drehmann et al. (2012).
3128 B. Arnold et al. / Journal of Banking & Finance 36 (2012) 3125–3132
the build-up of the financial cycle. Partly in response to the mone-
tary easing that followed, the credit-to-GDP ratio and property
prices continued to climb, only to collapse a few years further
down the road, generating a financial crisis and a much larger drop
in output. In fact, from a medium-term perspective, one could
characterize the economic slowdowns that went hand-in-hand
with the equity price crashes as ‘‘unfinished recessions’’.
Second, and generalizing the previous point, the financial cycle
has a much lower frequency than the traditional business cycle.
Since financial liberalization, the typical cycle length is of the order
of 16–20 years. This compares with typical business cycle frequen-
cies of up to 8 years. In other words, the financial cycle is a med-
ium-term phenomenon.
Third, peaks in the financial cycle tend to coincide with episodes
of financial distress. For example, in a sample of seven industrial
countries4 examined in Drehmann et al. (2012), all post-financial
liberalization financial cycle peaks are associated with either full-
blown crises or serious financial strains.
Fourth, and conversely, few crises do not occur at such peaks.
And these crises turn out to reflect exposures to financial cycles
abroad. Obvious recent examples include the recent crises in
Germany and Switzerland, in which banks incurred losses on their
exposures to financial cycles rooted mainly in the United States
and United Kingdom.
Fifth, it is possible to construct real-time indicators of banking
crises that provide fairly reliable signals with quite a good lead –
between 2 and 4 years, depending on the calibration (e.g., Borio
and Drehmann (2009)). Not surprisingly, such indicators are based
on (private-sector) credit-to-GDP and asset prices (especially prop-
erty prices) jointly exceeding certain thresholds, which fall outside
normal historical ranges. One can think of these indicators as prox-
ies for the build-up of financial imbalances, and as tools that help
policymakers distinguish sustainable from unsustainable booms.
The indicators appear to perform also quite well out of sample.
Finally, the amplitude and length of the financial cycle are
regime-dependent: they do not reflect by any means ‘‘natural con-
stants’’. Arguably, three key factors support financial cycles: finan-
cial liberalization, which weakens financing constraints; monetary
policy frameworks focused on near-term inflation control, which
can provide less resistance to the build-up of financial imbalances
as long as inflation remains low and stable; and positive supply-
side developments (e.g., the globalization of real economy), which
provide fuel for the financial boom while at the same time putting
downward pressure on inflation. It is not a coincidence, therefore,
that financial cycles have doubled in length since financial liberal-
ization in the early and mid-1980s and that they have been espe-
cially virulent since the early 1990s.
3.2. Policy lesson: the countercyclical capital buffer5
The countercyclical capital buffer envisaged in Basel III takes
fully on board the previous stylized facts (BCBS (2010a, 2010b),
Caruana (2010)), The scheme is designed to build up buffers during
the boom in the financial cycle and to draw them down as stress
materializes. Its main goal is to protect banks from the bust of
the cycle, thereby limiting the risk of banking crises. A collateral
benefit is that it may also restrain the boom in the first place. How-
ever, since capital is both cheap and plentiful during booms, this
collateral benefit may prove to be elusive, at least for the typical
size of the buffer contemplated in the scheme.
A constraint in the design of the scheme is that it should be sim-
ple and easily understandable. Ideally, therefore, one would like to
4
The countries are Australia, Japan, Germany, Norway, Sweden, the United
Kingdom and the United States.
5 6
This section draws, in particular, on Drehmann et al. (2011,2012).
have a single variable that could guide both the build-up and re-
lease phases of the buffer. But is this really possible? The problem
is that the best variable for the build-up phase would be the best
leading indicator of banking distress. Policymakers and market
participants should have sufficient time to build the system’s de-
fenses. By contrast, the best variable for the release phase would
be the best contemporaneous indicator of banking distress. The
buffer should be usable as soon as strains emerge. But it is hard
to imagine how the same variable could do both!
The design of the countercyclical buffer addresses this issue by
distinguishing between the indicators for the two phases. For the
build-up phase, the preferred reference guide described in the
BCBS documents is the credit-to-GDP gap, i.e. the deviation of
the ratio from a medium-term trend that is in fact consistent with
the average length of the financial cycle noted above. For the re-
lease phase, since it has not proved possible to identify any single
variable that would work consistently across countries, the trigger
is based on an evaluation that stress has materialized, which could
be rely on variables such as credit spreads, and bank losses.
Two additional issues concerning the design of the buffer are
especially important.
The first issue concerns the balance between rules and discre-
tion. Ideally, rules are important as pre-commitment devices and
to help the authorities resist the huge political economy pressures
to refrain from taking restrictive action during booms. But the
scheme is necessarily a simplification. Thus, for the build-up phase,
the credit-to-GDP gap acts purely as a reference guide, as a starting
point for a richer, far from mechanical assessment of the risks. The
second issue concerns how to address cross-border exposures.
There is an identification problem: what happens if losses are in-
curred on exposures to financial cycles abroad? And there is a con-
trol problem: how can national authorities offset the impact of
international lending? This form of lending may take place directly
from abroad or through domestic branches, which are often pri-
marily under the jurisdiction of the home authorities, i.e. those
where the headquarters of the transnational banks are located.
Home authorities may have little incentive to take action because
of a natural size asymmetry: typically, the exposures involved are
small relative to the size of the transnational banks’ portfolios, but
large relative to the host country’s financial system or GDP.
The scheme tackles these two problems head-on. To address the
identification problem, the buffer is calculated in relation to the
weighted average of the exposures of a bank to the various juris-
dictions. For example, based on the credit-to-GDP gap indicator,
German and Swiss banks with large exposures to the United States
would have seen a significant increase in their capital buffer on
those exposures ahead of the crisis. To address the control problem
the scheme envisages cooperative (reciprocity) arrangements. It is
the host authority that activates the buffer when signs of the build-
up in risks emerge in its jurisdiction; and the home authorities can
do more, but never less. All this implies a major shift in responsi-
bility from home to host authorities. And it could work as a model
of cross-border cooperation for a broader set of (macro-)prudential
tools.
3.3. Policy lesson: macro-stress tests6
The nature of the financial cycle has important implications for
what macro-stress tests can and cannot be expected to do. Given
current technology, macro-stress tests are ill suited as early warn-
ing devices. In fact, none flashed red ahead of the current crisis.
That said, if properly designed, they can be quite effective as crisis
This section is based on Borio et al. (2012).
 B. Arnold et al. / Journal of Banking & Finance 36 (2012) 3125–3132
management and resolution tools. The recent US experience prob-
ably comes closest to the use of the tool in such a context.
There are basically two reasons why macro-stress tests are inef-
fective as early warning devices: technical shortcomings and
context.
The technical shortcomings relate to the fact that the current
generation of models that underlie stress tests is unable to provide
a realistic picture of the dynamics of financial distress. In particu-
lar, the models and procedures cannot meaningfully capture the
relevant non-linearities and feed-backs, both within the financial
system and between the financial system and the macroeconomy.
No matter how hard you shake the box, little falls out! This means
that the action is shifted to the size of the ‘‘shocks’’, which end up
being unreasonably large. More generally, the very structure of
macro-stress tests in the antithesis of what financial instability is
all about. The essence of financial instability is that normal-sized
shocks cause the system to break down. An unstable financial sys-
tem is a fragile system; it is not one that would break down only if
hit by an extraordinarily large recession. The available empirical
evidence strongly supports this point. Financial crises occur before
output has contracted significantly and before asset prices (prop-
erty) or credit growth have fallen significantly.
The context is what might be called the ‘‘paradox of financial
instability’’: the system looks strongest precisely when it is most
fragile. Credit growth and asset prices are unusually strong, lever-
age measured at market prices artificially low, profits and asset
quality especially healthy, risk premia and volatilities unusually
low precisely when risk is highest. What looks like low risk is, in
fact, a sign of aggressive risk-taking. In other words, if interpreted
literally, these indicators show trouble far too late, only when the
risk that has built up during the boom materializes. All this com-
pounds the deficiencies in the models. Initial conditions look
unusually good. And, psychologically, the context fuels the ‘‘this-
time-is-different’’ syndrome (Reinhart and Rogoff (2009a)), some-
thing to which not even the authorities are immune.
The bottom line is that as early warning devices macro-stress
tests are more likely to be part of the problem than of the solution.
They risk lulling policymakers into a false sense of security. And
they did so in spades ahead of the recent crisis.
For much the same reasons, macro-stress tests can do consider-
ably better as crisis management and resolution tools. Now the
deck is stacked in their favor, or at least not obviously against.
The crisis has already erupted. As a result, vulnerabilities and
non-linearities have revealed themselves, balance sheets are weak,
banks are incurring losses or recording low profits, and hubris has
given way to prudence. In addition, the shocks envisaged are not
such unrealistic extrapolations of actual historical experience, as
would be required to generate a severe crisis in a model for a juris-
diction where no crisis has occurred for some time.
Macro-stress tests can then be helpful in identifying the need
for additional capital to prevent unnecessary credit crunches and
in sorting out strong from weak institutions to resolve them prop-
erly. This can help to address the excess capacity in the industry
generated during the preceding unsustainable boom.
That said, for macro-stress tests to work it is essential to design
them correctly. This means having the will to shake the system
hard. It means seeing them as complements, and never as substi-
tutes, for a tough inspection of the value of banks’ assets. And it
means having capital and liquidity backstops in place. These condi-
tions are necessary to get meaningful results and to secure the
credibility of the tests.
3.4. Summary
A better understanding of the financial cycle will help us devel-
op better analytical models and better policies. Belatedly, progress
3129
has been made. And, not uncharacteristically, policies have moved
ahead of academic research: waiting was not an option. It is high
time for research to catch up and move ahead of the curve (Borio
(2011)). The field is there for the taking.
4. Challenges of monitoring and policy-making
This section considers the challenges associated both with mon-
itoring financial systems and institutions and with constructing
policy. Section 4.1 addresses the issue of monitoring. Section 4.2
considers challenges involved in policy-making.
4.1. Challenges of monitoring and analysis
When implementing this array of policy proposals, financial sta-
bility policymakers have faced a number of practical challenges.
First among these is how best to monitor financial systems and
the broader economy to detect signs of vulnerabilities that might
lead to future bouts of financial instability. Policymakers had argu-
ably been held back in this regard because the canonical models
used in mainstream macroeconomics in recent years have not been
suitable guides for this work. The DSGE framework makes a
number of micro-foundational assumptions that are singularly
unhelpful in framing analysis in support of financial stability poli-
cymaking. In particular, the default assumption of a representative
agent makes it difficult to model the distributional issues that are
so important when detecting vulnerabilities, and relaxing this
assumption meaningfully can make the model intractable. The
usual rationality assumption, of a particular definition of rational
expectations where the agents have full information about the
workings of the model, likewise tends to rule out crucial behaviors
such as debt default and over-exuberance.
Instead, policymakers have looked to other parts of the litera-
ture for guidance on alternative models that could help frame their
monitoring and analysis. Three aspects of financial vulnerability
have motivated a great deal of new research, as well as much of
the post-crisis efforts to expand and enhance statistical collections
(IMF, 2009).
 Leverage: As became clear during the crisis, the most leveraged
banks and other players in the market are the most likely to
become distressed. Some of the banks at the center of the crisis
had increased their raw leverage significantly in the preceding
years, often in ways that the Basel II risk-weighted capital
adequacy rules would not detect. Work by Geanokoplos and
his co-authors, among others, has emphasized the role of lever-
age in creating and propagating distress amongst investors and
holders of financial products.
 Maturity transformation: Banks are in the business of transform-
ing maturity and taking on the attendant risks. In the years
leading up to the crisis, however, a range of ‘shadow banks’
and market participants took on such risks without having
either the capital or the access to central bank liquidity that
have long been known to be necessary to avoid runs. Banks
themselves also failed to allow for the possibility that funding
markets could become disorderly, and some began to rely on
repo markets and other very short-maturity funding. The role
of liquidity and short-term funding has been emphasized in
the work of Shin and his co-authors.
 Interconnectedness: One good example of work in this area is the
rapidly developing financial stability literature that borrows
from the theory of networks developed by researchers in other
fields. The paper by Fukuda in this volume takes another view of
interconnectedness, in this case between interbank money mar-
kets located in different countries.
3130 B. Arnold et al. / Journal of Banking & Finance 36 (2012) 3125–3132
Looking further back, four strands of older literature appear rel-
evant for practical implementation of monitoring and policymak-
ing to promote financial stability.7
The first is careful accounting of stock-flow imbalances in the
style of Wynne Godley. This has been the backbone of traditional
macro-financial analysis, focused on saving–investment behavior
and its balance sheet counterparts, both of sectors and whole econ-
omies. Analysis of the composition of household and business
balance sheets, funding requirements of the financial sector and
debt-servicing obligations all fall into this tradition, and are al-
ready frequently seen in central banks’ financial stability reviews
and similar publications.
The second older analytical tradition can be traced to the histor-
icism of Kindleberger, and related to that, Minsky’s insight that sta-
bility can breed future instability. The notion of a credit cycle,
somehow separate from the business cycle, that builds up as
risk-taking and asset prices rise, stems from Minsky’s work,
although his taxonomy of borrowing phases is usually only implicit
in more recent research. Policy-oriented analysis that belongs to
this tradition includes monitoring of credit flows by purpose, with
specific attention to debt-funded speculative asset purchases. Ana-
lysts and policymakers in this tradition would need information on
the fundamentals of different investment projects and asset mar-
kets, to determine when ‘hedge’ finance has turned into the spec-
ulative or Ponzi kind.
A third tradition focuses on governance, legal contract details
and the scope for fraud. Galbraith (2009) has linked this perspective
to the work of Galbraith senior; some others in this broad tradition
include Akerlof and Stiglitz, including their work on default and
asymmetric information. The paper by Aebi et al. (forthcoming)
with its focus on governance, falls partly into this strand of re-
search, as does much of the literature dissecting the specific failing
in the US mortagage market in the lead-up to the housing bust
there.
Aebi et al’s finding that corporate governance over risk manage-
ment was the most important element of governance determining
banks’ performance during the crisis points to the crucial role of
risk appetite and risk-taking behavior as an indicator of vulnerabil-
ities. Such a focus on behavior is another way of framing some of
the boom-bust dynamics that form the basis of the Minsky–Kind-
leberger tradition of literature; some more recent attempts to
quantify risk-taking as a key dynamic and source of vulnerability
in economic systems include Borio and Zhu (2008) and Altunbas
et al. (2009). The paper by Klomp and de Haan in this volume sug-
gests that prudential regulation and supervision does indeed affect
risk-taking and thus the risk profile of banks, most significantly
amongst the higher-risk banks.
Another signal of governance issues that might lead to vulnera-
bilities and future financial disruption is rent-seeking. Rent-seek-
ing is normally defined as using private information or other
advantages to obtain a private gain, without adding any economic
value. Such behavior thrives in complex, opaque markets or where
there are long chains of agents between the ultimate principal and
the underlying transaction; a perfect example of such an environ-
ment is the structured finance market during the years leading up
to the crisis. Rent-seeking is hard to quantify, but can often be
identified through careful analysis of contract terms, and supervi-
sory observation of market participants’ behavior and attitudes.
Two other, inter-related analytical perspectives in some senses
bring the focus of attention back to the micro-level, on the
behavior of individual agents, but without making the strong
7
The first three of these strands of literature were pointed out as being useful in
Galbraith (2009). Several other authors (e.g. White, 2008) have emphasized the
Minsky–Kindleberger tradition over alternatives.
assumptions of the mainstream DSGE literature. The first of these
emphasizes the role of emergent behavior from the interaction of
agents. Feedback effects pervade financial systems. These feedback
effects include the effects of decisions by lenders or borrowers on
their counterparties, by suppliers of credit on asset prices and thus
owners of those assets (even when they are not customers of that
lender), or by the creditors of a financial intermediary on that
intermediaries’ borrowing customers in turn. In particular, lenders’
decisions can affect the economic situation of their borrower cus-
tomers, which if it reaches a point of sufficient economic distress,
can feed back onto the viability of those lenders. Examining the
behavior of one institution in isolation, or of a representative agent,
is therefore often not helpful in understanding or predicting out-
comes at the level of the whole system, but neither is examining
the system without considering interaction effects.
Since mainstream DSGE-type models become analytically
intractable with multiple agent types, some researchers have
turned instead to agent-based modeling (ABM): see Tesfatsion
(2006) for a survey and Leijonhufvud (2006) for some reflections
on the usefulness of this methodology in macroeconomics and
financial stability analysis specifically. This strand of literature
and techniques emphasizes emergent behavior the possibility of
nonlinear phase transitions, for example between a ‘normal’ state
and a distressed or ‘crisis’ state. An example of this nonlinearity
might be seen in the extent of the collapse in world trade in late
2008, and slow subsequent recovery, compared with the faster
recoveries from the shocks to trade arising from natural disasters
in Japan and Thailand in 2011. Although it is not explicitly an
ABM, the paper by Affinito (forthcoming) certainly puts relation-
ships between different kinds of agents – in this case, banks and
their customers – front and center.
The second strand of micro-level focus emphasizes less the
interactions between the agents as the diversity amongst them.
Losses on a segment of a leverage investor’s assets can lead to dis-
tress and failure even while the ‘median’ asset performs well. Tra-
ditional prudential stress-testing of every institution in an industry
could be considered to fall into this stream of work, in a way that a
stress-test based on macro-level variables does not. By examining
results for individual banks or portfolios, supervisors and other
policymakers can detect (cross-sectional) tail risk: vulnerabilities
that are not apparent in macro-level data. Here some of the work
of the actuarial profession may be brought to bear, particularly
the parts of the risk-management literature focused on extreme va-
lue theory (e.g., MacNeil et al. (2005) and other papers by these
authors). While the mainstream finance literature has traditionally
focused on pricing and thus the whole distribution of possible out-
comes, the actuarial literature by definition has focused on ex-
treme outcomes and model risk.
4.2. Challenges of practical policymaking
As analytical models improve, policymakers will increasingly
face the question of whether and how they should respond if those
models imply that risks and vulnerabilities in the financial system
have risen. Possible responses range from issuing warnings – for
example in financial stability reviews or speeches – to more expli-
cit policy interventions such as controls on the property market.
Normal prudential supervision clearly can be included in the set
of possible interventions. Inappropriate mandates have con-
strained supervisors in some countries from taking systemic issues
into account – forcing them to take a purely ‘microprudential’ view
as noted above. However, supervisors in at least several countries
have shown a willingness to use the powers available to them in
the interests of the stability of the whole system.
While some supervisors have long been successfully manipulat-
ing microprudential tools for macroprudential reasons, designing
 B. Arnold et al. / Journal of Banking & Finance 36 (2012) 3125–3132
macroprudential policy rules seems more problematic. In large
part this is because of a lack of theoretical work that could be used
to guide good practice. Policymakers find themselves relying on
the experience of other countries, which might not be relevant to
their own systems. There is very little theoretical or other compar-
ative work on whether and where institutional differences can af-
fect the optimal policy framework or setting for financial stability.
In particular, there is nothing to suggest that ‘one size fits all’, and
that there is one set of best practice that all countries should
adopt.8
A further difficulty lies in the fact that the ultimate goals of the
policy are still the usual macroeconomic ones of output and wel-
fare. Financial stability matters because the lack of it imposes enor-
mous costs to society through output losses and impaired
economic functioning. Financial vulnerabilities, credit booms and
asset price imbalances are intermediate targets, which only matter
for policy because of what they imply for (future) output. In this
sense, macroprudential policy targeting rules more closely resem-
ble the monetary targeting regimes of the past than the more di-
rect mapping of policy instrument to ultimate goal in modern
inflation targeting. Attempts to specify macroprudential regimes
and rules too tightly are therefore likely to result in the same issues
of parameter instability that Goodhart’s (1984) Law implied for
monetary targeting. All these considerations suggest that it would
be imprudent to organize macroprudential policy around precise
quantitative rules or targets.
5. Evolution of capital adequacy
Financial-sector regulation has been focused on the idea of cap-
ital since the promulgation of the Basle Capital Accord (later reti-
tled ‘‘International convergence of capital measurement and capital
standards’’, and now referred to as Basel I). This focus has naturally
shaped the research agenda over the last two decades. In turn, this
research has given us insight into the workings of the financial cy-
cle; sophisticated approaches to modeling a highly interconnected
system of individual actors (an example of which is the work by
Affinito (forthcoming)); and, at least in hindsight, how much capi-
tal is actually required to cope with accumulated risk and unantic-
ipated interaction effects.
We now seem to know a lot about capital and the banking sys-
tem — so much so that we sometimes forget how little we knew
20 years ago. An introductory paragraph of Basel I noted the pur-
pose of the accord, ‘‘to establish minimum levels of capital for inter-
nationally active banks’’ (paragraph 7, emphasis in the original). So
we knew that more capital was good (The inference was clear,
though, that the then-current levels of capital might not have been
good enough). We formed the consensus that capital requirements
properly should be based on some measure of riskiness, rather
than ‘‘the simpler gearing ratio approach’’ (paragraph 28).
What we learned in the next decade was that the original scope
of the accord was too restrictive, and that the original methodology
for quantifying risk was deficient. Basel I was supplemented by the
1996 Market Risk Amendment and superseded by Basel II: A Revised
Framework, and so credit risk was joined by market risk, opera-
tional risk, and a host of so-called Pillar-2 risks. Credit risk weights
based on the obligor’s category type were replaced by weights
based on the obligor’s creditworthiness. More importantly, Basel II
opened the door to banks’ substituting their own models for
assessing risk, developed and validated through their own
8
See for example the discussion of the EU takeover directive in Humphery-Jenner
(2012b), the discussion of regulatory harmonization in Humphery-Jenner (2012a) and
Weatherill (2012), and the discussion of banking risk and regulation in Klomp and de
Haan (forthcoming).
3131
empirical research efforts, but subject to independent supervisory
evaluation.
From a research perspective, there is still work to be done in
refining methodologies and quantifying risk, and in assessing the
effectiveness of the various aspects of the capital-adequacy regime.
Research of this kind is represented in studies such as Duan and
Van Laereb (forthcoming) and Shi and Werkery (forthcoming).
But if these are the issues that now concern regulators when cali-
brating capital, just consider how far we have come from the start-
ing point represented by Basel I.
The other area in which our knowledge has grown relates to the
quality of capital. Due to ever greater creativity in the capital mar-
kets, the number of capital (or capital-like) instruments prolifer-
ated. As they did so, regulators felt compelled to make ever finer
distinctions. The original simple split between core and supple-
mental capital became more complex, with a third tier added
and ‘‘innovative residual Tier 1 capital’’ and other non-intuitive
categories somehow accommodated. Then the global financial cri-
sis taught us that some capital is better than other capital, and lo
Basel III: A global regulatory framework for more resilient banks and
banking systems. While Basel III is more complex than Basel I —
as it needs to be, given the evolution of the capital market — the
‘‘going-concern/gone-concern’’ conceit represents a return to first
principles.
In sum, we learned three lessons — more capital is good, the
amount of capital should be commensurate with the relevant risks,
and some capital is better than other capital — but it took us
20 years and a few false starts to do so.
In this light, consider the stablemate to the Basel III directive on
capital: International framework for liquidity risk measurement, stan-
dards and monitoring. In a phrase which echoes Basel I — even to
the point of italicizing the same word — ‘‘the standards establish
minimum levels of liquidity for internationally active banks’’ (par-
agraph 6, emphasis in the original). What we know now about
liquidity is what we knew two decades ago about capital: More
is good.9
There are other parallels between Basel I-on-capital and Base-
l III-on-liquidity. Like Basel I, the new accord on liquidity asserts
that liquidity requirements properly should be based on a formal
measure of liquidity risk. It will be interesting to see whether the
Basel III metrics — the inflow and outflow rates and ASF and RSF
factors — will someday be perceived as arbitrary, just as the Basel I
risk weights are perceived today. And the new accord, like Basel I,
distinguishes between good liquid assets and better liquid assets. It
will also be interesting to see how well the classifications of Level 1
and Level 2 high-quality liquid assets hold up to the creativity of
the capital markets. Research efforts like that underpinning the
Klomp and de Haan (forthcoming) are a welcome start, so perhaps
20 years need not pass for our understanding of liquidity to catch
up with our understanding of capital.
That said, the most important difference between our relative
understandings is not in the details of the models or the fine print
governing capital-market instruments. What has emerged over the
past 20 years is an accepted narrative on capital, and in particular,
of procyclicality. When times are good, our regulatory approach
and supporting models allow credit to expand, but when times
9
(It should be noted however that there is one key difference between the
minimum capital requirements and the minimum liquidity requirements. Banks are
prohibited from operating if their capital falls below the minimum. However, the
buffer which the liquidity provisions create is intended to be used ‘‘in times of stress’’.
(See the press release of the Group of Governors and Heads of Supervision (the
oversight body of the BCBS), http://www.bis.org/press/p120108.htm.) In this regard,
the liquidity buffer is similar to the countercyclical capital buffer of the Basel III
capital initiative. Like the capital buffer, though, the exact circumstances under which
banks can use the liquidity buffer have not yet been specified).
3132 B. Arnold et al. / Journal of Banking & Finance 36 (2012) 3125–3132
are bad, the same mechanism causes credit to contract, with brutal
consequences.
At present, the liquidity narrative has not yet been written.
Twenty years ago, a policy initiative — regulating capital — sparked
a generation of research. Today sees another policy initiative —
regulating liquidity — once again opening up a fertile and as of
yet largely unexplored field.
6. Conclusion
This paper has explored various issues that need to be tackled to
promote greater financial stability – one of the great challenges of
our time. It has highlighted that our understanding of financial
instability has made considerable progress in several areas: in
the measurement of systemic risk and of individual institutions
contribution to it; in the properties of the financial cycle and pro-
cyclicality; in the signals that point to the build-up of financial, at
the level of both individual institutions and the system as a whole,
and in the role that capital and liquidity play in making the system
more resilient. To varying degrees, this progress is being translated
into policy, as most clearly indicated by Basel III.
At the same time, there is still a lot we do not yet understand
about systemic risk and about how best to design policy to address
it. This will be a major challenge in the years ahead, for both
researchers and policymakers.
Disclaimer
The views expressed by Bruce Arnold, Claudio Borio, Luci Ellis in
this article are theirs and not necessarily those of the Australian
Prudential Regulatory Authority, the Bank for International Settle-
ments and the Reserve Bank of Australia respectively. Fariborz
Moshirian would like to acknowledge the support of the Australian
Research Council. He also would like to thank Mark Humphrey–
Jenner for his editorial assistance and Sidharth Sahgal for his input
on issues related to systemic risk.
References
Acharya, V.V., Richardson, M., 2009. Restoring Financial Stability: How to Repair a
Failed System. John Wiley and Sons, New York.
Acharya, V.V., Pedersen, L.H., Philippon, T., Richardson, M., 2010. Measuring
Systemic Risk (Working Paper). New York University.
Acharya, V.V., Engle, R., Richardson, M., 2012. Capital shortfall: A new approach to
ranking and regulating systemic risk. American Economic Review 102, 59–64.
Aebi, V., Sabato, G., and M. Schmid, M., forthcoming. Risk management, corporate
governance, and bank performance in the financial crisis. Journal of Banking and
Finance.
Affinito, M., forthcoming. Do interbank customer relationships exist? And how did
they function. Journal of Banking and Finance.
Allen, F., Gu, X., Kowalewski, O., forthcoming. Financial crisis, structure and reform.
Journal of Banking and Finance.
Altunbas, Y., Gambacorta, L., Marques-Ibanez, D., 2009. An empirical assessment of
the risk-taking channel. Available at SSRN: <http://ssrn.com/
abstract=1459627>.
Barth, J., Caprio, G., Levine, R., 2004. Bank regulation and supervision: what works
best? Journal of Financial Intermediation 13, 205–248.
BCBS, 2010a. Countercyclical capital buffer proposal – consultative document
(consultative document). Basel Committee for Banking Supervision.
BCBS, 2010b. Guidance for national authorities operating the countercyclical capital
buffer (REPORT). Basel Committee for Banking Supervision.
Billio, M., Getmansky, M., Lo, A.W., Pelizzon, L., 2012. Econometric measures of
connectedness and systemic risk in the finance and insurance sectors. Journal of
Financial Economics 104, 535–559.
Bisias, D., Flood, M.D., Lo, A.W., Valavanis, S., 2012. A survey of systemic risk analytic
(US Department of Treasury, Office of Financial Research Paper No. 0001). US
Department of Treasury, Office of Financial Research.
Borio, C., 2011. Rediscovering the macroeconomic roots of financial stability policy:
journey, challenges and a way forward. Annual Review of Financial Economics
3, 87–117.
Borio, C., Drehmann, M., 2009. Assessing the risk of banking crises – revisited. BIS
Quarterly Review, 29–46, March.
Borio, C., Drehmann, M., Tsatsaronis, K., forthcoming. Stress-testing macro-stress
tests: does it live up to expectations? Journal of Financial Stability.
Borio, C., Zhu, H., 2008. Capital regulation, risk-taking and monetary policy: A
missing link in the transmission mechanism? BIS Working Paper No 268.
December.
Boyd, J.H., De Nicolo, G., 2005. The Theory of bank risk taking and competition
revisited. Journal of Finance 60, 1329–1343.
Boyd, J.H., Chang, C., Smith, B.D., 1998. Moral hazard under commercial and
universal banking. Journal of Money Credit and Banking 30, 426–468.
Brownlees, C., Engle, R., 2010. Volatility, Correlation and tails for systemic risk
measurement (Working Paper). New York University.
Brunnermeir, M.K., Pedersen, L.H., 2009. Market liquidity and funding liquidity.
Review of Financial Studies 22, 2201–2238.
Caruana, J., 2010. Macroprudential policy: could it have been different this time?
(Speech at the People’s Bank of China seminar on Macroprudential Policy),
Shanghai.
Chari, V., Jagannathan, R., 1988. Banking panics, information, and rational
expectations equilibrium. Journal of Finance 43, 749–760.
De Jonghe, O., 2010. Back to the basics of banking? A micro-analysis of banking
system stability. Journal of Financial Intermediation 19, 387–417.
Demigurc-Kunt, A., Huizinga, H., 2010. Bank activity and funding strategies: the
impact on risk and returns. Journal of Financial Economics 98, 626–650.
Diamond, D., Rajan, R., 2005. Liquidity shortages and banking crises. Journal of
Finance 60, 615–647.
Drehmann, M., Tarashev, N., 2011. Measuring the systemic importance of
interconnected banks (BIS Working paper No. 342). Bank for International
Settlements.
Drehmann, M., Borio, C., Tsatsaronis, K., 2011. Anchoring countercyclical capital
buffers: the role of credit aggregates. International Journal of Central Banking 7,
189–240.
Drehmann, M., Borio, C., Tsatsaronis, K., 2012. Characterising the financial cycle: do
not lose sight of the medium term! (Working Paper No. 380). Bank for
International Settlements.
Duan, J.C., Van Laereb, E., forthcoming. A public good approach to credit ratings:
From conceptulalisation to implementation, Journal of Banking and Finance.
Engle, R., Moshirian, F., Sahgal, S., Zhang, B., 2012. Non-interest income and
systemic risk: The role of concentration (Working Paper). University of New
South Wales.
Galbraith, J., 2009. Who are these economists anyway? Thought and Action 2009,
85–97.
Gennaioli, N., Shleifer, A., Vishny, R., 2012. Neglected risks, financial innovation, and
financial fragility. Journal of Financial Economics 104, 452–468.
Goodhart, C.W.L., 1984. Monetary Theory and Practice. Macmillan, London.
Griffin, J., Kelly, P.J., Nardari, F., 2010. Do market efficiency measures yield correct
inferences? A comparison of developed and emerging markets. Review of
Financial Studies 23, 3225–3277.
Hellman, T., Murdock, K., Stiglitz, J.E., 2000. Liberalization, moral hazard in banking
and prudential regulation: are capital controls enough? American Economic
Review 90, 147–165.
Humphery-Jenner, M.L., 2012b. The impact of the EU takeover directive on takeover
performance and empire building. Journal of Corporate Finance 18, 254–272.
Humphery-Jenner, M., forthcoming. The desirability of ‘‘Weak’’ form legal
harmonization: Perspectives from statutory interpretation and legal
coherence. German Law Journal.
IMF, 2009. The Financial Crisis and Information Gaps (Report to G-20 Finance
Ministers and Central Bank Governors by the IMF Staff and the FSB Secretariat).
International Monetary Fund.
Klomp, J., de Haan, J., forthcoming. Banking risk and regulation: does one size fit all?
Journal of Banking and Finance.
Leijonhufvud, A., 2006. Agent-Based Macro. In: Handbook of Computational
Economics. Elsevier Publisher, North-Holland, Amsterdam, pp. 1625–1637.
Liu, M., Staum, J., 2010. Systemic risk components in a network model of contagion.
Northwestern University.
MacNeil, A.J., Frey, R., Embrechts, P., 2005. Quantitative risk management: concepts,
techniques and tools. Princeton University Press, Princeton.
Moshirian, F., 2011. Global financial crisis and the evolution of markets and
institutions. Journal of Banking and Finance 35, 502–511.
Reinhart, C., Rogoff, K., 2009a. This time is different: eight centuries of financial
folly. Princeton University Press, Princeton.
Reinhart, C.M., Rogoff, K., 2009b. The aftermath of financial crises. American
Economic Review 99, 466–472.
Rochet, J.C., Tirole, J., 1996. Interbank lending and systemic risk. Journal of Money,
Credit and Banking 28, 733–762.
Shi, Z., Werkery, B., forthcoming. Economic costs and benefits of imposing short-
horizon value-at-risk type regulation. Journal of Banking and Finance.
Tarashev, N., Borio, C., Tsatsaronis, K., 2010. Attributing systemic risk to individual
institutions (BIS Working Paper No. 308). Bank for International Settlements.
Tesfatsion, L., 2006. Agent-based computational economics: a constructive
approach to economic theory. In: Handbook of Computational Economics.
Elsevier Publisher, North-Holland, Amsterdam, pp. 831–880.
Trichet, J.-C., 2010. Macroprudential regulation as an approach to contain systemic
risk: Economic foundations, diagnostic tools and policy instruments (Speech at
the 13th conference of the ECB-CFS Research Network). European Central Bank.
Weatherill, S., 2012. The limits of legislative harmonization 10 years after tobacco
advertising: how the court’s case law has become a ‘‘Drafting Guide’’. German
Law Journal 12, 827–864.
White, W.R., 2008. Globalisation and the determinants of domestic inflation (BIS
Working Paper No. 250). Bank for International Settlements.