Measuring Systemic Risk

Viral V. Acharya
New York University, Stern School of Business, CEPR, and NBER

Lasse H. Pedersen
Copenhagen Business School, New York University, AQR Capital
Management, and CEPR

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

Thomas Philippon
New York University, Stern School of Business, CEPR, and NBER

Matthew Richardson
New York University, Stern School of Business, and NBER

We present an economic model of systemic risk in which undercapitalization of the financial

sector as a whole is assumed to harm the real economy, leading to a systemic risk externality.
Each financial institution’s contribution to systemic risk can be measured as its systemic
expected shortfall (SES), that is, its propensity to be undercapitalized when the system as
a whole is undercapitalized. SES increases in the institution’s leverage and its marginal
expected shortfall (MES), that is, its losses in the tail of the system’s loss distribution. We
demonstrate empirically the ability of components of SES to predict emerging systemic
risk during the financial crisis of 2007–2009. (JEL G01, G21, G28, D62, H23)

Received December 1, 2015; editorial decision August 5, 2016 by Editor Andrew Karolyi.

Widespread failures and losses of financial institutions can impose an

externality on the rest of the economy, and the global financial crisis of
2007–2009 provides ample evidence of the importance of containing this risk.
However, current financial regulations, such as Basel capital requirements, are
designed to limit each (or representative) institution’s risk seen in isolation;
they are not sufficiently focused on systemic risk even though systemic risk is

We would like to thank Rob Engle for many useful discussions. We are grateful to Christian Brownlees,
Farhang Farazmand, Hanh Le, and Tianyue Ruan for excellent research assistance. We also received useful
comments from Tobias Adrian, Mark Carey, Matthias Drehman, Dale Gray, and Jabonn Kim (discussants),
Andrew Karolyi (editor), and seminar participants at several central banks and universities where the current
paper and related systemic risk rankings at [Link]/welcome/risk have been presented. Pedersen
gratefully acknowledges support from the European Research Council (ERC grant no. 312417) and the
FRIC Center for Financial Frictions (grant no. DNRF102). Send correspondence to Viral Acharya, New York
University, Stern School of Business, 44 West 4th St., New York, NY 10012; telephone: (212) 998-0354.
E-mail: vacharya@[Link].

© The Author 2016. Published by Oxford University Press on behalf of The Society for Financial Studies.
All rights reserved. For Permissions, please e-mail: [Link]@[Link].
doi:10.1093/rfs/hhw088 Advance Access publication October 19, 2016
 Measuring Systemic Risk

often the rationale provided for such regulation. As a result, while individual
risks may be properly dealt with in normal times, the system itself remains, or
in some cases is induced to be, fragile and vulnerable to large macroeconomic
shocks.1
The goal of this paper is to propose and apply a useful and model-based
measure of systemic risk. To this end, we first develop a framework for
formalizing and measuring systemic risk. Using this framework, we derive
an optimal policy for managing systemic risk. Finally, we provide a detailed
empirical analysis of how our ex ante measure of systemic risk can predict the

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

ex post losses during the financial crisis of 2007–2009 as well as the regulators’
“stress test” in the spring of 2009.
The need for economic foundations for a systemic risk measure is more than
an academic concern since regulators around the world consider how to reduce
the risks and costs of systemic crises.2 It is of course difficult, if not impossible,
to find a systemic risk measure that is at the same time practically relevant and
completely justified by a general equilibrium model. In fact, the gap between
theoretical models and the practical needs of regulators has been so wide that
measures such as institution-level value-at-risk (VaR), designed to address the
risk of an individual institution, have persisted in regulation-assessing risks of
the financial system as a whole (Allen and Saunders 2002).
We “bridge this gap” by studying a theoretical model that is based on the
common denominator of various general equilibrium models yet simple enough
to provide clear recommendations relying on well-known statistical measures.
Our model is based on the basic idea that the main reasons for regulating
financial institutions are that (i) failing banks impose costs due to insured
creditors and bailouts; and (ii) undercapitalization of the financial system leads
to externalities that spill over to the rest of the economy.3 Interestingly, even a
relatively simple model is enough to obtain a rich new theory of systemic risk
regulation with strong empirical content.
Our theory considers a number of financial institutions (“banks”) that must
decide on how much capital to raise and which risk profile to choose in order
to maximize their risk-adjusted return. A regulator considers the aggregate
outcome of banks’ actions, additionally taking into account each bank’s insured

1 See Crockett (2000) and Acharya (2009) for a recognition of this inherent tension between micro-prudential and
macro-prudential regulation of the financial sector.
2 Some examples are the “crisis responsibility fee” proposed by the Obama administration (White House press
release, January 14, 2010) and the systemic risk levy advocated by the International Monetary Fund (Global
Financial Stability Report, International Monetary Fund, April 2010).
3 This assumption is consistent with models that spell out the exact nature of the externality, such as models of
(i) financial contagion through interconnectedness (e.g., Rochet and Tirole 1996); (ii) pecuniary externalities
through fire sales (e.g., several contributions in Allen and Gale 2007 and Acharya and Yorulmazer 2007), margin
requirements (e.g., Garleanu and Pedersen 2007), liquidity spirals (e.g., Brunnermeier and Pedersen 2009),
and interest rates (e.g., Acharya 2009; Diamond and Rajan 2005); (iii) runs (e.g., Diamond and Dybvig 1983;
Pedersen 2009); and (iv) time-inconsistency of regulatory actions that manifests as excessive forbearance and
induces financial firms to herd (Acharya and Yorulmazer 2007; Farhi and Tirole 2009).

3
 The Review of Financial Studies / v 30 n 1 2017

losses during an idiosyncratic bank failure and the externality arising in a

systemic crisis, that is, when the aggregate capital in the banking sector is
sufficiently low. The pure market-based outcome differs from the regulator’s
preferred allocations since, due to limited liability, banks do not take into
account the loss they impose in default on guaranteed creditors and the
externality they impose on the economy at large in a systemic crisis.
We show that to align incentives, the regulator optimally imposes a tax
on each bank that is related to the sum of its expected default losses and
its expected contribution to a systemic crisis, which we denote the systemic

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

expected shortfall (SES).4 Importantly, this means that banks have an incentive
to reduce their tax (or insurance) payments and thus take into account the
externalities arising from their risks and default. Additionally, it means that
they pay in advance for any support given to the financial system ex post.
We show that SES, the systemic-risk component, is equal to the expected
amount a bank is undercapitalized in a future systemic event in which the
overall financial system is undercapitalized. Said differently, SES increases in
the bank’s expected losses during a crisis. SES is therefore measurable, and we
provide theoretical justification for it being related to a financial firm’s marginal
expected shortfall, MES (i.e., its losses in the tail of the aggregate sector’s loss
distribution), and to its leverage.
We empirically investigate three examples of emerging systemic risk in
the financial crisis of 2007–2009 and analyze the ability of our theoretically
motivated measures to capture this risk ex ante.5 Specifically, we look at how
our measures of systemic risk estimated ex ante predict the ex post realized
systemic risk as measured, respectively, by (i) the capital shortfalls at large
financial institutions as assessed in the regulator’s stress tests during the spring
of 2009, (ii) the actual drop in equity values of large financial firms during the
crisis, and (iii) the increase in credit risk estimated from credit default swaps
(CDS) of large financial firms during the crisis.
We note that MES is very simple to estimate: one can simply calculate each
firm’s average return during the 5% worst days for the market. This measures
how exposed a firm is to aggregate tail shocks and, interestingly, together with
leverage, it has a significant explanatory power for which firms contribute
to a potential crisis, consistent with our theory. On the other hand, we find
that standard measures of institution-level risk such as expected loss in an
institution’s own left tail and volatility have little explanatory power. Moreover,

4 Using a variant of SES, called SRISK, the Volatility Institute at the NYU Stern School of Business publishes
Systemic Risk Rankings, providing estimates of the expected capital shortfall of global financial firms given a
systemic crisis (see [Link] For recent work either using or discussing SES,
see, among others, Acharya, Engle, and Pierret (2013), Acharya, Engle, and Richardson (2012), Allen, Bali,
and Tang (2012), Bostandzic and Weiss (2015), Brownlees and Engle (Forthcoming), Brownlees et al. (2015),
Brunnermeier, Dong, and Palia (2011), Cummins and Weiss (2014), Engle, Jondeau, and Rockinger (2014),
Giesecke and Kim (2011), Hansen (2014), and Huang, Zhou, and Zhu (2009, 2012).
5 Our systemic risk measure is provided in real time at [Link]

4
 Measuring Systemic Risk

the standard measure of covariance, namely beta, also has less explanatory
power than the measures we propose.
Turning to the literature, one strand of recent papers on systemic risk
takes a structural approach using contingent claims analysis of the financial
institutions’ assets (Lehar 2005; Gray, Merton, and Bodie 2008; Gray and
Jobst Forthcoming). There are complexities in applying the contingent claims
analysis in practice due to the strong assumptions that need to be made about
the liability structure of the financial institutions. As an alternative, some
researchers have used market data to back out reduced-form measures of

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

systemic risk.6 For example, Huang, Zhou, and Zhu (2009) use data on credit
default swaps (CDSs) of financial firms and stock return correlations across
these firms to estimate expected credit losses above a given share of the financial
sector’s total liabilities. Similarly, Adrian and Brunnermeier (Forthcoming)
measure the financial sector’s VaR given that a bank has had a VaR loss, which
they denote CoVaR, using quantile regressions. Their measure uses data on
market equity and book value of the debt to construct the underlying asset.
Adrian and Brunnermeier’s approach has the advantage of framing the analysis
using the standard regulatory tool of VaR, though regulators should also care
about expected losses beyond the VaR threshold. Billio et al. (2012) measure
systemic risk through Granger causality (i.e., autocovariances) across and
within different parts of the financial sector. de Jonghe (2010) presents estimates
of tail betas for European financial firms as their systemic risk measure. Borio,
Tarashev, and Tsatsaronis (2009) present a game-theoretic formulation that also
provides a possible allocation of capital charge to each institution based on its
systemic importance. Finally, Segoviano and Goodhart (2009) also view the
financial sector as a portfolio of individual financial firms, and look at how
individual firms contribute to the potential distress of the system by using the
CDSs of these firms within a multivariate setting.
We “bridge the gap” between the structural and reduced-form approaches by
considering a simple economic model that gives rise to a measure of systemic
risk contribution that depends on observable data and statistical techniques that
are related to those in the reduced-form approaches and easily applicable by
regulators. Since our systemic risk measure arises from a model, this ensures
that it is logically consistent and is measured in natural units that make it usable
as a basis for a systemic tax. For example, it has natural additivity properties if
firms merge or divisions are spun off, scales naturally with the size of the firm,
and so on—as opposed to many of the reduced-form approaches.
Our theoretical model potentially also provides an economic foundation
for the systemic risk measures proposed by de Jonghe (2010), Goodhart and
Segoviano (2009) and Huang, Zhou, and Zhu (2009). However, Adrian and
Brunnermeier’s (forthcoming) CoVaR measure is conceptually different from

6 See the survey of systemic risk methodologies by Bisias et al. (2012).

5
 The Review of Financial Studies / v 30 n 1 2017

our measure in that it examines the system’s stress conditional on an individual

firm’s stress, whereas we examine a financial firm’s stress conditional on
systemic stress. As a way of ranking the systemic risk of firms, our measure
has the advantage that the conditioning set is held constant for all firms (i.e.,
the existence of a financial crisis), whereas this is not the case with CoVaR
(i.e., conditional on a given firm’s stress, which varies cross-sectionally).
This can lead to some undesirable properties in the rankings. For example,
Acharya, Engle, and Richardson (2012) show that, under certain distributional
assumptions about firm’s returns, CoVaR treats two firms identically in terms

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

of systemic risk if the firms have the same return correlation with the aggregate
market even though they might have very different return volatilities.
In conclusion, we provide a simple economic framework for measuring
systemic risk, and our results have consequences for how macro-prudential
regulation can be achieved through a systemic tax, stress tests, or
recapitalization of financial firms during systemic crises.7

1. Systemic Risk in an Economic Model

1.1 Definitions and preliminary analysis
We start by reviewing the standard risk measures used inside financial firms and
discuss how these measures can be extended to apply for the whole financial
system.8 This preliminary analysis allows us to define some simple concepts
and generate an intuition that is useful in our model of systemic risk.
Two standard measures of firm-level risk are value-at-risk (VaR) and expected
shortfall (ES). These seek to measure the potential loss incurred by the firm as
a whole in an extreme event. Specifically, VaR is the most that the bank loses
with confidence 1−α, that is, Pr (R < −VaRα ) = α. The parameter α is typically
taken to be 1% or 5%. For example, with α = 5%, VaR is the most that the bank
loses with 95% confidence. The ES is the expected loss conditional on the loss
being greater than the VaR:
ESα = −E [R|R ≤ −VaRα ] (1)
Said differently, the expected shortfall is the average of returns on days when
the portfolio’s loss exceeds its VaR limit.
We focus on ES rather than VaR for several reasons. First, VaR is not robust in
the sense that asymmetric, yet very risky, bets may not produce a large VaR. The
reason is that if the negative payoff is below the 1% or 5% VaR threshold, then

7 Recent proposals (based among others on Raviv (2004), Flannery 2005; Kashyap, Rajan, and Stein 2008; Hart
and Zingales 2009; Duffie 2010) suggest requiring firms to issue “contingent capital,” which is debt that gets
automatically converted to equity when certain firm-level and systemic triggers are hit. Our systemic risk measures
correspond precisely to states in which such triggers will be hit, implying that it should be possible to use our
measures to predict which firms are more systemic and therefore will find contingent capital binding in more
states ex post.
8 See Lehar (2005) and Yamai and Yoshiba (2005) for a fuller discussion.

6
Measuring Systemic Risk

VaR does not capture it. Indeed, one of the concerns in the ongoing crisis has
been the failure of VaR to pick up potential “tail” losses in the AAA-tranches
of collateralized debt obligations (CDOs) and other structured products. In
contrast, ES does not suffer from this problem, since it measures all the losses
beyond the threshold. This distinction is especially important when considering
moral hazard of banks, because the large losses beyond the VaR threshold are
often borne by the government bailout. Second, VaR is not a coherent measure
of risk because the VaR of the sum of two portfolios can be higher than the sum
of their individual VaRs, which cannot happen with ES (Artzner et al. 1999).

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

For risk management, transfer pricing, and strategic capital allocation, banks
need to break down firm-wide losses into contributions from individual groups
or trading desks. To see how, let us decompose
 the bank’s return R into the sum
of each group’s return ri , that is, R = i yi ri , where yi is the weight of group i
in the total portfolio. From the definition of ES, we see that:

ESα = − yi E [ri |R ≤ −VaRα ]. (2)
i

From this expression we see the sensitivity of overall risk to exposure yi to

each group i:
∂ESα
= −E [ri |R ≤ −VaRα ] ≡ MESαi , (3)
∂yi
where MES i is group i’s marginal expected shortfall. The marginal expected
shortfall measures how group i’s risk taking adds to the bank’s overall risk. In
words, MES can be measured by estimating group i’s losses when the firm as
a whole is doing poorly.
These standard risk-management practices can be useful for thinking about
systemic risk. A financial system is constituted by a number of banks, just
like a bank is constituted by a number of groups. We can therefore consider the
expected shortfall of the overall banking system by letting R be the return of the
aggregate banking sector or the overall economy. Then each bank’s contribution
to this risk can be measured by its MES.

1.2 Banks’ incentives

We next present an economic model in which we consider the incentives of
financial firms and their systemic externalities. The economy has N financial
firms, which we denote as banks, indexed by i = 1,..N and two time periods
t = 0,1. Each bank i chooses how much xji to invest in each of the available
assets j = 1,..J , acquiring total assets a i of

J
ai = xji . (4)
j =1

These investments can be financed with debt or equity. In particular, the owner
of any bank i has an initial endowment w̄0i of which w0i is kept in the bank

7
 The Review of Financial Studies / v 30 n 1 2017

as equity capital and the rest is paid out as a dividend (and consumed or used
for other activities). The bank can also raise debt bi . Naturally the sum of the
assets a i must equal the sum of the equity w0i and the debt bi , giving the budget
constraint:
w0i +bi = a i . (5)

At time 1, asset j pays off rji per dollar invested for bank i (so the net return
is rji −1). We allow asset returns to be bank-specific to capture differences in
investment opportunities. The total market value of the bank assets at time 1

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

is y i = ŷ i −φ i where φ i captures the costs of financial distress and ŷ i is the
pre-distress income:

J
ŷ i = rji xji . (6)
j =1

The costs of financial distress depend on the market value of bank assets and
on the face value f i of the outstanding debt:
 
φ i =  ŷ i ,f i . (7)

Our formulation of distress costs is quite general. Distress costs can occur even
if the firm does not actually default. This specification captures debt overhang
problems as well as other well-known costs of financial distress. We restrict
the specification to φ ≤ ŷ so that y ≥ 0.
What is special about banks (versus other corporations) is that (i) they enjoy
government guarantees of parts of their debt, and (ii) their financial distress
can impose systemic-risk externalities. We first discuss the issue of guaranteed
debt and turn to systemic risk in the next section.
To capture various types of government guarantees, we assume that a fraction
α i of the debt is implicitly or explicitly guaranteed by the government. The face
value of the debt is set so that the debt holders break even, that is,
    
bi = α i f i + 1−α i E min f i ,y i . (8)

Although our focus is on systemic risk, we include government debt guarantees

because they are economically important and because we want to highlight
the different regulatory implications of deposit insurance and systemic risk.
The insured debt can be interpreted as deposits, but it can also cover implicit
guarantees.9

9 Technically, the pricing equation (8) treats the debt as homogeneous ex ante with a fraction being guaranteed ex
post. This is only for simplicity, and all of our results go through if we make the distinction between guaranteed
and non-guaranteed debt from an ex ante standpoint. In that case, the guaranteed debt that the bank can issue
would be priced at face value, while the remaining debt would be priced as above with α = 0. We adopt the general
formulation as it allows us to span the setting where a portion of bank debt, e.g., retail deposits up to a threshold
size, is guaranteed by a national deposit insurance agency.

8
Measuring Systemic Risk

The net worth of the bank, w1i , at time 1 is:

w1i = ŷ i −φ i −f i (9)
The owner of the bank equity is protected by limited liability so it receives
1wi >0 w1i and, hence, solves the following program:
1

 
max c × w̄0i −w0i −τ i +E u 1wi >0 ×w1i , (10)
w0i ,bi , xji 1
j

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

subject to Equations (5)–(9), where, ui (×) is the bank owner’s utility of time 1
income, w̄0i −w0i −τ i is the part of the initial endowment w̄0i that is consumed
immediately (or used for outside activities), and the remaining endowment is
kept as equity capital w0i or used to pay the bank’s tax τ i , which we describe
later. The parameter c has several interpretations. It can simply be seen as a
measure of the utility of immediate consumption, but, more broadly, it is the
opportunity cost of equity capital. We can think of the owner as raising capital
at cost c, or we can think of debt as providing advantages in terms of taxes or
incentives to work hard. What matters for us is that there is an opportunity cost
of using capital instead of debt.

1.3 Welfare, externalities, and the planner’s problem

The regulator wants to maximize the welfare function P 1 +P 2 +P 3 , which has
three parts: the first part is simply the sum of the utilities of all the bank owners,
 N 
N
 i  
P =1
c × w̄0 −w0 −τ +E
i i i 
u 1 wi >0 ×w1 . i
1
i=1 i=1

The second part,  


N
2
P =E g 1w1i <0
α i
w1i ,
i=1

is the expected cost of the debt insurance program, where the parameter g
captures administrative costs and costs of tax collection. The cost is paid
conditional on default by firm i and a fraction α i of the shortfall is covered.
The third part of the welfare function is the main focus of our analysis since
 
P 3 = E e ×1[W1 <zA] ×(zA−W1 )
captures the
externality of financial crisis, where each term is defined 
as follows.
First, A = N i
i=1 a are the aggregate assets in the system and W1 =
N i
i=1 w1 is
the aggregate banking capital to support it at time 1. A systemic crisis occurs
when the aggregate capital W1 in the financial system falls below a fraction z of
the assets A. The critical feature that we want to capture as simply as possible
is that of an aggregate threshold for capital needed to avoid early fire sales

9
 The Review of Financial Studies / v 30 n 1 2017

and restricted credit supply. The externality cost is zero as long as aggregate
financial capital is above this threshold and grows linearly when it falls below,
where the slope parameter e measures the severity of the externality imposed
on the economy when the financial sector is in distress.10
This formulation of a systemic crisis is consistent with the emphasis of the
stress tests performed by the Federal Reserve in the United States starting
in the spring of 2009,11 and in understanding the crucial difference between
systemic and institution-specific risk. It means that a bank failure occurring in a
well-capitalized system imposes no externality on the economy. This captures

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

well-known examples such as the idiosyncratic failure of Barings Bank in the
United Kingdom in 1995, which did not disrupt the global (or even the United
Kingdom’s) financial system. (That is, the Dutch bank ING purchased Barings
and assumed all of its liabilities with minimal government involvement and
no commitment of taxpayer money.) This stands in sharp contrast with the
failures of Bear Stearns or Lehman Brothers witnessed in 2008. When the
whole financial system has too little capital, then firms and consumers face
a credit crunch, which can lead to job losses, a recession, or perhaps even a
depression (see, for instance, the discussion in Acharya et al. 2009).
The planner’s problem is to choose a tax system τ i that maximizes the
welfare function P 1 +P 2 +P 3 subject to the same technological constraints as
the private agents. This ex ante (time 0) regulation is relevant for the systemic
risk debate, and this is the one we focus on. We do not allow the planner to
redistribute money among the banks at time 1 because we want to focus on
how to align ex ante incentives. Indeed, adding capital ex post to troubled
banks creates moral hazard, and here we focus on measuring and managing
systemic risk ex ante. In doing so, we follow the constrained efficiency analysis
performed in the liquidity provision literature. In this literature, the planner is
typically restricted to affect only the holding of liquid assets in the initial period
(see Lorenzoni 2008, for instance).
Lastly, we need to account for the taxes that the regulator collects at time 0
and the various costs borne at time 1. Since we focus on the financial sector

10 There is growing evidence on the large bailout costs and real economy welfare losses associated with banking
crises (see, for example, Caprio and Klingebiel 1996; Honohan and Klingebiel 2000; Hoggarth, Reis, and Saporta
2002; Reinhart and Rogoff 2008; Borio and Drehmann 2009; and more recently, Laeven and Valencia 2013;
Chodorow-Reich 2014; Acharya et al. 2015). The bottom line from these studies is that these crises represent
significant portions of GDP, on the order of 10–20%.
11 The Federal Reserve states on their website that “the Comprehensive Capital Analysis and Review (CCAR)
is an annual exercise by the Federal Reserve to assess whether the largest bank holding companies operating
in the United States have sufficient capital to continue operations throughout times of economic and financial
stress and that they have robust, forward-looking capital-planning processes that account for their unique risks.”
As part of this exercise, the Federal Reserve evaluates institutions’ capital adequacy, internal capital adequacy
assessment processes, and their individual plans to make capital distributions, such as dividend payments or
stock repurchases. Dodd-Frank Act stress testing (DFAST)—a complementary exercise to CCAR—is a forward-
looking component conducted by the Federal Reserve and financial companies supervised by the Federal
Reserve to help assess whether institutions have sufficient capital to absorb losses and support operations
during adverse economic conditions. For more details, see the Federal Reserve Board’s stress test website:
[Link]

10
Measuring Systemic Risk

and do not model the rest of the economy, we simply impose that the aggregate
taxes paid by banks at time 0 add up to a constant:

τ i = τ̄ . (11)
i

There are several interpretations for this equation. One is that the government
charges ex ante for the expected cost of the debt insurance program, making
it a self-funded entity. We can also add the expected cost of the externality. At
time 1, the government would simply balance its budget in each state of the

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

world with lump-sum taxes on the non-financial sector. We can also think of
Equation (11) as part of a larger maximization program, where a planner would
maximize utility of bank owners and other agents. This complete program
would pin down τ̄ , and we could then think of our program as solving the
problem of a financial regulator for any given level of transfer between the
banks and the rest of the economy.

1.4 Optimal taxation

Our optimal taxation policy depends on each bank’s expected capital shortfall
measured based on, respectively, institution-specific and systemic risk. First, it
depends on its expected shortfall (ES i ) in default:
 
ES i ≡ −E w1i |w1i < 0 (12)

Further, we introduce what we call a bank’s systemic expected shortfall (SES i ).

SES i is the amount a bank’s equity w1i drops below its “required” level—which
is a fraction z of assets a i —in case of a systemic crisis when aggregate banking
capital W1 is less than z times aggregate assets:
 
SES i ≡ E za i −w1i |W1 < zA (13)
Recall that a crisis happens when the aggregate capital W1 is below z times
aggregate assets A. This condition can be avoided if each bank keeps its own
capital above z times its own assets—hence, the “required” capital is the same
fraction z of assets for all banks. A bank that has positive SES is expected
to contribute to a future systemic crisis in the sense of failing to meet this
requirement during a future crisis. Therefore, SES is the key measure of each
bank’s expected contribution to a systemic crisis.
Using ES and SES, we can characterize a tax system that implements
the optimal allocation. The regulator’s problem is to choose the tax scheme
{τi }i=1,...N such as to mitigate systemic risk and inefficient effects of debt
guarantees. The timing of the implementation is that the banks choose their
leverage and asset allocations and then pay the taxes. The taxes are conditional
on choices made by the banks, which captures the idea that a regulator can
impose a higher tax on banks that take more systemic risk (or require certain
actions based on stress tests).

11
 The Review of Financial Studies / v 30 n 1 2017

Proposition 1. The efficient outcome is obtained by a tax

αi g e
τi = ×P r(w1i < 0)×ES i + ×P r(W1 < zA)×SES i +τ0 , (14)
c c
where τ0 is a lump sum transfer to satisfy Equation (11).

Proof. See Appendix A. 

This result is intuitive. Each bank must first be taxed based on its probability

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

of default Pr(w1i < 0), times the expected losses in default ES, to the extent
that those losses are insured by the government, where we recall that α i is
the fraction of insured debt. The tax should be lower if raising bank capital
is expensive (c > 1) and higher the more costly is government funding (g). A
natural case is simply to think of g/c = 1 so that this part of the tax is simply an
“actuarially fair deposit-insurance tax.”12 Hence, the first term in Equation
(14) corrects the underpricing of credit risk caused by the debt insurance
program. We note that this term is a measure of a bank’s own risk, irrespective
of its relation to the system, and it is similar to the current practice since
the calculation of the expected shortfall is similar to a standard value-at-risk
calculation.
The second part of the tax in Equation (14) depends on the probability
of a systemic crisis Pr(W1 < zA) and, importantly, the bank’s contribution to
systemic risk as captured by SES, namely the bank’s own loss during a potential
crisis. This tax is scaled by the severity e of the externality and scaled down
by the bank’s cost of capital c. This forces the private banks to internalize the
externality from aggregate financial distress.
We note that SES is based on a calculation that is similar to that of
marginal risk within financial firms discussed in Section 1.1. In a marginal
risk calculation, the risk managers ask how much a particular line of business
is expected to lose on days where the bank as a whole has a large loss (i.e.,
how much that particular line of business is expected to contribute to the overall
loss). Our formula applies this idea more broadly, namely to the financial system
as a whole.
The optimal tax system holds for all kinds of financial distress costs, and the
planner reduces its taxes when capital is costly at time 0 (c is high). The fact that
we obtain an expected shortfall measure comes from the shape of the externality
function. It is important to understand the information required to implement
the systemic regulation. The planner does not need to know the utility functions
and investment opportunity sets of the various banks. It needs to estimate two

12 Note that it is important for incentive purposes to keep charging this tax even if the deposit insurance reserve
fund collected over time has happened to become overfunded (in contrast to the current premium schedules of the
Federal Deposit Insurance Corporation [FDIC] in the United States). See, for example, the theoretical arguments
and the empirical evidence in Acharya, Santos, and Yorulmazer (2010).

12
 Measuring Systemic Risk

objects: the probability of an aggregate crisis, and the conditional loss of capital
of a particular firm if a crisis occurs. In practice, the planner may not be able
to observe or measure these precisely. Our empirical work to follow makes a
start in estimating one of the two objects, the conditional capital loss of a bank
in a crisis, using market-based data.

2. Measuring Systemic Risk

The optimal policy developed in Section 1.4 calls for a fee (i.e., a tax) equal

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

to the sum of two components: (i) an institution-risk component, that is, the
expected loss on its guaranteed liabilities, and (ii) a systemic-risk component,
namely, the expected systemic costs in a crisis (i.e., when the financial
sector becomes undercapitalized) times the financial institution’s percentage
contribution to this undercapitalization.
In practice, the planner needs to estimate the conditional expected losses
before a crisis occurs. Our theory says that the regulator should use any variable
that can predict capital shortfall in a crisis. In order to improve our economic
intuition and to impose discipline on our empirical analysis, it is important to
have a theoretical understanding of the variables that are likely to be useful for
these predictions. To this end, we explain the theoretical relationship between
SES and observed equity returns.
We can think of the systemic events in our model (W1 < zA) as extreme tail
events that happen once or twice a decade (or less), say. In the meantime, we
observe more “normal” tail events, that is, the more frequent “moderately bad
days.” Let us define these events as the worst 5% market outcomes at daily
frequency, which we denote by I5% . Based on these events, we can define a
marginal expected shortfall (MES) using net equity returns of firm i during
these bad market outcomes
 i 
w1
MES5% i
≡ −E −1 | I 5% .
w0i
A regulator needs to use the information contained in the “moderately bad days”
(MES5% i
) to estimate what would happen during a real crisis (SES).13 We can
use extreme value theory to establish a connection between the moderately bad
and the extreme tail. Specifically, let the return on security j for bank i follow
rji = ηji −δi,j εji −βi,j εm ,
where ηji follows a thin-tailed distribution (Gaussian, for instance), while εji and
εm follow independent normalized power law distributions with tail exponent ζ .

13 Note that if we assume returns are multivariate normal, then the drivers of the firm’s systemic risk would be
entirely determined by the expected return and volatility of the aggregate sector return and the firm’s return,
and their correlation. However, there is growing consensus that the tails of return distributions are not described
by multivariate normal processes and much more suited to that of extreme value theory (e.g., see Barro 2006;
Backus, Chernov, and Martin 2009; Gabaix 2009; Jiang and Kelly 2014). Our discussion helps clarify what
variables are needed to measure systemic risk in the presence of extreme values.

13
 The Review of Financial Studies / v 30 n 1 2017

The thin-tailed factor captures normal day-to-day changes, while the power
laws explain large events, both idiosyncratic (εji ) and aggregate (εm ). The
sensitivity to systemic risk of activity j in bank i is captured by the loading
βi,j . Since power laws dominate in the tail, we have the following simple
properties (Gabaix 2009). First, the VaR of rji at level α, for α sufficiently small,
 1/ζ
α −1/ζ , and the corresponding expected shortfall is
ζ ζ
α ≈ δi,j +βi,j
is VaRi,j
α ≈ ζ −1 VaRα . Second, the events I5% and (W1 < zA) correspond to the
ζ
ES i,j i,j

% S
critical values ε̄m and ε̄m of the systemic shock εm , and we can define the

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

relative severity as:
S
ε̄m
k≡ % .
ε̄m
Note that there is a direct link
 between1/ζ
the likelihood of an event and its tail
ε̄ S
size, since we have k = ε̄%m = Pr(W5%<zA) .
m 1
Then, using the power laws, we obtain the following proposition:

Proposition 2. The systemic expected shortfall is related to the marginal

expected shortfall according to

SES i za i −w0i i i
= +kMES5% + , (15)
w0i w0i
E [φ i |W1 <zA]−k×E [φ i |I5% ] (k−1)(f i −bi )
where i
≡ w0i
− w0i
.

Proof. See Appendix A. 

We see therefore that SES has three components: (i) the excess ex ante degree
of undercapitalization za i /w0i −1, (ii) the measured marginal expected shortfall
MES using pre-crisis data, scaled up by a factor k to account for the worse
performance in thetrue crisis, and
 (iii) an adjustment
 term i . The main part
of i
is the term E φ | W1 < zA −k ×E φ | I5% , which measures the excess
i i

costs of financial distress. The typical estimation sample contains bad market
days, but no real crisis. Weare therefore
 likely to miss most costs
 of financial

distress and to measure kE φ i | I5% ≈ 0. On the other hand, E φ i | W1 < zA is
probably significant, especially for highly levered large financial firms where
we expect large deadweight losses in a crisis.14
Based on this discussion, we therefore expect MES and leverage to be
predictors of SES. We now turn to the empirical analysis to test this prediction.

14 The second part of i measures the excess returns on bonds due to credit risk (f i −bi ). This second part is likely
to be quantitatively small because ex ante credit spreads are relatively small.

14
 Measuring Systemic Risk

3. Empirical Analysis of the Crisis of 2007–2009

We consider whether our model-implied measures of systemic risk—measured
before the crisis—can help predict which institutions actually did contribute to
the systemic crisis of 2007-2009. We are interested in predicting the systemic
expected shortfall, SES (see Section 1.4). Using the results of Section 2, we
show below that SES can be estimated using the marginal expected shortfall
MES and leverage.
To control for each bank’s size, we scale by initial equity w0i , which gives
the following cross-sectional variation in systemic risk SES:

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

 i 
SES i za i w1
= −1−E −1 | W 1 < zA .
w0i w0i w0i

The first part, za i /w0i −1, measures whether the leverage a i /w0i is initially
already “too high.” Specifically, since systemic crises happen when aggregate
bank capital falls below z times assets, z times leverage should be less
than 1. Hence, a positive value of za i /w0i −1 means that the bank is already
undercapitalized at time 0 in the sense that the capital w0i is low relative to the
assets a i .15 The second term is the expected equity return conditional on the
occurrence of a crisis. Hence, the sum of these two terms determine whether
the bank will be undercapitalized in a crisis and by what magnitude.
We estimate MES at a standard risk level of α=5% using daily data of equity
returns from the Center for Research in Security Prices (CRSP). This means
that we take the 5% worst days for the market returns (R) in any given year,
and we then compute the equal-weighted average return on any given firm (R b )
for these days:
1 
b
MES5% = Rtb (16)
#days
t: system is in its 5% tail

Even though the tail days in this average before the crisis do not capture the
tails of a true financial crisis, our power law analysis in Section 2 shows how
it is linked nevertheless.
It is not straightforward to measure true leverage due to limited and infrequent
market data, especially on the breakdown of off- and on-balance sheet financing.
We apply the standard approximation of leverage, denoted LVG:
quasi-market value of assets book assets – book equity + market equity
LVGb = =
market value of equity market value of equity
(17)
The book-value characteristics of firms are available at a quarterly frequency
from the CRSP-Compustat merged dataset.

15 We can think of z as being in the range of 8% to 12% if all assets have risk-weighting of close to 100% under
Basel I capital requirements.

15
The Review of Financial Studies / v 30 n 1 2017

As a first look at the data, Appendix C lists the U.S. financial firms with a
market capitalization of at least $5 billion as of June 2007. For each of these
firms, Appendix C provides the realized SES during the financial crisis, the
MES using the prior year of data, the leverage of the firm using Equation
(17), the quasi-market value of the assets, and the firm’s fitted SES rank from
a cross-sectional regression of realized SES on MES, leverage, and industry
characteristics. As an illustration, consider Bear Stearns, the first of the major
financial firms to effectively fail during the crisis. As of June 2007, Bear Stearns
ranked third in MES (i.e., its average loss on 5% worst-case days of the market

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

was 3.15%), first in leverage (i.e., its quasi-market assets to market equity ratio
was 25.62), and not surprisingly, first in fitted SES rank.
Going into the crisis, the next four highest ranked firms in terms of fitted SES
are Freddie Mac, Fannie Mae, Lehman Brothers, and Merrill Lynch. Aside from
the insurance giant AIG, these were the next four largest financial firms to run
aground during the crisis, either through government receivership, bankruptcy,
or sale. Of some note, unlike Bear Stearns, Lehman Brothers, and Merrill Lynch,
Freddie Mac and Fannie Mae rank high in leverage, yet less so in MES. This
observation highlights the importance of both MES and leverage in terms of
systemic risk. While MES measures the firm’s expected capital losses during
a crisis, these losses matter most in the aggregate to the extent the firm is
poorly capitalized as Freddie Mac and Fannie Mae were. As a final comment,
note that SES is on a per-dollar basis. The level of systemic risk is scaled up
by the firm’s assets. For example, while Bear Stearns had considerable assets
(i.e., $423 billion), Freddie Mac, Fannie Mae, and Merrill Lynch all had twice
the assets of Bear Stearns (i.e., $822 billion, $858 billion, and $1.076 billion,
respectively), leading to a higher level of absolute systemic risk.
These observations are suggestive of the potential use of the methodology
of Section 2. In this section, we take a more thorough look at systemic risk
using this methodology. Specifically, we analyze the ability of our theoretically
motivated measure to capture realized systemic risk in three ways: (i) the capital
shortfalls at large financial institutions estimated via stress tests performed by
bank regulators during the spring of 2009; (ii) the realized systemic risk that
emerged in the equity of large financial firms from July 2007 through the end
of 2008; and (iii) the realized systemic risk that emerged in the credit default
swaps of large financial firms from July 2007 through the end of 2008. As we
will see, the simple measures of ex ante systemic risk implied by the theory
have useful information for which firms ran aground during the financial crisis.

3.1 The stress test: Supervisory Capital Assessment Program

At the peak of the financial crisis, in late February 2009, the government
announced a series of stress tests were to be performed on the 19 largest
banks over a two-month period. Known as the Supervisory Capital Assessment
Program (SCAP), the Federal Reserve’s goal was to provide a consistent
assessment of the capital held by these banks. The question asked of each

16
 Measuring Systemic Risk

bank was how much of an additional capital buffer, if any, each bank would
need to make sure it had sufficient capital if the economy got “even worse” in
the sense of specific stress scenarios defined by the Fed and then supervised by
its examiners. In early May of 2009, the results of the analysis were released
to the public at large. A total of 10 banks were required to raise $74.6 billion
in capital. The SCAP was generally considered to be a credible test with bank
examiners imposing severe loss estimates on residential mortgages and other
consumer loans, not seen since the Great Depression. The market appeared to
react favorably to having access to this information on the extent of systemic

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

risk.
At first glance, the “bottom-up” risk assessment of the SCAP would seem to
be very different from our SES measure. However, this stress test is very much
in the spirit of SES since it aims at estimating each bank’s capital shortfall in a
common potential future crisis and the total shortfall across banks. Hence, it is
interesting to consider how our simple statistical measures of systemic capital
shortfall compare to the outcome of the regulator’s in-depth analysis. Of course,
we must naturally recognize that our measures are based on much less data than
the detailed data available to regulators. It nevertheless is interesting to compare
and, more broadly, to note that the regulators are essentially computing systemic
risk as in our model when they perform stress tests, just based on other data
and statistical methods.
The regulators spent two months examining the portfolios and financing of
the largest banks with a particular emphasis on creating consistent valuations
across these banks. Table 1, Panel A, provides a summary of each bank,
including (i) its shortfall (if any) from the SCAP at the end of April 2009, (ii) its
Tier 1 capital (so-called core capital including common shares, preferred shares,
and deferred tax assets), (iii) its tangible common equity (just its common
shares), along with our measured MES (from April 2008 to March 2009),
and (iv) its quasi-market leverage. Five banks, as a percentage of their Tier 1
capital, had considerable shortfalls, namely Regions Financial (20.66%), Bank
of America (19.57%), Wells Fargo (15.86%), Keycorp (15.52%), and Suntrust
Banks (12.50%).16
The SCAP can be considered as close as possible to an ex ante estimate of
expected losses of different financial firms in a financial crisis in the spirit of
our measure of systemic risk. Panel B of Table 1 provides the correlation across
firms between the banks’ SCAP/Tier 1 and the banks’ MES and leverage. The
correlations are large and positive, 59.5% and 31.6%, respectively. Consistent

16 The interested reader might be surprised to see that, although it required additional capital, Citigroup was not
one of the most undercapitalized. It should be pointed out, however, that toward the end of 2008 (and thus prior
to the SCAP), Citigroup received $301 billion of federal asset guarantees on their portfolio of troubled assets.
Conversations with the Federal Reserve confirm that these guarantees were treated as such for application of the
stress test. JPMorgan and Bank of America also received guarantees (albeit in smaller amounts) through their
purchase of Bear Stearns and Merrill Lynch, respectively. We also note that the SCAP exercise also included
GMAC, but it only had preferred stock trading over the period analyzed.

17
 The Review of Financial Studies / v 30 n 1 2017

Table 1
Banks included in the stress test, descriptive statistics

Panel A
Bank Name SCAP Tier1 Tier1 SCAP/ SCAP/Tier1 MES LVG
Comm Tier1 Comm
REGIONS FINANCIAL CORP NEW 2.5 12.1 7.6 20.66% 32.89% 14.8 44.42
BANK OF AMERICA CORP 33.9 173.2 75 19.57% 45.50% 15.05 50.38
WELLS FARGO & CO NEW 13.7 86.4 34 15.86% 40.41% 10.57 20.58
KEYCORP NEW 1.8 11.6 6 15.52% 30.00% 15.44 24.36
SUNTRUST BANKS INC 2.2 17.6 9.4 12.50% 23.40% 12.91 39.85
FIFTH THIRD BANCORP 1.1 11.9 4.9 9.24% 22.45% 14.39 67.16

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

CITIGROUP INC 5.5 118.8 23 4.63% 24.02% 14.98 126.7
MORGAN STANLEY DEAN WITTER & CO 1.8 47.2 18 3.81% 10.11% 15.17 25.39
P N C FINANCIAL SERVICES GRP INC 0.6 24.1 12 2.49% 5.13% 10.55 21.58
AMERICAN EXPRESS CO 0 10.1 10 0.00% 0.00% 9.75 7.8
B B & T CORP 0 13.4 7.8 0.00% 0.00% 9.57 14.78
BANK NEW YORK INC 0 15.4 11 0.00% 0.00% 11.09 6.46
CAPITAL ONE FINANCIAL CORP 0 16.8 12 0.00% 0.00% 10.52 33.06
GOLDMAN SACHS GROUP INC 0 55.9 34 0.00% 0.00% 9.97 18.94
JPMORGAN CHASE & CO 0 136.2 87 0.00% 0.00% 10.45 20.43
METLIFE INC 0 30.1 28 0.00% 0.00% 10.28 26.14
STATE STREET CORP 0 14.1 11 0.00% 0.00% 14.79 10.79
U S BANCORP DEL 0 24.4 12 0.00% 0.00% 8.54 10.53
Panel B: Correlation matrix
SCAP/Tier1 SCAP/Tier1Comm MES LVG
SCAP/Tier1 100.00%
SCAP/Tier1Comm 95.42% 100.00%
MES 59.48% 61.47% 100.00%
LVG 31.58% 48.20% 53.70% 100.00%
Panel A of this table contains the values of SCAP shortfall, Tier1 capital, Tier1Comm (tangible common equity),
all in U.S. $billion, and SCAP Shortfall/Tier1, SCAP Shortfall/Tier1Comm, MES, and LVG for the 18 banks
who underwent stress testing. MES is the marginal expected shortfall of a stock given that the market return is
below its 5th percentile. Leverage (LVG) is measured as quasi-market value of assets divided by market value of
equity, where quasi-market value of assets is book value of assets minus book value of equity + market value of
equity. All stock market data are from Datastream and book value of equity is from the merged CRSP-Compustat
database. MES was measured for each individual company’s stock using the period April 2008 to March 2009
and the S&P 500 as the market portfolio. LVG is as of first quarter 2009. Panel B shows the correlation between
SCAP Shortfall/Tier1, SCAP Shortfall/Tier1Comm, MES, and LVG.

with this result, Figure 1 shows our measure of MES is linked positively in
the cross-section of stress-tested financial institutions to their capital shortfall
assessed by the stress test.17
To further test the link between the capital shortfall assessed by the stress test
and our measures of systemic risk, Table 2 provides an OLS regression analysis
of explaining SCAP shortfall as a percent of Tier 1 capital (Panel A) and Tier
1 common or tangible common equity (Panel B) with MES and leverage as the
regressors. Because a number of firms have no shortfall, and thus there is a mass
of observations at zero, we also extend the OLS regressions to a probit analysis
(which is identical for both panels and hence is shown only in Panel A).
MES is strongly significant in both the OLS and probit regressions. For
example, in the OLS regressions of MES on SCAP shortfall relative to Tier

17 Appendix E provides the map between abbreviated and full financial institution names.

18
Measuring Systemic Risk

.5
.4 BAC

WFC
SCAP/Tier1Comm

RF
.3

KEY

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

C
STI
FITB
.2 .1

MS

PNC
0

USB MET
BBT STT GSJPM
AXP COF
BK

4 5 6 7 8 9
MES5 measured Oct06-Sep08

Figure 1
MES predicts the stress tests
The marginal expected shortfall measure (MES), a measure of ex ante systemic risk, plotted against the stress
tests’ assessed capital shortfall, SCAP/Tier1comm. MES is stock return given that the market return is below
its 5th percentile, measured for each individual company stock using the period October 2007–September 2008.
The sample consists of 18 U.S. financial firms included in the Federal Reserve’s stress tests of spring of 2009.

1 capital and tangible common equity, respectively, the t-statistics are 3.00
and 3.12 with adjusted R 2 s of 32.03% and 33.19%. When leverage is added,
the adjusted R 2 s either drop or are marginally larger. The (pseudo) R 2 s jump
considerably for the probit regressions, with the SCAP shortfall by Tier 1
capital regressions reaching 40.68% and, with leverage included, 53.22%. The
important point is that the systemic risk measures seem to capture quite well
the SCAP estimates of percentage expected losses in a crisis.
The above regressions use information up to March 2009 to coincide with
the timing of the Federal Reserve’s SCAP. As an additional analysis, the same
regressions are run in the right columns of Panels A and B using MES and
leverage measured prior to the failure of Lehman Brothers, that is, using
information from October 2007 to September 2008. While MES remains
statistically significant, the adjusted R 2 drops considerably for both measures
of capital and for both the OLS and probit regressions as expected.

3.2 The financial crisis: July 2007 to December 2008

We next consider how MES and leverage estimated using data from the year
prior to the crisis (June 2006 through June 2007) explain the cross-sectional

19
The Review of Financial Studies / v 30 n 1 2017

Table 2
OLS regression and probit regression analyses

Panel A: Dependent variable is SCAP Shortfall/Tier1

April 2008–March 2009 October 2007–September 2008

OLS Probit OLS Probit

(I) (II) (III) (IV) (V) (VI) (VII) (VIII) (IX) (X) (XI) (XII)
Intercept −17.29 3.14 −17.33 −5.44 −2.43 −6.04 −13.46 3.94 −14.19 −2.4 −0.95 −2.03
(−2.2) (1.16) (−2.00) (−2.72) (−2.26) (−2.24) (−1.50) (1.12) (−1.50) (−1.37) (−1.40) (−1.14)
MES 1.91 1.91 0.45 0.34 3 3.29 0.37 0.21
(3.00) (2.46) (2.72) (1.65) (2.19) (2.04) (1.40) (0.67)
0.09 −0.001 0.15 −0.09

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

LVG 0.10 0.09 0.08 0.06
(1.35) (−0.01) (2.16) (1.61) (0.66) (−0.37) (1.50) (1.05)
Adj. R2 32.03% 4.65% 27.5% 40.68% 45.09% 53.22% 18.27% −3.46% 13.61% 11.06% 15.17% 17.3%
No. obs 18 18 18 18 18 18 18 18 18 18 18 18

Panel B: Dependent variable is SCAP Shortfall/Tier1Comm

April 2008-March 2009 October 2007-September 2008
OLS OLS

(I) (II) (III) (VII) (VIII) (IX)

Intercept −36.24 4.41 −30.86 −25.72 9.02 27.13
(−2.25) (0.85) (−1.79) (−1.37) (1.24) (−1.37)
4.05 3.29 6.00 6.57
(3.12) (2.13) (2.09) (1.94)
0.27 0.12 0.31 −0.17
(2.20) (0.90) (0.64) (−0.34)

Adj. R2 33.19% 18.44% 33.17% 16.57% −3.56% 11.69%

No. obs 18 18 18 18 18 18
In Panel A the dependent variable is SCAP Shortfall/Tier1, and in Panel B it is SCAP Shortfall/Tier1Comm. Models
(I)–(III) are regression analyses based on MES and LVG computed respectively, during and at end of the period, April
2008–March 2009. Models (IV)–(VI) are the equivalent Probit regression results. In Panels A and B, Models (VII)–
(XII) repeat the analysis using the period October 2007–September 2008. t-stats are reported in brackets for the OLS
regression coefficient estimates. In the probit regressions the dependent variable is converted into a binary variable by
only considering non-zero or zero values. The reported R 2 is then the Pseudo R2 .

variation in equity performance during the crisis (July 2007 through December
2008). To put the explanatory power of MES and LVG in perspective, we also
check their incremental power relative to other measures of risk. For this, we
focus on (i) two measures of firm-level risk — the expected shortfall, ES (i.e.,
the negative of the firm’s average stock return in its own 5% left tail) and the
annualized standard deviation of returns based on daily stock returns, Vol, and
(ii) the standard measure of systemic risk, Beta, which is the covariance of a
firm’s stock returns with the market divided by variance of market returns. The
difference between our systemic risk measure and Beta arises from the fact that
systemic risk is based on tail dependence rather than average covariance. We
want to compare these ex ante risk measures to the realized SES, that is, the ex
post return of financial firms during the period July 2007–December 2008.
Table 3 describes the summary statistics of all these risk measures for the 102
financial firms in the U.S. financial sector with equity market capitalization as of
the end of June 2007 in excess of U.S.$5 [Link] B lists these firms and
their “type” based on two-digit SIC code classification (Depository Institutions,

20
Measuring Systemic Risk

Table 3
Summary statistics and correlation matrix of stock returns during the crisis, risk of financial firms, their
systemic risk and other firm characteristics

Panel A: Descriptive statistics of the measures Realized SES, ES, MES, Vol, Beta,
LVG, Log-Assets and ME
Realized SES ES MES Vol Beta LVG Log-AssetsME(blns)
Average – 47% 2.73% 1.63% 21% 1.00 5.25 10.84 31.25
Median – 46% 2.52% 1.47% 19% 0.89 4.54 10.88 15.85
Std. dev. 34% 0.92% 0.62% 8% 0.37 4.40 1.78 42.88
Min – 100% 1.27% 0.39% 10% 0.34 1.01 6.43 5.16
Max 36% 5.82% 3.36% 49% 2.10 25.62 14.61 253.70

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

Panel B: Sample correlation matrix of the measures Realized SES, ES, MES, Vol, Beta, LVG,
Log-Assets and ME
Realized SES 1.00
ES −0.17 1.00
MES −0.30 0.71 1.00
Vol −0.07 0.95 0.64 1.00
Beta −0.25 0.76 0.92 0.72 1.00
LVG −0.47 −0.09 0.24 −0.17 0.18 1.00
Log-Assets −0.38 −0.32 −0.07 −0.40 −0.07 0.75 1.00
ME −0.19 −0.24 −0.08 −0.25 −0.07 0.27 0.65 1.00
Panel C: Descriptive statistics of the average of the measures
Realized SES, ES, MES, Vol, Beta, LVG for different industry types
Depository institutions –42% 2.23% 1.42% 17% 0.87 6.21
Other: Non-depository —52% 3.35% 1.92% 26% 1.22 3.68
Insurance –44% 2.44% 1.28% 18% 0.78 4.44
Security dealers –59% 3.61% 2.68% 27% 1.61 9.58

This table contains overall descriptive statistics (Panel A) and sample correlation matrix (Panel B) for the
following measures: (i) Realized SES: the stock return during July 2007 to December 2008. (ii) ES: the Expected
Shortfall of an individual stock at the 5th percentile. (iii) MES is the marginal expected shortfall of a stock
given that the market return is below its 5th percentile. (iv) Vol is the annualized daily individual stock return
volatility. (v) Beta is the estimate of the coefficient in a regression of a firm’s stock return on that of the market’s.
(vi) Leverage (LVG) is measured as quasi-market value of assets divided by market value of equity, where
quasi-market value of assets is book value of assets minus book value of equity + market value of equity. (vii)
Log-Assets is the natural logarithm of total book assets. (viii) ME is the market value of equity. We used the
value-weighted market return as provided by CRSP. ES, MES, Vol, and Beta were measured for each individual
company’s stock using the period June 2006 to June 2007. LVG, log-assets, and ME are of end of June 2007.
The summary statistics are also shown in Panel C by different institution types as described in Appendix B.

Securities Dealers and Commodity Brokers, Insurance, and Others). The

realized SES in Panel A illustrates how stressful this period was for the financial
firms, with mean (median) return being −46% (−47%) and several firms losing
their entire equity market capitalization (Washington Mutual, Fannie Mae, and
Lehman Brothers). It is useful to compare ES and MES. While the average
return of a financial in its own left tail is −2.73%, it is −1.63% when the
market is in its left tail. Average volatility of a financial stock’s return is 21%
and average beta is 1.0. The power law application in Section 2 suggests that an
important component of systemic risk is LVG, the quasi-market assets to market
equity ratio. This measure is on average 5.26 (median of 4.59), but it has several
important outliers. The highest value of LVG is 25.62 (for Bear Stearns) and
the lowest is just 1.01 (for CBOT Holdings Inc.). All these measures, however,
exhibit substantial cross-sectional variability, which we attempt to explain later.

21
 The Review of Financial Studies / v 30 n 1 2017

Panel B shows that individual firm risk measures (ES and Vol) are highly
correlated, and so are dependence measures between firms and the market (MES
and Beta). Naturally, the realized returns during the crisis (realized SES) are
negatively correlated to the risk measures and, interestingly, realized SES is
most correlated with LVG, Log-Assets, and MES, in that order.
We also examine the behavior of risk and systemic risk across types of
institutions based on the nature of their business and capital structure. As
mentioned above, in Appendix B, we rely on four categories of institutions:
(i) Depository institutions (29 companies with two-digit SIC code of 60);

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

(ii) Miscellaneous non-depository institutions including real estate firms whom
we often refer to as “Other” (27 companies with codes of 61, 62 except 6211,
65, or 67); (iii) Insurance companies (36 companies with codes of 63 or 64);
and (iv) Security and Commodity Brokers (10 companies with four-digit SIC
code of 6211).18 These risk measures are reported in Panel C of Table 3.
When these risk measures are observed across institution type, there are
several interesting observations to be made. Depository institutions and
insurance firms have lower absolute levels of risk, measured both by ES and
Vol. These institutions also have lower dependence with the market, MES
and Beta. Financial leverage, that is, quasi-market assets to equity ratio, is,
however, higher for depository institutions than for insurance firms. When all
this is in theory combined into our estimate of the systemic risk measure,
in terms of realized SES, insurance firms are overall the least systemically
risky, next are depository institutions, and most systemically risky are the
securities dealers and brokers. Importantly, by any measure of risk, individual
or systemic, securities dealers and brokers are always the riskiest. In other
words, the systemic risk of these institutions is high not just because they are
riskier in an absolute risk sense, but they have greater tail dependence with the
market (MES) as well as the highest leverage (LVG). In particular, both their
MES and leverage are about twice the median of other financial firms.
Table 4 shows the power of MES and leverage in explaining the realized
performance of financial firms during the crisis, both in absolute terms as well
as relative to other measures of risk. In particular, it contains cross-sectional
regressions of realized returns during July 2007–December 2008 on the pre-
crisis measures of risk: ES, Vol, MES, Beta, LVG, and Log Assets. (As described
earlier, we also note that Appendix C provides the firm-level data on MES and
LVG.)
Figure 2 shows that MES does a reasonably good job of explaining the
realized returns, and naturally a higher MES is associated with a more negative
return during the crisis. A few cases illustrate the point well. We can see that

18 Note that Goldman Sachs has an SIC code of 6282, but we classify it as part of the Security and Commodity
Brokers group. Some of the critical members of “Other” category are American Express, Black Rock, various
exchanges, and Fannie Mae and Freddie Mac, the latter firms being of course significant candidates for
systemically risky institutions.

22
Measuring Systemic Risk

Table 4
Stock returns during the crisis, risk of financial firms, and their systemic risk

Panel A, OLS regression analysis: The dependent variable is Realized SES, the company stock returns
during the crisis
(1) (2) (3) (4) (5) (6) (7) (8)
ES −0.05
(−1.14)
Vol 0.04 −0.07
(0.07) (−0.12)
MES −0.21∗∗∗ −0.15∗∗ −0.17∗∗
(−2.90) (−2.25) (−2.08)

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

Beta −0.29∗∗
(−2.24)
LVG −0.04∗∗∗ −0.04∗∗∗ −0.03∗∗
(−5.73) (−5.43) (−2.29)
Log Assets −0.09∗∗∗ −0.05∗
(−4.86) (−1.69)
Industry dummies
Constant −0.32∗∗∗ −0.44∗∗∗ −0.13 −0.18 −0.18∗∗ 0.02 0.61∗∗∗ 0.50
(−2.71) (−3.81) (−1.09) (−1.42) (−2.50) (0.20) (2.75) (1.61)
Other −0.04 −0.09 0.01 0.012 −0.20∗∗ −0.12 −0.25∗∗∗ −0.15
(−0.33) (−0.91) (0.14) (0.12) (−2.44) (−1.35) (−2.87) (−1.61)
Insurance(×100) 0.43 −0.68 −3.63 −2.95 −8.86 −10.17 −0.09 −0.11
(0.05) (−0.08) (−0.45) (−0.36) (−1.19) (−1.39) (−1.13) (−1.55)
Broker-dealers −0.09 −0.16 0.11 0.06 −0.02 0.16 −0.17 0.14
(−0.65) (−1.20) (0.71) (0.3) (−0.18) (1.19) (−1.56) (1.02)
Adj. R2 0% −1.36% 6.72% 3.62% 24.27% 27.34% 18.46% 28.02%
No. obs. 102 102 102 102 101 101 101 101
(continued)

Bear Stearns, Lehman Brothers, CIT, and Merrill Lynch have relatively high
MES and these firms lose a large chunk of their equity market capitalization.
There are, however, also some reasons to be concerned. For example, exchanges
(NYX, ICE, ETFC) have relatively high MES, but we do not think of these as
systemic primarily because they are not as leveraged as, say, investment banks
are. Similarly, while AIG and Berkshire Hathaway have relatively low MES,
AIG’s leverage at 6.12 is above the mean leverage, whereas that of Berkshire is
much lower at 2.29. Thus, the two should be viewed differently from a systemic
risk standpoint. As described in the beginning of Section 3, combining MES and
leverage of financial firms helps explain systemic risk better since, as predicted
by the theory, financial distress costs of leveraged firms can be large in a crisis.
To understand this point, consider the estimated systemic risk ranking of
financial firms (i.e., Model 6 in Table 4, which coincides with the label “Fitted
Rank,” in Appendix C). In this light, when combining MES and LVG using
the estimated regression coefficients, exchanges are no longer as systemic as
investment banks and AIG looks far more systemic than Berkshire Hathaway.
The five investment banks rank in the top ten by both their MES and leverage
rankings, so they clearly appear systemically risky (Appendix C). Countrywide
is ranked 24th by MES given its MES of 2.09%, but due to its high leverage
of 10.39, it has a combined systemic risk ranking of 6th using the estimated

23
The Review of Financial Studies / v 30 n 1 2017

Table 4
Continued

Panel B, Tobit Analysis: The dependent variable is Realized SES, the company stock returns
during the crisis
ES −0.05
(−1.06)
Vol 0.10 −0.26
(0.17) (−0.42)
MES −0.23∗∗∗ −0.001∗∗ −0.001∗
(−2.85) (−2.03) (−1.69)
Beta −0.32∗∗
(−2.24)

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

LVG −0.07∗∗∗ −0.06∗∗∗ −0.05∗∗∗
(−6.40) (−6.14) (−3.18)
Log Assets −0.12∗∗∗ −0.04
(−5.48) (−1.18)
Industry dummies
Constant −0.35∗∗∗ −0.48∗∗∗ −0.14 −0.18 −0.06 0.12 0.87∗∗∗ 0.5
(−2.66) (−3.93) (−1.02) (−1.02) (−0.69) (1.01) (3.48) (1.48)
Other −0.01 −0.08 0.04 0.04 −0.26∗∗∗ −0.18∗ −0.28∗∗∗ −0.18∗
(−0.10) (0.70) (0.41) (0.40) (−2.92) (−1.82) (−2.90) (−1.82)
Insurance(×100) 0.03 0.01) −0.02 −0.01 −0.11 −0.12 −0.09 −0.13
(0.27) (0.14) (−0.21) (−0.14) (−1.42) (−1.58) (−1.03) (−1.60)
Broker-dealers −0.14 −0.22 0.08 0.03 −0.07 0.10 −0.23∗ 0.10
(−0.87) (−1.42) (0.49) (0.18) (−0.58) (0.68) (−1.85) (0.68)
Pseudo R2 3.95% 2.95% 10.21% 7.49% 43.95% 47.70% 28.87% 49.05%
No. obs. 102 102 102 102 101 101 101 101
This table contains the results of the cross-sectional regression analyses (Panel A) and Tobit analyses (Panel
B) of individual company stock returns (Realized SES) on risk (ES, Vol, LVG) and systemic risk (MES, Beta)
measures. Realized SES, risk measures, and leverage are as described in Table 3. In the tobit regression analyses
the following firms were assumed to have a Realized SES of –1: AIG, Bear Stearns, Citi-Group, Countrywide
Financial Corp., Freddie Mac, Fannie Mae, Lehman Brothers, Merrill Lynch, National City Corp., Washington
Mutual, and Wachovia. All balance sheet data are based on quarterly CRSP-Compustat merged data as of end of
June 2007. The industry-type dummies are employed for Other, Insurance, and Broker-Dealers as classified in
Appendix B.
t -statistics are given in parentheses. ***, **, and * indicate significance at 1%, 5%, and 10% levels, respectively.

coefficients from Model 6 in Table 4. Similarly, Freddie Mac is ranked 61st by

its MES, but given its high leverage of 21 (comparable to that of investment
banks), it ranks 2nd in terms of its combined ranking. On the flip side, CB
Richard Ellis, a real-estate firm, ranks 5th in MES, but given its low leverage
of 1.55, it ranks only 24th in terms of combined ranking. Investment banks,
Countrywide, and Freddie Mac all collapsed or nearly collapsed, whereas CB
Richard Ellis survived, highlighting the importance of the leverage correction
in systemic risk measurement.
In contrast to the statistically significant role of MES in explaining cross-
sectional returns, traditional risk measures—Beta, Vol, and ES—do not perform
that well. The R 2 with Beta is just 3.62%, and R 2 s with Vol and ES are 0.0%.
It is also interesting to note that, in the regressions that include LVG and MES
together, the institutional characteristics no longer show up as significant. This
suggests that the systemic risk measures do a fairly good job of capturing, for
example, the risk of broker dealers. Regarding the size of banks, we see that
the log of assets is significant when included alone in the regression (Model

24
 Measuring Systemic Risk

.5
HCBK
UB
Return during crisis: July07 to Dec08

SAF
AOC BOT
ATPBCT
0

CB AGE
CBH CG BER
AFL CBSS
BRK
BRK WFC
TRV BLK
MA UNPNTRS SCHW
USB
MMCFNF
NYB

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

BBT
PNC UNM
CINF FIS AMTD
BK
JPMWUTROW
TMK PGR NMX
SNV HUM
ALL AET
MTB STT
METLTRLUK SEIC ICE
UNHWLP AIZ
-.5

EVBEN
COFCME PFG
STIMIHBAN CMA GSAMP NYX
ZIONCNAPRU CI
BACAXP
KEY LNC JNS
CVH RF MS
HNT FITB LM
SLM HIG
C SOVCFC MER
WB CBG
GNW ACAS CIT
NCC MBI BSCETFC
AIG ABK
-1

FRE WM FNM LEH

0 .01 .02 .03 .04

MES5 measured June06 to June07

Figure 2
MES predicts realized equity returns during the crisis
MES estimated ex ante over the period June 2006–June 2007 plotted against the stock return during the crisis
July 2007 to December 2008. The sample consists of 102 U.S. financial firms with a market cap in excess of $5
billion as of June 2007.

(7)), and while its significance drops substantially once MES and leverage
are included, it remains borderline significant (Model (8)). The negative sign
on log of assets suggests that size may affect not only the dollar systemic
risk contribution of financial firms but also the percentage systemic risk
contribution as well. That is, large firms may create more systemic risk than a
likewise combination of smaller firms, according to this regression, though
the significance of this result is weak (and our theory does not have this
implication).19
As is clear from Table 3 and Figure 2, there are a number of firms for which
the realized stock return during the crisis period was −100%. This introduces
a potential truncation bias in the dependent variable and in turn will affect
the model’s estimated regression coefficients. To control for this bias, Panel
B of Table 4 runs a Tobit analysis where 11 firms (listed in the caption of
Table 4) that had returns worse than −90% are assumed to have in fact had
returns of −100%. In all likelihood, these firms would have all reached that

19 The R 2 s from Columns (1), (3), and (6) of Table 4, Panel A, imply that the explanatory power of leverage is
about four times as high as that of MES in explaining the realized SES .

25
The Review of Financial Studies / v 30 n 1 2017

.05
ICE
MES5 measured June06 to June07
.04

NYX
.03

AMTD

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

LUK LEH
CBG ETFC
GS BSC
.02

LM SCHW
MS
TROW MER
FNF FIS EV AGE
BEN
CBHCGCME BER AMP
SEIC FNM
AIG CVH ALL AXP
NCC
HIG
CBABK BOT JPMSTT
LTR JNS
ATMITRV
HBAN LNC
MTB
CI C NTRS
UNP MBI
.01

RF
UBCOFPRU
HUM
MET
CMAPFG
ZION CNA
AOC
STI
TMK BBT
WFCAET BK CFC CIT
BLK
UNH
NYBHNT
MMCSNV
USB
AFL
HCBK
CBSSWB
KEYBAC
UNMGNWWM
PGR
PBCT SAF
FITB
FRE SOV
WLP AIZ
SLM CINFMA
PNC ACAS

BRK
BRK
0

0 .01 .02 .03 .04

MES5 measured June05 to June06

Figure 3
Stability of MES
The graph depicts a scatter plot of the MES, marginal expected shortfall measure at the 5% level, computed
during the June 2006–June 2007 period versus that computed during June 2005–June 2006. MES is the marginal
expected shortfall of a stock given that the market return is below its 5th percentile

outcome but were bailed out in advance, as with Fannie Mae, Freddie Mac, AIG,
and Citigroup, or were merged through government support, as in the case of
Bear Stearns. Our results are qualitatively unaffected though the coefficient
on leverage increases almost twofold, which is unsurprising given the high
leverage of the firms that ran aground in the crisis.
We consider several robustness checks. Figure 3 graphs a scatter plot of
the MES computed during June 2006–June 2007 versus that computed during
June 2005–June 2006. Even though there is no overlap between the return
series, the plot generally shows a fair amount of stability from year to year with
this particular systemic risk measure. Wide time-series variation in relative
MES would make the optimal policy more difficult to implement. It is of
interest therefore to examine how early MES and LVG predict the cross-
section of realized returns during the crisis. We compute MES and SES over
several periods other than the June 2006-June 2007 estimation period: June
2006–May 2007, May 2006–April 2007, April 2006–March 2007, and March
2006–February 2007. In each period, we use the entire data of daily stock
returns on financial firms and the market, and the last available data on book

26
 Measuring Systemic Risk

assets and equity to calculate the quasi-market measure of the assets to equity
ratio. Once the measures are calculated for each of these periods, the exercise
involves explaining the same realized returns during the crisis period of July
2007 to December 2008.
Panel A of Table 5 shows that the predictive power of MES progressively
declines as we use lagged data for computing the measure. The overall
predictive power, however, remains high as leverage has certain persistent,
cross-sectional characteristics across financial firms. The coefficients on LVG
remain unchanged throughout these periods. To better understand the MES

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

decline, we repeat the Panel A regressions using two alternative measures of
MES: (i) W-MES, a weighted MES, which uses exponentially declining weights
(λ = 0.94 following the Risk Metrics parameter) on past observations to estimate
the average equity returns on the 5% worst days of the market, and (ii) D-MES,
a dynamic approach to estimating MES, which uses a dynamic conditional
correlation (DCC) model with fat idiosyncratic tails.20 Panels B and C provide
the results for W-MES and D-MES, respectively. The adjusted R 2 s are generally
higher, and the alternative measures of MES better hold their predictive power
even with lagged measurement. For example, the coefficients remain strongly
significant using the April 2006–Mar 2007 data. These results suggest there is
some value to exploring more sophisticated methods for estimating MES and
to including the most recent data in estimates.
Finally, Panel D of Table 5 considers F-MES, which is calculated as our
benchmark MES, but instead of using the CRSP value-weighted index return
as the “market return,” we instead use the financial industry return series
obtained from the data on 30 industry portfolios provided by Kenneth French.
The financial industry return maps closer to our economic model of systemic
risk because the externality arises when the financial sector experiences
undercapitalization rather than the market as a whole. Also, F-MES might
capture better tail dependence induced between a financial firm and other
financial firmns due to contagion-based systemic risk. We find that the results
using F-MES are virtually identical to the benchmark results in Panel A,
implying little difference in using stock market or financial sector as the relevant
market for computing MES.

3.3 Using CDS to measure systemic risk

We have seen the ability of the MES and leverage of financial firms to forecast
the outcome of the stress test and the equity performance during the financial
crisis. We add to this evidence by considering the credit default swaps (CDS)
data from Bloomberg for these financial firms.21 On the one hand, CDS data

20 We are grateful to Christian Brownlees and Robert Engle of New York University Stern School of Business for
sharing with us their dynamic measures of MES for our sample firms, using the methodology they develop in
Brownlees and Engle (Forthcoming).
21 Our results are robust to the sample of firms for which data are available from Markit, and the sample of
overlapping firms between Bloomberg and Markit.

27
The Review of Financial Studies / v 30 n 1 2017

Table 5
Stock returns during the crisis and systemic risk measured with different leads

Panel A (MES): The dependent variable is Realized SES, the company stock returns during the crisis
June 2006– May 2006– April 2006– March 2006–
May 2007 April 2007 March 2007 February 2007
Intercept −0.14∗ −0.20∗∗ −0.20∗∗ −0.23∗∗∗
(−1.75) (−2.42) (−2.48) (−3.09)
MES −0.10∗∗ −0.05 −0.05 −0.04
(−2.30) (−1.26) (−1.24) (−0.98)
LVG −0.04∗∗∗ −0.04∗∗∗ −0.04∗∗∗ −0.04∗∗∗
(−5.06) (−5.09) (−5.21) (−5.20)

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

Adj. R2 24.87% 21.84% 22.61% 21.00%
No. obs. 101 101 101 101
Panel B (W-MES): The dependent variable is Realized SES, the company stock returns during the crisis
Intercept −0.21∗∗∗ −0.09 −0.09 −0.18∗
(−3.22) (−1.11) (−1.15) (−1.96)
W-MES −0.07∗ −0.10∗∗∗ −0.10∗∗∗ −0.03
(−1.73) (−2.96) (−2.94) (−1.30)
LVG −0.04∗∗∗ −0.03∗∗∗ −0.03∗∗∗ −0.04∗∗∗
(−5.01) (−4.49) (−4.61) (−5.25)
Adj. R2 23.15% 27.11% 27.76% 21.97%
No. obs. 101 101 101 101
Panel C (D-MES): The dependent variable is Realized SES, the company stock returns during the crisis
Intercept −0.12 −0.06 −0.11 −0.18∗
(−1.40) (−0.66) (−1.24) (−2.27)
D-MES −0.12∗ −0.13∗∗ −0.12∗ −0.08
(−2.23) (−2.86) (−2.36) (−1.92)
LVG −0.03∗∗ −0.03∗∗ −0.03∗∗ −0.03∗∗
(−5.25) (−4.82) (−4.13) (−5.02)
Adj. R2 24.14% 26.44% 24.58% 23.15%
No. obs. 101 101 101 101
Panel D (F-MES): The dependent variable is Realized SES, the company stock returns during the crisis
Intercept −0.15∗ −0.19∗∗ −0.19∗∗ −0.22∗∗∗
(−1.84) (−2.30) (−2.35) (−2.82)
F-MES −0.09∗ −0.06 −0.06 −0.04
(−1.82) (−1.43) (−1.41) (−0.94)
LVG −0.04∗∗∗ −0.04∗∗∗ −0.04∗∗∗ −0.04∗∗∗
(−4.80) (−5.03) (−5.15) (−5.18)
Adj. R2 23.24% 22.19% 22.96% 21.33%
No. obs. 101 101 101 101
This table contains the results of the cross-sectional regression analyses of individual company stock returns
(Realized SES) on systemic risk: MES (Panel A), W-MES (Panel B), D-MES (Panel C), and F-MES (Panel D)
measure. All measures are as described in Table 3 and Table 4, except for W-MES, which is the exponentially
weighted MES; D-MES, which is the dynamic MES; and F-MES, which is MES computed using the return on
the financial industry* as the market portfolio. All three variants of MES are measured over different pre-crisis
periods as indicated below. The stock return during the crisis is always measured during July 2007 to December
2008. Leverage is based on data available at the end of each period. Hence, for Columns 1 through 3 we use
2007Q1 data, and for the last column we use 2006Q4 balance sheet data.
* The financial industry return series are obtained from the 30 industry portfolios available on Kenneth French’s
website, [Link] [Link].
t –statistics are given in parentheses. ***, **, and * indicate significance at 1%, 5%, and 10% levels, respectively.

might be preferred to equity data because CDS might better capture estimates
of losses of the market value of the financial firm’s assets, as opposed to just
its equity. On the other hand, CDS data reflects the underlying value of the

28
 Measuring Systemic Risk

financial firm’s debt, which may be subject to government guarantees. Of the

102 financial firms we have looked at so far, 40 of them have enough unsecured
long-term debt to warrant the existence of CDS in the credit derivatives market.
Appendix D provides a list of the 40 firms and their type of institution.
A question arises how to operationalize the CDS data for calculating MES.
The CDS premium resembles the spread between risky and riskless floating rate
debt. We denote this spread by s. To garner some intuition as to how changes
in the spread are related to MES, note that if P is the bond price, V the value
of the firm’s assets, ξ is the elasticity of the bond price to firm value, and D

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

is the bond’s duration, then dP/P = −Dds and dP/P = ξ dV /V . Combining the
two relationships, we obtain that ds = −(ξ/D)(dV /V ). Ignoring the duration
term changes across firms/days means that measuring the firm’s losses, that
is, dV /V , using the spread change ds is proportional to its bond elasticity ξ .
Since we know that ξ is approximately 0 when the bond is close to risk-free
and approximately one when the bond is virtually in default, ds attaches close
to zero weight to the firm value return dV /V for safe firms (when leverage is
very low) and high weight (equal to 1/D) to dV /V for very risky firms (when
leverage is very high). Therefore, firm value changes can be approximated
better than using the arithmetic change in spread ds by using instead the log
change, ds/s = −(ξ/(Ds))(dV /V ), where s is small when ξ is close to zero and
large when ξ is close to one. Further, from an econometric standpoint, the log
change is more stationary and less driven by outliers. Therefore, when using
CDS data, we empirically estimate MES at a standard risk level of 5% using
daily data of CDS returns, ds/s.
Thus, practically, we take the 5% worst (i.e., highest CDS return) days for
an equally weighted portfolio of CDS returns on the 40 financial firms from
June 2006 to July 2007, and we then compute the CDS return for any given
firm for these days. This measure is used against realized SES during the crisis.
For comparison purposes, we also show results that use arithmetic changes in
the CDS spread as a measure of CDS return.
As a first pass at the data, Appendix D presents stylized facts about the
financial firms’ MES based on the CDS market, including ranking, MES%,
and realized CDS spread returns during the crisis period.22 Consider the top
three financial institutions in terms of highest CDS MES in each institutional
category:
• The three insurance companies are Genworth Financial (16.40%),Ambac
Financial (8.05%), and MBIA (6.71%). All of these companies were
heavily involved in providing financial guarantees for structured products
in the credit derivatives area.

22 We can compare the MES CDS ranking of Appendix D to the MES equity ranking of Appendix C for the 40
coincident firms. The rank correlation is 23%, which suggests there is different potential information in CDS and
equity markets. This could be possibly due to CDS MES better capturing asset losses from a positive viewpoint
or being a biased measure due to government guarantees from a negative point of view.

29
 The Review of Financial Studies / v 30 n 1 2017

• The top three depository institutions are Wachovia (7.21%), Citigroup

(6.80%), and Washington Mutual (6.15%). These institutions are
generally considered to ex post have been most exposed to the nonprime
mortgage area, with two of them, Wachovia and Washington Mutual,
actually failing.
• The top three broker dealers are Merrill Lynch (6.3%), Lehman Brothers
(5.44%), and Morgan Stanley (4.86%). Two of these three institutions
effectively failed.23
• The top three others, SLM Corp (6.82%), CIT Group (6.80%), and Fannie

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

Mae (5.70%), also ran into trouble due to their exposure to credit markets,
with CIT going bankrupt and Fannie Mae being put into conservatorship.
The above anecdotal observations illustrate the potential for using MES based
on CDS data. That said, as mentioned previously, note that CDS in a pre-crisis
period may not relate well to the realized losses of financial firms during a
crisis if some firms receive greater government guarantees, for example, deposit
institutions, the government-sponsored enterprises, and the so-called too-big-
to-fail firms.24 To address this issue in part, we analyze the ability of CDS MES
to forecast systemic risk in both the July 2007 to December 2008, and the July
2007 to June 2008 periods (i.e., prior to many government guarantees being
made explicit). To further address this issue, we also investigate the ability
of CDS MES to forecast not only future CDS returns, but also equity returns
(which accrue the benefit of government guarantees to a much lesser extent
than creditor returns).
In terms of a more thorough analysis using CDS, Figure 4 shows that there is
a positive relation between the ex ante measure of systemic risk based on CDS
and the ex post CDS return during the crisis (as measured from July 2007 to June
2008). To test this relation more rigorously, Table 6 reports regressions in which
the regressors are, respectively, CDS MES based on CDS returns (Panel A)
and CDS spread changes (Panel B). The dependent variables are, respectively,
the realized CDS returns and changes during different periods covering the
crisis related to government action on creditor guarantees (July 2007-June
2008/September 14, 2008/September 30, 2008/October 10, 2008/December 30,
2008). Table 6 demonstrates that our ex ante measure significantly predicts the
realized systemic risk. First, putting aside the date of TARP capital assistance
in October, the R 2 s are between 17.86% to 19.94%. Second, in terms of CDS
MES versus leverage, CDS MES is generally the more significant variable.
Because CDS reflects the claim on the underlying debt, this is consistent with

23 We note here that if Bear Stearns CDS return were measured until the point of its arranged merger with JPMorgan
in mid-March 2008, its realized CDS return would be higher than having measured it until dates thereafter.
24 Equity also suffers from this problem to the extent government guarantees delay bankruptcy preferentially for
some financial firms, extending the option of their equity to continue relative to the option for some other firms.
It is more likely a second-order effect, however, compared to the pricing of the underlying debt and CDS of
financial firms in distress.

30
 Measuring Systemic Risk

ABK

4
measured during 1 July 07- 30 June 08 MBI
Total realised return in CDS spread

AIG
WM WB
PRU
LNC AXP C
CIT
MET HIG WFC
FREBAC
2

LEHMER
FNM
ALL
CB JPM MS
SCHW GNW

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

L GS
CIAETSAF
UNMSTA
CNA
1

UNP
BSC
SLM
JNS
MMC AOC
TMK
0 -1

AT

0 .05 .1 .15 .2
CDS MES

Figure 4
MES predicts realized CDS returns during the crisis
MES estimated ex ante from CDS returns July 2006–June 30, 2007, plotted against the total realized return on
CDS spread during July 1, 2007–June 30, 2008.

CDS MES capturing more of the tail behavior and thus being less reliant on the
leverage arguments provided in Section 2. Third, there are substantive drops in
explanatory power when CDS spread changes are used instead of CDS returns
(Panel B). This is consistent with the aforementioned argument on the need to
be careful with respect to operationalizing CDS MES.25
As final evidence, Table 7 shows how CDS MES based on CDS returns
(Panel A) or CDS spread changes (Panel B) predicts the realized equity returns
during the same periods as Table 6. The results are quite strong, with both CDS
MES and leverage coming in at very high significant levels with adjusted R 2 s of
50% or higher using CDS returns (and 30% plus using CDS spread changes).
The important point is that the ex ante systemic risk measures (i.e., prior to
the crisis) have information for which firms might run into trouble. Therefore,
by inference, these are the firms that should, according to our derived optimal
policy, be taxed in order to induce them to reduce their systemic risk.
In summary, these results are also strongly supportive of the ability of CDS
MES to forecast future changes in firm value during a financial crisis, whether
estimated by CDS or equity returns. While CDS MES may have been useful

25 Note that unlike in Table 4, leverage is statistically insignificant in explaining realized SES in the presence of
CDS MES.

31
 The Review of Financial Studies / v 30 n 1 2017

Table 6
CDS MES vs. realized CDS SES

Panel A: The dependent variable is total realized return on CDS spread during the crisis; CDS MES is
measured as log returns
July 1, 2007– July 1, 2007– July 1, 2007– July 1, 2007– July 1, 2007–
June 30, September 14, September 30, October 10, December 30,
2008 2008 2008 2008 2008
CDS MES 10.21∗∗ 9.67∗ 13.11∗∗ 10.72 11.56∗
(2.06) (1.83) (2.15) (1.65) (2.02)
LVG 0.05 0.05 0.05 0.06 0.03
(1.43) (1.41) (1.33) (1.45) (0.81)
1.34∗∗ 1.75∗∗ 1.80∗∗∗ 1.90∗∗∗ 1.71∗∗∗

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

Constant
(2.68) (3.28) (2.93) (2.91) (2.96)
Other −0.95∗ −1.29∗∗ −1.22∗ −0.97 −1.09∗
(−1.93) (−2.46) (−2.02) (−1.52) (−1.92)
Insurance −0.14 −0.48 −0.44 −0.03 0.35
(−0.32) (−1.01) (−0.81) (−0.04) (0.68)
Broker dealers −0.87 −0.91 −0.72 −0.80 −0.63
(−1.52) (−1.49) (−1.02) (−1.07) (−0.96)
Adj. R2 17.86% 19.94% 19.37% 10.80% 19.30%
No. obs. 40 40 40 40 40
Panel B: The dependent variable is total change in CDS spread during the crisis; CDS MES is
measured as changes in CDS spreads
CDS MES 90.41∗∗ 91.04∗∗ 201.35∗∗∗ 239.08∗∗ 228.27∗∗
(2.63) (2.16) (2.82) (3.12) (2.70)
LVG −2.07 5.80 12.24 25.50 23.76
(−0.20) (0.45) (0.56) (1.09) (0.92)
Constant 46.51 236.00 433.10 289.63 240.62
(0.30) (1.24) (1.35) (0.84) (0.63)
Other −131.56 −387.37∗ −693.51∗ −573.43 −738.60∗
(−0.78) (−1.87) (−1.98) (−1.52) (−1.78)
Insurance 104.02 −52.03 −233.95 4.30 77.11
(0.72) (−0.29) (−0.78) (0.01) (0.22)
Broker dealers −25.49 −183.60 −435.61 −489.86 −606.80
(−0.14) (−0.80) (−1.11) (−1.17) (−1.31)
Adj. R2 7.21% 5.13% 11.67% 14.09% 12.45%
No. obs. 40 40 40 40 40
This table contains the results of the cross-sectional regression analyses of 40 companies’ realized CDS SES on
CDS MES. Panel A provides the results where CDS MES and realized CDS SES are measured in log return.
Panel B provides the results where CDS MES and realized CDS SES are measured using arithmetic changes in
CDS spreads. All measures are as described in Table 3 and Table 4, except for CDS MES, which is the average
CDS returns on the worst 5% days during July 1, 2006–June 30, 2007, where the average return on CDS spreads
of the 40 companies are the highest. Leverage is based on data available at end of each period. All CDS data are
from Bloomberg.
t -statistics are given in parentheses. ***, **, and *, indicate significance at 1%, 5%, and 10% levels, respectively.

prior to the start of the crisis, it is an open question whether this will continue
in the future with all the government guarantees in place.

4. Discussion
Before we conclude, it is useful to compare our optimal policy of Section 1.4
for regulating systemic risk to some of the proposals put forward by regulators
and policymakers. We then end with a discussion of important features not
analyzed in this paper.

32
Measuring Systemic Risk

Table 7
CDS MES vs. realized stock SES

Panel A: The dependent variable is realized stock return during the crisis;
CDS MES is measured as log returns
July 1, 2007– July 1, 2007– July 1, 2007– July 1, 2007– July 1, 2007–
June 30, September 14, September 30, October 10, December 30,
2008 2008 2008 2008 2008
CDS MES −4.38∗∗∗ −5.20∗∗∗ −6.05∗∗∗ −4.48∗∗∗ −4.11∗∗∗
(−3.33) (−3.52) (−3.83) (−3.19) (−2.77)
LVG −0.03∗∗∗ −0.04∗∗∗ −0.04∗∗∗ −0.04∗∗∗ −0.03
(−3.82) (−4.31) (−4.13) (−4.17) (−3.64)
−0.03 −0.007 −0.14

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

Constant 0.19 0.25
(−0.26) (1.29) (1.57) (−0.05) (−0.91)
Other 0.09 −0.11 −0.16 −0.13 −0.09
(0.69) (−0.76) (−0.99) (−0.90) (−0.62)
Insurance 0.03 −0.08 −0.17 −0.19 −0.06
(0.24) (−0.62) (−1.19) (−1.53) (−0.44)
Broker dealers 0.19 0.07 0.03 0.03 0.07
(1.26) (0.43) (0.19) (0.21) (0.39)
Adj. R2 46.79% 51.66% 50.94% 45.52% 40.76%
No. obs. 40 40 40 40 40
Panel B: The dependent variable is realized stock return during the crisis;
CDS MES is measured as changes in CDS spreads
CDS MES −0.06∗∗ −0.07∗ −0.07∗ −0.04 −0.02
(−2.04) (−2.00) (−2.02) (−1.21) (−0.71)
LVG −0.04 −0.05∗∗∗ −0.05∗∗∗ −0.05∗∗∗ −0.04∗∗∗
(−4.48) (−4.90) (−4.70) (−4.60) (−4.04)
Constant −0.17 0.03 0.06 −0.17 −0.30∗
(−1.26) (0.19) (0.35) (−1.16) (−1.98)
Other 0.20 0.02 −0.006 −0.03 −0.02
(1.42) (0.12) (−0.03) (−0.21) (−0.11)
Insurance 0.12 0.03 −0.04 −0.09 0.04
(0.96) (0.19) (−0.26) (−0.67) (0.28)
Broker dealers 0.33∗∗ 0.24 0.23 0.17 0.18
(2.06) (1.29) (1.12) (0.95) (1.00)
Adj. R2 37.16% 40.98% 37.31% 32.15% 28.49%
No. obs. 40 40 40 40 40
This table contains the results of the cross-sectional regression analyses of 40 companies’ realized stock returns
(Realized SES) on CDS MES (measured as log returns in Panel A and changes in CDS spreads in Panel B). All
measures are as described in Table 3 and Table 4, except for CDS MES, which is the average CDS returns on
the worst 5% days during July 1, 2006–June 30, 2007, where the average changes in CDS spreads of the 40
companies are the highest. Leverage is based on data available at end of each period. All CDS data are from
Bloomberg.
t -statistics are given in parentheses. ***, **, and *, indicate significance at 1%, 5%, and 10% levels, respectively.

There is much discussion among regulators, policymakers, and academics

of the need for a resolution fund that could be used to bail out large, complex
financial institutions. This fund, paid for by the institutions themselves, would
be akin to the Federal Deposit Insurance Corporation (FDIC). This resolution
fund is essentially the institution-risk component of the above tax and reflects
the costs of the government guarantees in the system (e.g., deposit insurance and
too-big-to-fail). It does not, however, fully address the systemic-risk component
since it does not differentiate between different macro-economic states and does
not recognize that the costs associated with the failure of a particular firm are
significantly higher in a crisis.

33
 The Review of Financial Studies / v 30 n 1 2017

Another important topic in the discussion of systemic risk has been the
size of financial institutions’ assets and/or liabilities. The theory described
in Section 1.4 gives some support for this approach. Almost trivially, ceteris
paribus, the expected losses of a financial firm conditional on a crisis are tied
one-for-one to the size of the firm’s assets.26 Of course, even though a firm that
doubles its size would pay, to a first approximation, twice the systemic tax, the
firm would also have twice the cash flow to cover the tax. Therefore, from an
economic point of view, the interesting question is what variables help explain
the percentage of expected losses (as opposed to losses in dollars).

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

Our theory says that the regulation of systemic risk should be based on each
firm’s SES and the overall probability of a systemic event Pr (W1 < zA). We
focused our empirical analysis on the cross-section of systemic risk, SES. The
risk of a systemic event Pr (W1 < zA) can be measured using historical research,
as in Reinhart and Rogoff (2008), who show that there are consistent leading
indicators of banking crises, for example, an asset price bubble, a corresponding
credit boom, and large capital inflows into the economy (see also Laeven and
Valencia 2013). The conditional risk of a systemic event can then be inferred
from dynamic long-run volatility models (Engle 2009; Brownlees and Engle
Forthcoming).
Finally, there have been recent discussions as to whether non-banking
institutions such as insurance companies and asset managers can be
systemically important. In our model, we did not introduce specific features
distinguishing banks from such non-banking entities. It is best to think of
our model as being most applicable to banks because we treat all liabilities
as short-term, whereby there is no distinction in the model between liquidity
and solvency risks. Acharya, Philippon, and Richardson (Forthcoming) extend
our model where some of the financial firm liabilities are long-term, allowing
for a distinction between liquidity and solvency risks. Their extended model
decomposes the negative externality of the financial sector’s distress into two
components: (i) a fire-sale externality that arises due to asset liquidations in
response to runs on short-term liabilities; and (ii) a going-concern externality
that is a disruption of new intermediation activities and which arises from
solvency risk. While banks likely contribute to systemic risk via both of these
externalities, some non-banking entities such as insurance companies that have
less short-term debt are likely to contribute to systemic risk primarily through
the going-concern externality.

26 In fact, Appendix C of the paper provides the contribution of each firm’s average dollar loss in market
capitalization during the worst 5% of market return days as a percentage of the average dollar loss across all of
the 102 largest financial firms (i.e., firms with over $5 billion of market equity). The top 6 in terms of contribution
(Citigroup (8.81%), JPMorgan (6.70%), Bank of America (6.87%), Morgan Stanley (4.39%), Goldman Sachs
(4.48%) and Merrill Lynch (4.06%)) are also in the top 7 in terms of total number of assets.

34
 Measuring Systemic Risk

5. Conclusion
Current financial regulations seek to limit each institution’s risk. Unless the
external costs of systemic risk are internalized by each financial institution,
the institution will have the incentive to take risks that are borne by all. An
illustration is the current crisis in which financial institutions had levered up
on similar large portfolios of securities and loans that faced little idiosyncratic
risk, but large amounts of systematic risk.
In this paper, we argue that financial regulation be focused on limiting
systemic risk, that is, the risk of a crisis in the financial sector and its spillover

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

to the economy at large. We provide a simple and intuitive way to measure each
bank’s contribution to systemic risk, suggesting ways to limit it. In a variety of
tests (e.g., stress test outcomes of 2009 and firm performance during the crisis of
2007-08) using market data from equity and CDS, our systemic risk measures
appear to be able to predict the financial firms with the worst contributions in
systemic crises.
Several extensions of our work are worthy of pursuit in future. While we
estimated and tested our proposed systemic risk measure using equity and
CDS data, another way to obtain such information is through prices of out-of-
the-money equity options and insurance contracts against losses of individual
firms when the system as a whole is in stress.27 While such insurances are not
yet traded, data on firm equity options as well as market options is available
and can be used to construct measures of tail dependence such as MES.
Finally, we investigated the role of leverage (measured as the ratio of assets to
common equity) in determining the systemic risk of firms. The form of leverage
that had the most pernicious effect in the crisis of 2007-08 was arguably short-
term debt, such as the overnight secured borrowing (“repo”) against risky assets
employed heavily by the investment banks (Adrian and Shin 2010), and the
short-term (overnight to week maturity) asset-backed commercial paper issued
by conduits that were backed by commercial banks (Acharya, Schnabl, and
Suarez 2013). In contrast, even though deposits are in principle demandable and
thus short-term too, the presence of deposit insurance meant that commercial
banks with access to insured deposits were in fact relatively stable in the crisis.
It seems important therefore to empirically understand how short-term leverage
contributes to market-based measures of systemic risk of financial firms.

27 Based on the theory presented here, Acharya et al. (2009) propose regulation of systemic risk based on mandatory
purchase of such insurance contracts by financial firms, partly from private sources (insurance companies), and
the rest from a systemic risk regulator.

35
The Review of Financial Studies / v 30 n 1 2017

Appendix A
Proof of Proposition 1
Using the definition of τ i in Equation (14), the bank’s problem is
 
 
max × w̄0i −w0i −τ0 +E u 1wi >0 ×w1i
w0i ,bi , xji 1
j

−α i g ×P r(w1i < 0)ES i −e ×P r(W1 < zA)SES i ,

and using Equations (12) and (13), this becomes
 
 

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

max c × w̄0i −w0i −τ0 +E u 1wi >0 ×w1i
w0i ,bi , xji 1
j
 
+E α i g1wi <0 w1i +e1[W1 <zA] (za i −w1i ) .
1

The set of programs for i = 1,...,N is equivalent to the planner’s program and the budget constraint
can be adjusted with τ0 . 
i J
Proof of Proposition 2 Equity value satisfies: w1 −w0 = j =1 rj xj −φ i −f i −w0 . This allows
i i i i

us to write  
J
xji  i E φ i | I5% f i −bi
MES5% i
= E −r j | I 5% + +
wi
j =1 0
w0i w0i
 
ζ
In expectations we have E −rji | I5% = βi,j ζ −1 % and therefore E −r i | W < zA =
ε̄m j 1

kE −rji | I5% . Using the definition of SES we can write:
 
za i  xj  i
J i
SES i za i w1i
1+ = i −E −1 | W 1 < zA = + E −rj | W1 < zA
w0 w0 w0i w0i j =1 w0i
 
E φ i | W1 < zA f i −bi
+ +
w0i w0i
Hence, under the power law assumption:
   
SES i za i E φ i | W1 < zA −k ×E φ i | I5% f i −bi
1+ −k ×MES i = i + +(1−k) .
w0 w0 w0 w0i


Appendix B
This appendix contains the names of the U.S. financial institutions used in the analysis of the recent
crisis. The institutions have been selected according to their inclusion in the U.S. financial sector
and their market cap as of end of June 2007 where all firms had a market cap in excess of U.S.$5
billion.
The companies can be categorized into the following four groups: Depositories (JPMorgan,
Citigroup, WAMU, ...), Broker-Dealers (Goldman Sachs, Morgan Stanley, etc.), Insurance (AIG,
Berkshire Hathaway, Countrywide, etc.) and Insurance Agents, Brokers, Service (Metlife, Hartford
Financial, etc.) and a group called Other consisting of non-depository institutions, real estate, and
so on.
The total number of firms in the sample is 102.
Note that although Goldman Sachs has a SIC code of 6282, thus initially making it part of the
group called Others, we have nonetheless chosen to put in the group of Broker-Dealers.

36
Measuring Systemic Risk

Table A.1

Depositories: 29 Other: Non-depository Insurance: 36 Broker-Dealers: 10

companies, 2-digit institutions etc.: 27 companies, 2-digit companies, 4-digit
SIC code = 60 companies, 2-digit SIC SIC code = 63 and 64 SIC code = 6211
code = 61, 62 (except
6211), 65, 67
1. B B & T CORP 1. ALLTEL CORP 1. A F L A C INC 1. BEAR STEARNS
COMPANIES INC
2. BANK NEW 2. AMERICAN CAPITAL 2. AETNA INC NEW 2. E TRADE

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

YORK INC STRATEGIES LTD FINANCIAL CORP
3. BANK OF 3. AMERICAN 3. ALLSTATE CORP 3. EDWARDS A G
AMERICA CORP EXPRESS CO INC
4. CITIGROUP INC 4. AMERIPRISE 4. AMBAC 4. GOLDMAN
FINANCIAL INC FINANCIAL GROUP SACHS GROUP
INC AMERICAN INC
5. COMERICA INC 5. BLACKROCK INC 5. INTERNATIONAL 5. LEHMAN
GROUP INC BROTHERS
HOLDINGS INC
6. COMMERCE 6. C B O T HOLDINGS 6. AON CORP 6. MERRILL LYNCH
BANCORP INC NJ INC ASSURANT INC & CO INC
7. HUDSON CITY 7. C B RICHARD ELLIS 7. BERKLEY W R 7. MORGAN
BANCORP INC GROUP INC CORP STANLEY DEAN
WITTER & CO
8. HUNTINGTON 8. C I T GROUP INC 8. BERKSHIRE 8. NYMEX
BANCSHARES NEW HATHAWAY INC HOLDINGS INC
INC DEL
9. JPMORGAN 9. CAPITAL ONE 9. BERKSHIRE 9. SCHWAB
CHASE & CO FINANCIAL CORP HATHAWAY INC CHARLES CORP
DEL NEW
10. KEYCORP NEW 10. CHICAGO 10. C I G N A CORP 10. T ROWE PRICE
MERCANTILE EXCH GROUP INC
HLDG INC
11. M & T BANK 11. COMPASS 11. C N A FINANCIAL
CORP BANCSHARES INC CORP
12. MARSHALL & 12. EATON VANCE 12. CHUBB CORP
ILSLEY CORP CORP
13. NATIONAL CITY 13. FEDERAL HOME 13. CINCINNATI
CORP LOAN MORTGAGE FINANCIAL CORP
CORP
14. NEW YORK 14. FEDERAL 14. COUNTRYWIDE
COMMUNITY NATIONAL FINANCIAL CORP
BANCORP INC MORTGAGE ASSN
15. NORTHERN 15. FIDELITY 15. COVENTRY
TRUST CORP NATIONAL INFO HEALTH CARE INC
SVCS INC
16. P N C 16. FIFTH THIRD 16. FIDELITY
FINANCIAL BANCORP NATIONAL FINL
SERVICES GRP INC NEW
INC
17. PEOPLES 17. FRANKLIN 17. GENWORTH
UNITED RESOURCES INC FINANCIAL INC
FINANCIAL INC
18. REGIONS 18. INTERCONTINEN- 18. HARTFORD
FINANCIAL CORP TALEXCHANGE FINANCIAL
NEW INC
19. SOVEREIGN 19. JANUS CAP GROUP 19. SVCS GROUP IN
BANCORP INC INC
(continued)

37
The Review of Financial Studies / v 30 n 1 2017

Table A.1
Continued

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

Depositories: 29 Other: Non-depository Insurance: 36 Broker-Dealers: 10
companies, 2-digit institutions etc.: 27 companies, 2-digit companies, 4-digit
SIC code = 60 companies, 2-digit SIC SIC code = 63 and 64 SIC code = 6211
code = 61, 62 (except
6211), 65, 67
20. STATE STREET 20. LEGG MASON INC 20. HEALTH NET INC
CORP
21. SUNTRUST 21. LEUCADIA 21. HUMANA INC
BANKS INC NATIONAL CORP
22. SYNOVUS 22. MASTERCARD INC 22. LINCOLN
FINANCIAL CORP NATIONAL CORP
IN
23. U S BANCORP 23. N Y S E EURONEXT 23. LOEWS CORP
DEL
24. UNIONBANCAL 24. S E I INVESTMENTS 24. LOEWS CORP
CORP COMPANY
25. WACHOVIA 25. S L M CORP 25. M B I A INC
CORP 2ND NEW
26. WASHINGTON 26. T D AMERITRADE 26. MARSH &
MUTUAL INC HOLDING CORP MCLENNAN COS
INC
27. WELLS FARGO 27. UNION PACIFIC 27. METLIFE INC
& CO NEW CORP
28. WESTERN 28. PRINCIPAL
UNION CO FINANCIAL GROUP
INC
29. ZIONS 29. PROGRESSIVE
BANCORP CORP OH
30. PRUDENTIAL
FINANCIAL INC
31. SAFECO CORP
32. TORCHMARK
CORP
33. TRAVELERS
COMPANIES INC
34. UNITEDHEALTH
GROUP INC
35. UNUM GROUP
36. WELLPOINT INC

38
 Appendix C

Table C.1
Systemic risk ranking of financial firms during June 2006 to June 2007

Realized SES = 0.02−0.15×MES−0.04×LVG−0.12×1[Other]−0.01×1[Insurance]+0.16×1[Broker −Dealers]

MES ranking Name of company Realized SES MES Avg. $Loss(bln) Avg. contribution LVG Fitted rank Assets (bln) ME(bln)
Measuring Systemic Risk

1. INTERCONTINENTAL EXCHANGE INC – 44.24% 3.36% 0.24 0.50% 1.12 16 2.55 10.40
2. E TRADE FINANCIAL CORP – 94.79% 3.29% 0.33 0.69% 7.24 21 62.98 9.39
3. BEAR STEARNS COMPANIES INC – 93.28% 3.15% 0.55 1.16% 25.62 1 423.30 16.66
4. N Y S E EURONEXT – 61.48% 3.05% 0.43 0.90% 1.43 19 16.93 19.44
5. C B RICHARD ELLIS GROUP INC – 88.16% 2.84% 0.20 0.42% 1.55 24 5.95 8.35
6. LEHMAN BROTHERS HOLDINGS INC – 99.82% 2.83% 1.08 2.27% 15.83 4 605.86 39.51
7. MORGAN STANLEY DEAN WITTER & CO – 76.21% 2.72% 2.09 4.39% 14.14 9 1199.99 88.40
8. AMERIPRISE FINANCIAL INC – 62.41% 2.68% 0.35 0.74% 7.72 7 108.13 14.95
9. GOLDMAN SACHS GROUP INC – 60.59% 2.64% 2.13 4.48% 11.25 15 943.20 88.54
10. MERRILL LYNCH & CO INC – 85.21% 2.64% 1.93 4.06% 15.32 5 1076.32 72.56
11. SCHWAB CHARLES CORP NEW – 15.95% 2.57% 0.59 1.24% 2.71 88 49.00 25.69
12. NYMEX HOLDINGS INC – 34.46% 2.47% 0.28 0.59% 1.23 98 3.53 11.57
13. C I T GROUP INC NEW – 91.08% 2.45% 0.26 0.55% 8.45 8 85.16 10.52
14. T D AMERITRADE HOLDING CORP – 28.75% 2.43% 0.24 0.50% 2.40 26 18.53 11.92
15. T ROWE PRICE GROUP INC – 29.83% 2.27% 0.27 0.57% 1.03 101 3.08 13.76
16. EDWARDS A G INC – 0.71% 2.26% 0.11 0.23% 1.46 100 5.24 6.43
17. FEDERAL NATIONAL MORTGAGE ASSN – 98.78% 2.25% 1.24 2.61% 14.00 3 857.80 63.57
18. JANUS CAP GROUP INC – 71.12% 2.23% 0.09 0.19% 1.34 35 3.76 5.16
19. FRANKLIN RESOURCES INC – 51.23% 2.20% 0.62 1.30% 1.08 40 9.62 33.07
20. LEGG MASON INC – 76.98% 2.19% 0.29 0.61% 1.25 38 10.08 12.97
21. AMERICAN CAPITAL STRATEGIES LTD – 91.08% 2.15% 0.15 0.32% 1.73 32 12.15 7.75
22. STATE STREET CORP – 41.07% 2.12% 0.46 0.97% 5.54 28 112.27 23.01
23. WESTERN UNION CO – 30.84% 2.10% 0.36 0.76% 1.34 83 5.33 16.09
24. COUNTRYWIDE FINANCIAL CORP – 87.46% 2.09% 0.48 1.01% 10.39 6 216.82 21.57
25. EATON VANCE CORP – 51.20% 2.09% 0.09 0.19% 1.03 47 0.62 5.54
26. S E I INVESTMENTS COMPANY – 45.61% 2.00% 0.11 0.23% 1.08 50 1.12 5.69
27. BERKLEY W R CORP – 3.57% 1.95% 0.13 0.27% 3.07 31 16.63 6.32
28. SOVEREIGN BANCORP INC – 85.77% 1.95% 0.21 0.44% 8.34 20 82.74 10.11
29. JPMORGAN CHASE & CO – 31.48% 1.93% 3.19 6.70% 9.09 17 1458.04 165.51
30. BANK NEW YORK INC – 29.05% 1.90% 0.54 1.13% 4.64 48 126.33 31.43
31. M B I A INC – 93.34% 1.84% 0.16 0.34% 5.47 25 43.15 8.14
32. BLACKROCK INC – 12.07% 1.83% 0.23 0.48% 1.60 53 21.99 18.18
33. LEUCADIA NATIONAL CORP – 43.54% 1.80% 0.12 0.25% 1.28 61 6.38 7.63
34. WASHINGTON MUTUAL INC – 99.61% 1.80% 0.72 1.51% 8.67 23 312.22 37.63

39
(continued)

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

 Table C.1

40
Continued

MES ranking Name of company Realized SES MES Avg. $Loss(bln) Avg. contribution LVG Fitted rank Assets (bln) ME(bln)
35. NORTHERN TRUST CORP – 16.84% 1.75% 0.23 0.48% 4.92 52 59.61 14.14
36. C B O T HOLDINGS INC 10.12% 1.71% 0.13 0.27% 1.01 69 0.89 10.92
37. PRINCIPAL FINANCIAL GROUP INC – 59.75% 1.71% 0.27 0.57% 10.15 12 150.76 15.61
38. CITIGROUP INC – 85.86% 1.66% 4.19 8.81% 9.25 22 2220.87 253.70
39. LOEWS CORP – 44.08% 1.63% 0.39 0.82% 3.28 44 79.54 27.38
40. GENWORTH FINANCIAL INC – 91.43% 1.59% 0.25 0.53% 7.62 18 111.94 14.96
41. LINCOLN NATIONAL CORP IN – 72.08% 1.59% 0.29 0.61% 10.15 13 187.65 19.21
42. UNION PACIFIC CORP – 15.14% 1.58% 0.45 0.95% 1.70 65 37.30 31.03
43. AMERICAN EXPRESS CO – 69.00% 1.56% 1.08 2.27% 2.70 51 134.37 72.66
44. COMERICA INC – 63.00% 1.55% 0.16 0.34% 6.77 36 58.57 9.27
45. CIGNA CORP – 67.69% 1.54% 0.21 0.44% 3.50 46 41.53 15.03
46. FIDELITY NATIONAL INFO SVCS INC – 27.15% 1.54% 0.14 0.29% 1.42 72 7.80 10.45
47. METLIFE INC – 44.06% 1.52% 0.71 1.49% 11.85 10 552.56 47.82
The Review of Financial Studies / v 30 n 1 2017

48. PROGRESSIVE CORP OH – 31.52% 1.51% 0.28 0.59% 1.89 73 21.07 17.42
49. M & T BANK CORP – 43.46% 1.49% 0.19 0.40% 5.47 60 57.87 11.57
50. NATIONAL CITY CORP – 94.28% 1.48% 0.34 0.71% 7.70 29 140.64 19.18
51. CHICAGO MERCANTILE EXCH HLDG INC – 59.88% 1.47% 0.27 0.57% 1.19 78 5.30 18.64
52. UNUM GROUP – 27.21% 1.46% 0.11 0.23% 5.99 27 52.07 8.95
53. HARTFORD FINANCIAL SVCS GROUP IN – 82.02% 1.46% 0.45 0.95% 11.48 11 345.65 31.19
54. AMBAC FINANCIAL GROUP INC – 98.47% 1.45% 0.13 0.27% 2.69 64 21.06 8.89
55. AETNA INC NEW – 42.17% 1.45% 0.34 0.71% 2.58 66 49.57 25.31
56. LOEWS CORP – 4.54% 1.44% 0.10 0.21% 1.29 82 2.84 8.38
57. BANK OF AMERICA CORP – 68.05% 1.44% 3.27 6.87% 7.46 33 1534.36 216.96
58. PRUDENTIAL FINANCIAL INC – 67.16% 1.43% 0.60 1.26% 10.75 14 461.81 45.02
59. SAFECO CORP 13.56% 1.42% 0.10 0.21% 2.51 68 13.97 6.61
60. HUMANA INC – 38.79% 1.40% 0.14 0.29% 1.97 76 13.33 10.24
61. FEDERAL HOME LOAN MORTGAGE CORP – 98.75% 1.36% 0.60 1.26% 21.00 2 821.67 40.16
62. CHUBB CORP – 2.24% 1.36% 0.30 0.63% 2.74 67 51.73 21.74
63. WELLS FARGO & CO NEW – 10.88% 1.34% 1.58 3.32% 5.19 71 539.87 117.46
64. KEYCORP NEW – 73.09% 1.31% 0.20 0.42% 7.41 41 94.08 13.47
65. WACHOVIA CORP 2ND NEW – 88.34% 1.31% 1.32 2.77% 7.64 37 719.92 98.06
66. B B & T CORP – 26.22% 1.30% 0.30 0.63% 6.23 59 127.58 22.07
67. FIFTH THIRD BANCORP – 77.61% 1.29% 0.29 0.61% 5.33 30 101.39 21.30
68. CAPITAL ONE FINANCIAL CORP – 57.90% 1.28% 0.38 0.80% 4.70 39 145.94 32.60
69. REGIONS FINANCIAL CORP NEW – 73.55% 1.27% 0.30 0.63% 6.06 63 137.62 23.33
70. HUNTINGTON BANCSHARES INC – 62.50% 1.27% 0.07 0.15% 7.23 45 36.42 5.35
71. MASTERCARD INC – 13.49% 1.27% 0.13 0.27% 1.21 85 5.61 13.23

(continued)

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

 Table C.1
Continued

MES ranking Name of company Realized SES MES Avg. $Loss(bln) Avg. contribution LVG Fitted rank Assets (bln) ME(bln)
72. TRAVELERS COMPANIES INC – 12.32% 1.26% 0.45 0.95% 3.54 62 115.36 35.52
73. COMMERCE BANCORP INC NJ – 4.42% 1.26% 0.08 0.17% 7.40 43 48.18 7.08
74. HUDSON CITY BANCORP INC 35.63% 1.26% 0.10 0.21% 6.39 58 39.69 6.50
75. P N C FINANCIAL SERVICES GRP INC – 27.35% 1.24% 0.28 0.59% 5.50 74 125.65 24.69
Measuring Systemic Risk

76. C N A FINANCIAL CORP – 64.73% 1.22% 0.14 0.29% 4.92 42 60.74 12.95
77. UNIONBANCAL CORP 29.14% 1.22% 0.11 0.23% 6.88 54 53.17 8.25
78. AON CORP 9.48% 1.20% 0.14 0.29% 2.55 80 24.79 12.51
79. MARSHALL & ILSLEY CORP – 60.34% 1.20% 0.15 0.32% 5.20 79 58.30 12.34
80. ASSURANT INC – 47.98% 1.18% 0.08 0.17% 4.08 57 25.77 7.13
81. CINCINNATI FINANCIAL CORP – 28.29% 1.17% 0.10 0.21% 2.53 81 18.26 7.46
82. PEOPLES UNITED FINANCIAL INC 5.77% 1.16% 0.07 0.15% 2.75 96 13.82 5.33
83. COMPASS BANCSHARES INC – 6.70% 1.16% 0.11 0.23% 4.48 49 34.88 9.17
84. TORCHMARK CORP – 32.18% 1.15% 0.07 0.15% 2.85 77 15.10 6.40
85. SYNOVUS FINANCIAL CORP – 36.53% 1.12% 0.11 0.23% 3.92 90 33.22 10.04
86. ALLSTATE CORP – 43.63% 1.10% 0.40 0.84% 4.72 55 160.54 37.36
87. FIDELITY NATIONAL FINL INC NEW – 16.80% 1.09% 0.04 0.08% 1.73 87 7.37 5.25
88. ALLTEL CORP 5.98% 1.08% 0.25 0.53% 1.25 89 17.44 23.23
89. SUNTRUST BANKS INC – 62.60% 1.08% 0.34 0.71% 6.35 70 180.31 30.58
90. HEALTH NET INC – 79.37% 1.04% 0.06 0.13% 1.47 91 4.73 5.93
91. ZIONS BANCORP – 66.42% 1.02% 0.09 0.19% 6.26 75 48.69 8.31
92. COVENTRY HEALTH CARE INC – 74.19% 0.99% 0.09 0.19% 1.39 94 6.41 9.01
93. MARSH & MCLENNAN COS INC – 17.94% 0.92% 0.16 0.34% 1.67 93 17.19 17.15
94. S L M CORP – 84.54% 0.92% 0.18 0.38% 6.40 34 132.80 23.69
95. NEW YORK COMMUNITY BANCORP INC – 23.11% 0.92% 0.05 0.11% 5.81 84 29.62 5.33
96. WELLPOINT INC – 47.23% 0.88% 0.43 0.90% 1.60 95 54.19 48.99
97. U S BANCORP DEL – 17.56% 0.88% 0.53 1.11% 4.55 92 222.53 57.29
98. A F L A C INC – 8.52% 0.85% 0.21 0.44% 3.07 86 60.11 25.14
99. UNITEDHEALTH GROUP INC – 47.94% 0.72% 0.49 1.03% 1.47 97 53.15 68.53
100. AMERICAN INTERNATIONAL GROUP INC – 97.70% 0.71% 1.22 2.56% 6.12 56 1033.87 181.67
101. BERKSHIRE HATHAWAY INC DEL(A) – 11.76% 0.41% 0.49 1.03% 2.29 99 269.05 119.00
102. BERKSHIRE HATHAWAY INC DEL(B) – 10.85% 0.39% 49.29

This table contains the list of U.S. financial firms with a market cap in excess of $5 billion as of June 2007. The firms are listed in descending order according to their marginal expected
shortfall at the 5% level (MES) measured over the period June 2006 to June 2007. Realized SES is the return during the crisis. Avg $Loss of an individual firm is the average day-to-day loss
in market cap during days in which the market return was below its 5th percentile. Avg Contribution of an individual firm is the ratio of day-to-day loss in market cap of an individual firm
relative to that of all financial firms, averaged over days where the market was below its 5th percentile. LVG is the market leverage, Fitted Rank is the ranking of firms based on the fitted
values of Realized SES as obtained by the regression given below, Log-Assets is the natural logarithm of total book assets, and ME is market value of equity all as of June 2007. All data are
from CRSP and CRSP merged Compustat.

41
Downloaded from [Link] by University of Bristol Library user on 26 May 2025
The Review of Financial Studies / v 30 n 1 2017

Appendix D

Table D.1
CDS MES ranking of financial firms during June 2006 to June 2007

Name of Type of CDS MES Realized CDS Realized CDS CDS

company institution ranking SES (July 2007– SES (July 2007– MES
June 2008) December 2008)
GENWORTH FINANCIAL INC Insurance 1 145.38% 403.03% 16.40%
AMBAC FINANCIAL GROUP INC Insurance 2 424.10% 389.12% 8.05%

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

WACHOVIA CORP 2ND NEW Depository 3 266.11% 219.94% 7.21%
S L M CORP Other 4 48.88% 113.08% 6.82%
CITIGROUP INC Depository 5 243.16% 278.96% 6.80%
C I T GROUP INC NEW Other 6 243.16% 278.96% 6.80%
M B I A INC Insurance 7 383.11% 303.44% 6.71%
MERRILL LYNCH & CO INC Broker-Dealer 8 200.27% 160.20% 6.37%
WASHINGTON MUTUAL INC Depository 9 261.19% 436.42% 6.15%
WELLS FARGO & CO NEW Depository 10 227.79% 233.43% 6.00%
FEDERAL NATIONAL MORTGAGE ASSN Other 11 194.89% 78.69% 5.70%
LEHMAN BROTHERS HOLDINGS INC Broker-Dealer 12 199.25% 282.25% 5.44%
BANK OF AMERICA CORP Depository 13 207.86% 215.70% 5.23%
MORGAN STANLEY DEAN WITTER & CO Broker-Dealer 14 166.88% 248.96% 4.86%
ALLTEL CORP Other 15 −119.93% −103.25% 4.80%
AMERICAN EXPRESS CO Other 16 237.53% 293.40% 4.36%
FEDERAL HOME LOAN MORTGAGE CORP Other 17 210.58% 94.57% 4.20%
BEAR STEARNS COMPANIES INC Broker-Dealer 18 68.72% 84.96% 4.18%
GOLDMAN SACHS GROUP INC Broker-Dealer 19 135.50% 213.68% 3.87%
UNION PACIFIC CORP Other 20 86.69% 123.56% 3.69%
JPMORGAN CHASE & CO Depository 21 166.95% 182.80% 3.49%
AMERICAN INTERNATIONAL GROUP INC Insurance 22 277.42% 369.20% 3.40%
ALLSTATE CORP Insurance 23 183.66% 271.38% 2.97%
LOEWS CORP1 Insurance 24 136.79% 175.47% 2.67%
PRUDENTIAL FINANCIAL INC Insurance 25 240.25% 394.44% 2.33%
LINCOLN NATIONAL CORP IN Insurance 26 234.94% 403.58% 2.27%
AON CORP Insurance 27 32.41% 55.10% 2.26%
HARTFORD FINANCIAL SVCS GROUP IN Insurance 28 212.09% 368.41% 2.03%
TRAVELERS COMPANIES INC Insurance 29 124.68% 171.62% 1.95%
CHUBB CORP Insurance 30 164.91% 192.52% 1.73%
UNUM GROUP Insurance 31 118.33% 165.43% 0.98%
SAFECO CORP Insurance 32 123.95% 155.92% 0.85%
C N A FINANCIAL CORP Insurance 33 105.34% 218.89% 0.84%
METLIFE INC Insurance 34 220.59% 362.62% 0.75%
TORCHMARK CORP Insurance 35 24.69% 182.45% 0.34%
JANUS CAP GROUP INC Broker-Dealer 36 38.36% 202.27% 0.00%
SCHWAB CHARLES CORP NEW Other 37 149.45% 191.31% 0.00%
AETNA INC NEW Insurance 38 127.42% 192.96% –0.12%
C I G N A CORP Insurance 39 124.73% 267.69% –0.56%
MARSH & MCLENNAN COS INC Insurance 40 31.82% 33.43% –0.63%

This table contains the list of 40 U.S. financial firms with a market cap in excess of $5 billion as of June 2007. The
firms are listed in descending order according to their CDS marginal expected shortfall at the 5% level (MES).
Realized SES is the return on CDS spread during the crisis. CDS data are from Bloomberg.

42
Measuring Systemic Risk

Appendix E

Table E.1
List of Institutions’ names and tickers

TICKER Institution’s Name

ABK AMBAC FINANCIAL GROUP INC
ACAS AMERICAN CAPITAL STRATEGIES LTD
AET AETNA INC NEW

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

AFL A F L A C INC
AGE EDWARDS A G INC
AIG AMERICAN INTERNATIONAL GROUP INC
AIZ ASSURANT INC
ALL ALLSTATE CORP
AMP AMERIPRISE FINANCIAL INC
AMTD T D AMERITRADE HOLDING CORP
AOC AON CORP
AT ALLTEL CORP
AXP AMERICAN EXPRESS CO
BAC BANK OF AMERICA CORP
BBT B B & T CORP
BEN FRANKLIN RESOURCES INC
BER BERKLEY W R CORP
BK BANK NEW YORK INC
BLK BLACKROCK INC
BOT C B O T HOLDINGS INC
BRK BERKSHIRE HATHAWAY INC DEL(A)
BRK BERKSHIRE HATHAWAY INC DEL(B)
BSC BEAR STEARNS COMPANIES INC
C CITIGROUP INC
CB CHUBB CORP
CBG C B RICHARD ELLIS GROUP INC
CBH COMMERCE BANCORP INC NJ
CBSS COMPASS BANCSHARES INC
CFC COUNTRYWIDE FINANCIAL CORP
CG LOEWS CORP2
CI C I G N A CORP
CINF CINCINNATI FINANCIAL CORP
CIT C I T GROUP INC NEW
CMA COMERICA INC
CME CHICAGO MERCANTILE EXCH HLDG INC
CNA C N A FINANCIAL CORP
COF CAPITAL ONE FINANCIAL CORP
CVH COVENTRY HEALTH CARE INC
ETFC E TRADE FINANCIAL CORP
EV EATON VANCE CORP
FIS FIDELITY NATIONAL INFO SVCS INC
FITB FIFTH THIRD BANCORP
FNF FIDELITY NATIONAL FINL INC NEW
FNM FEDERAL NATIONAL MORTGAGE ASSN
FRE FEDERAL HOME LOAN MORTGAGE CORP
GNW GENWORTH FINANCIAL INC
GS GOLDMAN SACHS GROUP INC
HBAN HUNTINGTON BANCSHARES INC
HCBK HUDSON CITY BANCORP INC
HIG HARTFORD FINANCIAL SVCS GROUP IN
(continued)

43
The Review of Financial Studies / v 30 n 1 2017

Table E.1
Continued

TICKER Institution’s Name

HNT HEALTH NET INC
HUM HUMANA INC
ICE INTERCONTINENTALEXCHANGE INC
JNS JANUS CAP GROUP INC
JPM JPMORGAN CHASE & CO
KEY KEYCORP NEW

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

L LOEWS CORP1
LEH LEHMAN BROTHERS HOLDINGS INC
LM LEGG MASON INC
LNC LINCOLN NATIONAL CORP IN
LUK LEUCADIA NATIONAL CORP
MA MASTERCARD INC
MBI M B I A INC
MER MERRILL LYNCH & CO INC
MET METLIFE INC
MI MARSHALL & ILSLEY CORP
MMC MARSH & MCLENNAN COS INC
MS MORGAN STANLEY DEAN WITTER & CO
MTB M & T BANK CORP
NCC NATIONAL CITY CORP
NMX NYMEX HOLDINGS INC
NTRS NORTHERN TRUST CORP
NYB NEW YORK COMMUNITY BANCORP INC
NYX N Y S E EURONEXT
PBCT PEOPLES UNITED FINANCIAL INC
PFG PRINCIPAL FINANCIAL GROUP INC
PGR PROGRESSIVE CORP OH
PNC P N C FINANCIAL SERVICES GRP INC
PRU PRUDENTIAL FINANCIAL INC
RF REGIONS FINANCIAL CORP NEW
SAF SAFECO CORP
SCHW SCHWAB CHARLES CORP NEW
SEIC S E I INVESTMENTS COMPANY
SLM S L M CORP
SNV SYNOVUS FINANCIAL CORP
SOV SOVEREIGN BANCORP INC
STA TRAVELERS COMPANIES INC
STI SUNTRUST BANKS INC
STT STATE STREET CORP
TMK TORCHMARK CORP
TROW T ROWE PRICE GROUP INC
UB UNIONBANCAL CORP
UNH UNITEDHEALTH GROUP INC
UNM UNUM GROUP
UNP UNION PACIFIC CORP
USB U S BANCORP DEL
WB WACHOVIA CORP 2ND NEW
WFC WELLS FARGO & CO NEW
WLP WELLPOINT INC
WM WASHINGTON MUTUAL INC
WU WESTERN UNION CO
ZION ZIONS BANCORP
This appendix contains the tickers and names of the 102 U.S. financial
institutions used in the analysis of the recent crisis.

44
Measuring Systemic Risk

References

Acharya, V. V. 2009. A theory of systemic risk and design of prudential bank regulation. Journal of Financial
Stability 5(3):224–55.

Acharya, V. V., T. Cooley, M. Richardson, and I. Walter. 2010. Manufacturing tail risk: A perspective on the
financial crisis of 2007-09. Foundations and Trends in Finance 4(4):247–325.

Acharya, V. V., T. Eisert, C. Eufinger, and C. W. Hirsch. 2015. Real effects of the sovereign debt crisis in Europe:
Evidence from syndicated loans. CEPR Discussion Paper No. DP10108.

Acharya, V. V., R. Engle, and D. Pierret. 2013. Testing macroprudential stress tests: The risk of regulatory weights.
Journal of Monetary Economics 65:36–53.

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

Acharya, V. V., R. Engle, and M. Richardson. 2012. Capital shortfall: A new approach to ranking and regulating
systemic risks. American Economic Review, Proceedings & Papers 102:59-64.

Acharya, V. V., L. H. Pedersen, T. Philippon, and M. Richardson. 2009. Regulating systemic risk. In V. V. Acharya
and M. Richardson (Eds.), Restoring financial stability: How to repair a failed system (283–304). Hoboken, NJ:
John Wiley and Sons.

Acharya, V. V., T. Philippon, and M. Richardson. Forthcoming. Measuring systemic risk for insurance companies.
In F. Hufeld, R. S. J. Koijen, and C. Thimann (Eds.), The economics, regulation, and systemic risk of insurance
markets (100–23). Oxford: Oxford University Press, 2016.

Acharya, V. V., T. Philippon, M. Richardson, and N. Roubini. 2009. Prologue: A bird’s-eye view: The financial
crisis of 2007–2009. In V. V. Acharya and M. Richardson (Eds.), Restoring financial stability: How to repair a
failed system (1–56). Hoboken, NJ: John Wiley and Sons.

Acharya, V. V., J. Santos and T. Yorulmazer. 2010. Systemic risk and deposit insurance premium. Economic
Policy Review 16(1):89–99.

Acharya, V. V., P. Schnabl and G. Suarez. 2013. Securitization without risk transfer. Journal of Financial
Economics 107:515–36.

Acharya, V. V., and T. Yorulmazer. 2007. Too many to fail: An analysis of time-inconsistency in bank closure
policies. Journal of Financial Intermediation 16(1):1–31.

Adrian, T., and M. Brunnermeier. Forthcoming. CoVaR. American Economic Review.

Adrian, T., and H. Shin. 2010. Leverage and liquidity. Journal of Financial Intermediation 19(3):418–37.

Allen, F., and D. Gale. 2007. Understanding financial crises. New York: Oxford University Press.

Allen, L., T. G. Bali, and Y. Tang. 2012. Does systemic risk in the financial sector predict future economic
downturns? Review of Financial Studies 25(10):3000–36.

Allen, L., and A. Saunders. 2002. Credit risk measurement: New approaches to value at risk and other paradigms.
Hoboken, NJ: Wiley.

Artzner, P., F. Delbaen, J. M. Eber, and D. Heath. 1999. Coherent measures of risk. Mathematical Finance
9(3):203–28.

Backus, D., M. Chernov, and I. Martin. 2009. Disasters implied by equity index options. Working Paper, New York
University.

Barro, R. 2006. Rare disasters and asset markets in the twentieth century, Quarterly Journal of Economics.
121:823–66.

Billio, M., M. Getmansky, A. Lo, and L. Pelizzon. 2012. Econometric measures of connectedness and systemic
risk in the finance and insurance sectors. Journal of Financial Economics 104(3):535–59.

Bisias, D., M. Flood, A. Lo, and S. Valavanis. 2012. A survey of systemic risk analytics. Office of Financial
Research Working Paper No. 0001.

45
The Review of Financial Studies / v 30 n 1 2017

Borio, C., and M. Drehmann. 2009. Assessing the risk of banking crises—revisited. BIS Quarterly Review.
March, 29–46.

Borio, C., N. Tarashev, and K. Tsatsaronis. 2009. The systemic importance of financial institutions. BIS Quarterly
Review. September, 75–87.

Bostandzic, D., and G. N. F. Weiss. 2015. Why do some banks contribute more to global systemic risk? Working
Paper.

Brownlees, C., B. Chabot, E. Ghysels, and C. Kurz. 2015. Back to the future: Backtesting systemic risk measures
during the Great Depression and historical bank runs. Working Paper, Federal Reserve.

Brownlees, C., and R. Engle. Forthcoming. SRISK: A conditional capital shortfall index for systemic risk

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

assessment. Review of Financial Studies.

Brunnermeier, M. K., G. N. Dong, and D. Palia. 2011. Banks’ non-interest income and systemic risk. Working
Paper, Princeton University.

Brunnermeier, M., and L. H. Pedersen. 2009. Market liquidity and funding liquidity. Review of Financial Studies
22:2201–38.

Caprio, G., and D. Klingebiel. 1996. Bank insolvencies: cross country experience. World Bank, Policy Research
Working Paper No. 1620.

Chodorow-Reich, G. 2014. The employment effects of credit market disruptions: Firm-level evidence from the
2008–09 financial crisis. Quarterly Journal of Economics. 129:1–59.

Crockett, A. 2000. Marrying the micro- and macro-prudential dimensions of financial stability, BIS Speeches.
September 21.

Cummins, J. D., and M. Weiss. 2014. Systemic risk and regulation of the U.S. Insurance Industry. In J. Biggs,
M. Richardson, and I. Walter (Eds.), Modernizing insurance regulation (85–136). Hoboken, NJ: John Wiley &
Sons.

de Jonghe, O. 2010. Back to the basics in banking? A micro-analysis of banking system stability. Journal of
Financial Intermediation 19(3):387–417.

Diamond, D., and P. Dybvig. 1983. Bank runs, deposit insurance, and liquidity. Journal of Political Economy
91(3):401–419.

Diamond, D., and R. G. Rajan. 2005. Liquidity shortages and banking crises. Journal of Finance 60(2):615–47.

Duffie, D. 2010. How large banks fail and what to do about it. Princeton, NJ: Princeton University Press.

Engle, R. F. 2009. Long run skewness and systemic risk. Presentation at International Association of Financial
Engineers (IAFE). Mimeo, New York University Stern School of Business.

Engle, R., E. Jondeau, and M. Rockinger. 2014. Systemic risk in europe. Review of Finance 19:145–190.

Farhi, E., and J. Tirole. 2009. Collective moral hazard, maturity mismatch and systemic bailouts. Working Paper,
Harvard University.

Flannery, M. J. 2005. No pain, no gain? Effecting market discipline via reverse convertible debentures. In H.
S. Scott (Ed.), Capital adequacy beyond Basel: banking, securities, and insurance. Oxford: Oxford University
Press.

Gabaix, X. 2009. Power laws in economics and finance. Annual Review of Economics 1(1):255–94.

Garleanu, N., and L. H. Pedersen. 2007. Liquidity and risk management. American Economic Review 97(2):
193–97.

Giesecke, K., and B. Kim. 2011. Systemic risk: What defaults are telling us. Management Science 57:1387–405.

Goodhart, C., and M. Segoviano. 2009. Banking stability measures. IMF Working Paper No. WP/09/4.

46
Measuring Systemic Risk

Gray, D., and A. A. Jobst. Forthcoming. Tail dependence measures of systemic risk using equity options data—
implications for financial stability. International Monetary Fund (IMF), Washington, DC.

Gray, D. F., R. C. Merton, and Z. Bodie. 2008. New framework for measuring and managing macrofinancial risk
and financial stability. Working Paper No. 09–015, Harvard Business School.

Hansen, L. P. 2014. Challenges in identifying and measuring systemic risk. In M. Brunnermeier and A.
Krishnamurthy (Eds.), Risk topography: Systemic risk and macro modeling (15–30). Chicago: University of
Chicago Press.

Hart, O., and L. Zingales. [Link] capital regulation for large financial institutions. Working Paper, University
of Chicago.

Downloaded from [Link] by University of Bristol Library user on 26 May 2025

Hoggarth, G., R. Reis, and V. Saporta. 2002. Costs of banking system instability: Some empirical evidence.
Journal of Banking and Finance 26(5):825–55.

Honohan, P., and D. Klingebiel. 2000. Controlling fiscal costs of bank crises. Working Paper #2441, World Bank.

Huang, X., H. Zhou, and H. Zhu. 2009. A framework for assessing the systemic risk of major financial institutions.
Journal of Banking and Finance 33(11):2036–49.

———. 2012. Systemic risk contributions. Journal of Financial Services Research 42(1):55–83.

Inui, K., and M. Kijima. 2005. On the significance of expected shortfall as a coherent risk measure. Journal of
Banking and Finance 29:853–864.

Jiang, H., and B. Kelly. 2014. Tail risk and asset prices. Review of Financial Studies 27(10):2841–71.

Kashyap, A., R. Rajan, and J. Stein. 2008. Rethinking capital regulation. Paper presented at Kansas City
Symposium on Financial Stability.

Laeven, L., and F. Valencia. 2013. Systemic banking crises database. IMF Economic Review 61(2):225–70.

Lehar, A. 2005. Measuring systemic risk: A risk management approach. Journal of Banking and Finance 29:
2577–603.

Lorenzoni, G. 2008. Inefficient credit booms. Review of Economic Studies 75(3):809–33.

Pedersen, L. H. 2009. When everyone runs to the exit. International Journal of Central Banking 5(4):177–99.

Raviv, A. 2004. Bank stability and market discipline: Debt-for-equity swap versus subordinated notes. Working
Paper, Brandeis University.

Reinhart, C. M., and K. Rogoff. 2008. Is the 2007 U.S. sub-prime financial crisis so different? An international
historical comparison. American Economic Review: Papers & Proceedings 98(2):339–44.

Rochet, J., and J. Tirole. 1996. Interbank lending and systemic risk. Journal of Money, Credit and Banking
28(4):733–62.

Segoviano, M., and C. Goodhart. 2009. Banking stability measures. IMF Working Paper 09/04, International
Monetary Fund.

Yamai, Y., and T. Yoshiba. 2005. Value-at-risk versus expected shortfall: A practical perspective. Journal of
Banking and Finance 29:997–1015.

47