Full text of Current Issues in Economics and Finance : Robust Capital Regulation, Volume 18, Issue 4

View original document

The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.

www.newyorkfed.org/research/current_issues
F

2012
F

Volume 18, Number 4

in Economics and Finance

current issues
FEDERAL RESERVE BANK OF NEW YORK

Robust Capital Regulation
Viral Acharya, Hamid Mehran, Til Schuermann,
and Anjan Thakor
Regulators and markets can find the balance sheets of large
financial institutions difficult to penetrate, and they are mindful of
how undercapitalization can create incentives to take on excessive
risk. This study proposes a novel framework for capital regulation
that addresses banks’ incentives to take on excessive risk and
leverage. The framework consists of a special capital account in
addition to a core capital requirement. The special account would
accrue to a bank’s shareholders as long as the bank is solvent, but
would pass to the bank’s regulators—rather than its creditors—if
the bank fails. By design, this special account thus limits risk taking,
but ensures that creditors’ disciplining incentives are preserved.
In early 2009, the largest U.S. bank holding companies, including the nineteen that
would undergo federally mandated stress tests later that year, were all considered
adequately capitalized according to government regulatory standards. The market,
however, had a different view: At the time, most of those institutions were trading at
less than book value and all were at or near record highs for their credit default swap
spreads—an indicator of a company’s likelihood of default. Indeed, some of them
might have failed in fall 2008 had there not been a systemwide capital injection by the
government.
For the regulators and the markets, it was difficult to penetrate the balance sheets
of those financial institutions, let alone assess which ones needed additional capital,
the amount required, and the cost of that capital. The bank stress test in 2009 offered
some temporary clarity, at least in regard to the nineteen participating institutions,
and its success prompted Congress to enshrine stress testing in the law with the
Dodd-Frank Wall Street Reform and Consumer Protection Act. In fact, the most recent
round of stress tests covered thirty institutions. Nonetheless, we continue to grapple
with the broader questions of financial fragility and capital adequacy.1
In this edition of Current Issues, we provide a fresh perspective on the forces that
shape banks’ capital structure choices, showing how these choices are distorted by
regulatory safety nets that give banks incentives to take on excessive risk and leverage.
To address the negative aspects of these incentives, specifically the privatization of
banks’ profits and the socialization of their losses, we offer a novel approach to capital
regulation. Our proposal involves a two-part capital requirement: a core capital
requirement (much like existing requirements) and a special capital account requirement.
1 The 2009 stress test is more formally known as the Supervisory Capital Assessment Program, or SCAP,
while the recent incarnation is called the Comprehensive Capital Analysis and Review (or CCAR, for the
original nineteen banks) and Capital Analysis and Review (or CapPR, for eleven banks with more than
$50 billion in assets). For more information on SCAP, see http://www.federalreserve.gov/newsevents/
press/bcreg/20090507a.htm; on CCAR and CapPR, see http://www.federalreserve.gov/newsevents/press/
bcreg/20120313a.htm.

CURRENT ISSUES IN ECONOMICS AND FINANCE v Volume 18, Number 4

The special account would involve capital that must be invested
in Treasury securities or their equivalents. The assets segregated
in the special account would accrue to the bank’s shareholders as
long as the bank is solvent and to the regulators, rather than the
creditors, if the bank fails.

Chart 1

Capital Ratios by Sector: Year-End 2009
Percent
80
70

The basic idea, formally developed in Acharya, Mehran, and
Thakor (2010), is to exploit the role of equity in reducing the risk
appetite of banks by requiring them to hold additional capital as
well as the role of uninsured debt in encouraging the monitoring
of bank management by ensuring that creditors have enough
“skin in the game” to find such monitoring desirable.

The Treasury securities in the special capital account could be
used by regulators to support the financial system if it is threatened by failing banks. For instance, the securities could be used
to reduce the banks’ cost of lending to households and the real
sector, or they could be saved as a cushion for future crises.

In addition, the quantification of the capital requirement need
not depend exclusively on the use of historical data for calibration of the bank’s risks; instead, it would rely on several different
approaches, such as market-based signals of bank-level and
systemic risk as well as regulatory intelligence gathered through
periodic stress tests of the financial sector. Besides being “robust”
in the sense of calibration, our proposal is also robust in the
sense that it does not rely solely on bank equity to provide the
right incentives; it also recognizes market discipline provided by
uninsured creditors.

The Capital Structure Decision
How does any firm decide on its capital structure? In other words,
how much equity, and debt, should it use? And why might the
answer be different for a bank than for a nonfinancial firm? In
particular, why do banks tend to have such high levels of debt?
A typical nonfinancial firm has equity (capital) that exceeds
50 percent of its assets. Chart 1 shows the ratios of equity to assets—also called capital ratios—for 5,988 firms across a broad
range of sectors at year-end 2009. Credit intermediaries have
by far the lowest capital ratio, at 12.1 percent, less than half the
capital of the next sector, insurance, at 29 percent, which itself is
less than half the ratio for most nonfinancial sectors.2
Academic studies of corporate finance entered the debate with
Modigliani and Miller’s famous (1958) “leverage indifference”
theorem, and the debate has been reinvigorated with contributions
by Mehran and Thakor (2011) and Admati et al. (2010). In a world
without frictions (no taxes, no bankruptcy costs, no safety net,
such as a lender of last resort or a deposit insurer), Modigliani
and Miller show that the capital structure decision—that is, the
decision on how to finance the balance sheet—for a given size
of firm and a given asset portfolio composition matters only if it
2 Credit intermediaries are depository institutions and nondepository credit
institutions.

50
40
30
20

int C
erm red
ed it
iat
Ins ion
ur
an
an Uti ce
d t liti
ran es
Co spor
ns
t
tr u
cti
on
Tra
de
Al
l se
cto
rs
Re
al
est
ate
Se
rv
Ma
ice
nu
s
fac
tur
ing
Mi
Ag ning
ric
ul
ad P ture
mi ub
nis lic
tra
tio
n

Source: Compustat.
Notes: Equity is the residual of total (book) assets less total (book) liabilities.
Firms in the sample total 5,988.

affects the value of the firm. Their argument takes the balance
sheet and thus the investment decisions that formed the balance
sheet (projects, machines, buildings, or, in the case of a bank, loans
made or securities bought) as given, implying that the financing
mix decision is separate from the firm’s investment decision.
The real world, of course, looks quite different from the one
created by Modigliani and Miller, particularly for banks. It has
been argued that banks tend to be highly levered because the
conditions imposed by Modigliani and Miller do not apply to
them. And indeed, a number of reasons have been cited to
support this view.
The standard argument against applying the Modigliani and
Miller theorem is that deposits are a factor of production in banking: Banks not only use deposits to make loans, the institutions
also provide liquidity and transaction services to depositors. As a
result, we should expect banks to be highly levered, since deposits
are a form of debt. However, given the constrained supply of core
deposits, it is not obvious why banks cannot add large amounts
of equity to whatever deposits they gather, which would then
make it difficult to explain the paucity of equity in banking.
A second popular argument is that banks prefer high leverage
because interest payments on debt are tax deductible while shareholder dividends are not. This statement is true, but it cannot
explain why banks are more levered than nonfinancial firms that
enjoy the same debt tax shield.
A third argument, one that we generally favor, has appeared in
theories that emphasize the monitoring and disciplining role of
leverage. As debt increases, the loss-absorption capacity provided
by equity capital in the event of bankruptcy shrinks, inducing

creditors to monitor the activities of management more closely
and to raise the price of debt to compensate for the increased risk.
This effect is present for all firms, but bank funding appears
to be unique because it is a liability that comes in the form of
demand deposits. Calomiris and Kahn (1991) were the first to
note that uninsured depositors who monitor the bank can decide
to withdraw their deposits if they come to suspect managerial
inefficiency or fraud. Observing these withdrawals, uninformed
depositors may follow suit. Their action precipitates a full-scale
bank run that may force liquidation of the institution. So fear
of such a run can induce the bank’s management to stay on the
straight and narrow.

Although the disciplining role of debt can help
reduce certain agency costs in banks, it can also have
deleterious consequences.
In this framework, leverage is needed in order for market
discipline to control agency problems.3 The effect extends beyond
deposits, however. As borrower (or, in this case, bank) risk increases, lenders tend to shorten maturities when the option of not
renewing the debt becomes more valuable. Indeed, such maturity
shortening was in broad evidence throughout the financial crisis.
Since this line of reasoning is meant to justify the heavy use
of demandable debt by banks, the potential discipline imposed
by such debt is substantial (at least in theory), because the bank
can be shut down very quickly by creditors who refuse to roll over
debt. To explain why nonfinancial firms, which also stand to benefit from the disciplining role of leverage, do not use this form of
debt, one must invoke the argument that the potential for agency
problems, and hence the need for market discipline, is much
greater in banking than in nonfinancial firms. The greater ease
with which banks can change their asset mix and keep it
hidden from all but the most diligent and skilled monitors is
likely an important reason.
The recent financial crisis has in fact provided many examples of creative manufacturing of assets whose “tail risks” were
far from transparent, even to some insiders. As the riskiness of
those assets became apparent to outsiders, the market reacted
quickly by either shortening the maturity of credit or refusing
to roll it over altogether.
3 See Diamond and Rajan (2000, 2001), who justify demandable bank debt given

the inability of bankers to pledge their relationship-specific rents to depositors.
Some (for example, Admati et al. [2010]) contend that the market discipline
argument in favor of bank leverage has been overstated, and that the cost of
bank equity is not as high as many have asserted. While it is true that, in many
cases, creditors may have been ineffective in disciplining banks during this crisis,
we believe it is because their incentives to do so were diluted by the de facto
protection afforded to them by the regulatory safety net.

Although the disciplining role of debt can help reduce certain
agency costs in banks, it can also have deleterious consequences.
Jensen and Meckling (1976) argue that sufficiently high leverage
creates asset-substitution or risk-shifting moral hazard: Bank
managers and shareholders prefer riskier gambles to safer ones,
simply out of a desire to maximize the value of their equity option on bank assets. Coping with this moral hazard requires one
to limit the use of leverage unless the discipline achieved through
liabilities is able to keep pace with the potential for asset substitution. So for banks, the discipline has to be particularly harsh.
“Run”-able demand deposits provide just that discipline, but in
general asset-substitution moral hazard can dominate the value
of discipline when bank leverage is high.
This tension between the run-based disciplining role of leverage and the risk-inducing role of debt has been formally examined
in a recent paper by Acharya, Mehran, and Thakor (2010), which
considers a model in which creditors can incur a fixed monitoring cost to detect managerial inefficiency or fraud.4 Upon finding
either, they could threaten to liquidate the bank, thereby creating a
“creditor run.” At the same time, however, the presence of leverage
can give bank managers the opportunity to gamble and expose
the bank to risky outcomes. Acharya, Mehran, and Thakor show
theoretically that the bank is caught between a rock and a hard
place in its choice of a privately optimal capital structure. If the
bank does not choose a sufficiently high amount of leverage, then
its creditors do not have enough “skin in the game” to find the
investment in monitoring worthwhile. They could threaten the
bank with liquidation for observed underperformance, thereby
imposing the necessary market discipline. However, if the leverage
ratio is too high, asset-substitution moral hazard is triggered, and
the bank may be induced to take excessive risk at the creditors’
expense, thereby expropriating wealth from the creditors and
depositors to the benefit of the shareholders.5
Acharya, Mehran, and Thakor show that the bank’s privately
optimal capital structure must navigate between these two forms
of moral hazard. In particular, leverage must be high enough to
induce the discipline imposed by creditors, but low enough to
ensure that the bank’s risk taking is not excessive. This balance
4 Acharya and Thakor (2010) also see an inherent conflict between the market
discipline of an individual bank having a fragile capital structure and the financial
stability of the system, when the fragility of an individual bank in the form of a
depositor or creditor run can induce (potentially inefficient) information-based
runs on other banks.
5 For financial firms, this asset-substitution moral hazard takes on particular
importance, as it is far easier to reallocate financing across different transactions
and alter risks at a high frequency before creditors can discern the problem—in
contrast to, say, an automobile company that would face immediate customer
outrage were it to make riskier cars (a point referred to as the “paradox of
liquidity” by Myers and Rajan [1998]). In Acharya, Mehran, and Thakor, the
manager’s compensation is perfectly correlated with the wealth of the initial
shareholders, so the manager has the same incentive as shareholders to take risk.
This is consistent with the idea that bank managers may be incentivized to take
high risk because of the nature of their compensation, a sentiment reflected in
the greater regulatory oversight of bank executive compensation that occurred
following the recent crisis.

www.newyorkfed.org/research/current_issues

CURRENT ISSUES IN ECONOMICS AND FINANCE v Volume 18, Number 4

ensures that bankers are taking economically attractive risks
(such as making loans to positive net-present-value projects),
but, at the same time, they are not making excessively risky bets
(such as funding undercapitalized mortgages).
Since bank-level agency problems are adequately taken into
account by bank-level capital structure problems, this argument
does not provide a reasonable case for regulatory setting of capital
requirements. Acharya, Mehran, and Thakor show, however, that
the argument for a privately optimal capital structure can break
down completely in the presence of regulatory distortions, which
we now consider.

The Role of Regulatory Safety Nets and a Step toward
Robust Capital Regulation
So far, we have deliberately excluded the role of regulatory safety
nets in the bank’s leverage choice. These safety nets mean that, if
a bank fails, its creditors do not have to take all of the losses on
their claims that they otherwise would. The bulk of a commercial bank’s deposits is insured, whereas its equity is not. Deposit
insurance—as well as other safety-net initiatives, such as ex post

[Safety nets] help prevent a wide-scale collapse
of the intermediation services provided by the
banking sector and can avert various forms of
contagion that could hurt the economy.
bailouts of some failing banks—turns overnight debt financing,
which would ordinarily be very risk sensitive, into financing that
is more tolerant of changes in a bank’s riskiness. A similar argument applies to undercapitalized over-the-counter derivatives
exposures that large financial firms have to one another.6
Another financial safety net is provided by the central bank in
its role as lender of last resort. Through its discount window, the
central bank plays a liquidity transformation role by providing
banks with access to short-term liquidity. Discount window access enables banks to turn illiquid assets (the collateral pledged)
into liquid assets minus a “haircut.”7 The discount window
complements deposit insurance. While deposit insurance allows
6 See Song and Thakor (2007), who show that deposit insurance adds to the
“stickiness” of a bank’s core deposits. That is, deposit insurance can induce a sort
of self-selection among investors: Those who are more interested in the bank’s
transaction services but less able or willing to monitor the bank choose to become
insured depositors, whereas the more active monitors become suppliers of
uninsured (purchased) money. Consequently, core deposits, which are covered
by deposit insurance, are less subject to withdrawal risk.
7

We subsequently discuss the practical difficulties of distinguishing between
insolvency and illiquidity.

banks to obtain cheaper funding and subsidizes the liability side
of the balance sheet, the discount window gives banks access to
short-term liquidity when the market is unwilling to provide it.
There are many ways of rationalizing these safety nets, but
certainly they help prevent a wide-scale collapse of the intermediation services provided by the banking sector and can avert
various forms of contagion that could hurt the economy (such
as a severe recession or worse). In other words, they are part and
parcel of the desire for a safe, sound, and stable banking system.
Moreover, they help banks engage in effective maturity transformation: Liabilities can be of shorter maturity in the presence
of deposit insurance, and assets can be of longer maturity (and
hence less liquid) in the presence of the discount window. In
short, there are valid economic reasons to have regulatory safety
nets in banking, when viewed purely from an ex post standpoint
in midcrisis.
However, it is now becoming abundantly clear—both in
theory and in practice—that regulatory safety nets can come at
a fairly substantial cost, not just ex post in terms of fiscal outlays
(Ireland’s sovereign credit risk following bank bailouts being
a prime example),8 but also ex ante in terms of moral hazard.
The most obvious moral hazard takes the form of banks being
encouraged to become more highly levered. Because creditors
do not face the same risk exposure as they would in the absence
of safety nets, the credit disciplining effect discussed earlier
is dampened, and the pricing of bank debt becomes relatively
insensitive to the amount of leverage. As a result, leverage appears
“cheap” to banks, even as they take on increasing amounts that
make the bank riskier and riskier.9
The presence of the safety net—deposit insurance and the
role of lender of last resort—upsets the balance of a finely tuned
capital structure, as described by Acharya, Mehran, and Thakor
(2010): enough equity capital to attenuate asset-substitution
moral hazard, yet not so much as to water down the market
discipline provided by uninsured creditors. In addition to
describing this bank-specific effect, Acharya, Mehran, and
Thakor argue that bank risk taking has an important collective
or systemic dimension. Banks can choose to take not only excessive idiosyncratic risk, but also risk that is highly correlated

8 See, for example, Acharya, Drechsler, and Schnabl (2011) for a theoretical
analysis of bank bailouts in a world of limited fiscal resources and for supporting
empirical evidence provided by the euro zone’s sovereign credit woes.
9 Merton (1977) shows that deposit insurance essentially gives the bank an
option to “put” its assets to the deposit insurer in the event that the assets fall
below liabilities, and that the value of this option increases as the bank’s leverage
goes up. The discount window has a similar effect. The availability of financing
through the discount window significantly reduces the refinancing risk in
maturity transformation. Moreover, as Farhi and Tirole (2009) point out, the
central bank may be unable to tell whether a bank is illiquid or insolvent. This
means that insolvent banks may also be able to stay alive by tapping the discount
window, an action that in turn encourages banks to become more highly levered.

across banks (for example, by engaging in “herding” behavior on
similar asset classes for lending or investment purposes).10

the extent that some guarantees are implicit in nature, appropriate
pricing of deposit insurance premiums may not suffice.

If all banks choose excessive and highly correlated risks, they
are likely to fail together. And faced with industrywide failures,
regulators are more likely to step in and bail out banks because
such an industry collapse would have potentially devastating real
economic effects.11 Acharya, Mehran, and Thakor show that the
mere anticipation of this forbearance when banks fail en masse
may cause them to choose highly correlated, excessively risky
projects. In pricing the uninsured debt, creditors will not “punish”
banks ex ante for systemic risk in their portfolio choices because
they—the creditors—anticipate being bailed out ex post. All

Acharya, Mehran, and Thakor argue that to prevent the
“looting” of taxpayer funds (to borrow a term from Akerlof and
Romer [1993]) through excessive leverage and correlated risk
taking by banks, the regulator needs to impose a well-designed
scheme of capital regulation that is robust in the following sense.
The capital regulation must be such that the bank’s leverage ratio
stays below the upper bound beyond which the banks collectively
wish to take excessive, correlated risks and so extract subsidies
from the safety net. At the same time, creditors should not
perceive banks to be so safe that they fail to discipline their asset
choices through monitoring and timely pricing of credit risks.
Acharya, Mehran, and Thakor propose two important measures
to deal with this trade-off.

It is now becoming abundantly clear—both in
theory and in practice—that regulatory safety nets
can come at a fairly substantial cost, not just ex post
in terms of fiscal outlays . . . , but also ex ante in
terms of moral hazard.
market discipline of debt is lost, and banks end up choosing
much higher leverage ex ante.
The channel of moral hazard is interesting. Ex post, it is the
creditors of banks who get bailed out, typically not bank shareholders, but this means that, ex ante, creditors do not price the
correlated risk of bank projects adequately. For bank shareholders,
this situation increases the attractiveness of riskier gambles, and
banks are often inclined to pursue these until the bets (almost inevitably) go bad. When this happens, the lender of last resort bails
out banks, and taxpayer funds get transferred to bank creditors.
Because these transfers are reflected in the ex ante pricing of debt,
it is effectively an ex ante wealth transfer from taxpayers to bank
shareholders, managers, and employees.
One avenue available for mitigating this correlation-induced
systemic risk is through appropriate pricing of deposit insurance.
Specifically, deposit insurance premiums should cover not just
the expected loss (to the deposit insurance fund) for a given bank
but, more important, its contribution to overall banking system
risk, which is a combination of size and correlation.12 However, to
10

Acharya (2009) models this collective agency problem and refers to it as
“systemic risk-shifting.” For supporting evidence on correlated bank exposures,
see Schuermann and Stiroh (2006), who show that, among firms that make up
the S&P 500, the average equity return correlation among banks is higher than the
correlation for firms in any other industry (energy firms come in second).
11

See Acharya and Yorulmazer (2007) and Farhi and Tirole (2009) for formal
analyses of this time-inconsistency problem facing regulators when they have
discretion over bailouts and expansionary monetary policy, respectively.
12

See, among others, Acharya, Santos, and Yorulmazer (2010) and Kuritzkes,
Schuermann, and Weiner (2005).

One measure is a regular core capital requirement guaranteeing that the bank’s leverage never exceeds the upper bound and
so keeps risk-shifting incentives in check. The other—and more
innovative—measure is a “special capital account” that is built up
through retained earnings made possible by restricting dividend
payouts by the bank. An important purpose of this special capital
account is to provide the bank with a readily available resource
that can be tapped to instantaneously replenish a diminished core
capital account to its desired level. In other words, an automatic
and mechanical transfer from the special capital account to the
core capital account would occur whenever the bank suffers an
income shock that depletes the core account. Restrictions on
dividend payouts are then imposed on the bank to ensure that
the special capital account is rebuilt to its original level over time
through retained earnings.
This special account needs to have several noteworthy
features. First, the capital must be invested in predesignated
liquid securities such as Treasuries in order to remove managerial discretion over the use of that capital. This action eliminates
the potential moral hazard of bank managers being less efficient
because they have excess cash that is not needed to run the bank.
Although managers clearly have limited control rights over this
capital account, it does have value that can be monetized—for
instance, through the sale of the bank.
Second, the capital account accrues to the shareholders as long
as the bank is solvent—for instance, it can be used to meet special
capital account requirements in the next period. However, if the
bank becomes insolvent and there is no industrywide rescue of
banks by the lender of last resort, the capital account accrues to
the regulator rather than to the bank’s creditors. The idea is that,
in an industrywide rescue, there is a scarcity of bank capital, and
since the regulator is implicitly recapitalizing the system, the
special capital account also accrues indirectly to creditors.
However, in the case of individual bank failures, the assets of
these institutions can be acquired by well-capitalized players in

www.newyorkfed.org/research/current_issues

CURRENT ISSUES IN ECONOMICS AND FINANCE v Volume 18, Number 4

the financial system. Here, the creditors can be forced to take a
haircut without the system suffering substantial repercussions.13
That creditors do not benefit from the special capital account
in the event of individual bank failures means that this capital is
“invisible” to creditors and ensures that they have enough “skin
in the game” to discipline the banks—that is, their incentives to
credibly threaten withdrawal of financing and premature liquidation are not diluted by having this additional capital in the bank.
Regulators would need to be explicitly directed, by regulation and
law, to take possession of the special capital account in the event of
bank insolvency. They could deploy the acquired capital to aid parts
of the financial sector affected by the failing institution, or even to
directly assist affected parts of the household and real sectors.
Alternatively, this capital could be saved as a buffer against a fullblown future crisis.
Thus, our overall proposal is a form of “capital preservation,”
whose goal is to ensure that the probability of the bank falling
into an insolvent state is minimized ex ante. But it also provides
for “market discipline preservation,” whose goal is to ensure
that creditors have sufficient incentives to intervene in underperforming banks.14 In this way, the capital account acts as a
deductible on explicit and implicit government insurance claims
(prepaid by the shareholders) and serves to reduce systemwide
losses given default, though not necessarily so for creditors of
any particular bank.
Third, since the special capital account is built up gradually
through retained earnings, the bank typically does not have to
raise equity in order to satisfy its capital requirement. Furthermore, since such a transfer occurs mechanically based on marketobserved performance variables, no new information is released
to the market—in contrast to a bank’s voluntary issuance of
equity, which reveals private information and will generally be
perceived by the market as negative news. Thus, the bank is able to
avoid the information costs associated with issuing equity (as in
Myers and Majluf [1984]), which often make bank managers and
CEOs reluctant to issue equity in the first place. Meanwhile, the
bank is not put in a position of having to raise equity when it
is in financial distress and doing so might be difficult or costly.
In this sense, Acharya, Mehran, and Thakor’s proposal can
be thought of as a mechanism to enforce countercyclical capital
requirements, which have been proposed as an important part
of the regulatory toolkit for macroprudential regulation of the
financial sector.

While the theoretical argument of Acharya, Mehran, and Thakor implies
that too much deposit insurance may not be desirable because it compromises
depositors’ monitoring of banks, our proposal in practice would not pass
haircuts to insured depositors.
14

Note that our proposal focuses on reducing the bank’s likelihood of getting
into trouble rather than on improving methods for resolving bank distress,
which is an important regulatory topic in itself.

Fourth, in the absence of a systemwide rescue, the transfer
from the special capital account to the core capital account and the
accompanying dividend restrictions are mechanically triggered,
based on prespecified rules (linked, for example, to the financial
sector’s total market capitalization loss in the last year). In other
words, regulators have no discretion in the matter. This way, there
is no bank-specific information conveyed by these actions, and the
issue of the trigger somehow becoming a self-fulfilling prophecy
of failure for an individual bank would not arise.15 Also, because
the special capital account is invested in Treasury securities or
other cash-like instruments, the bank always has a buyer of liquid
assets in the event of a liquidity crunch.
Finally, because the special capital account restricts the ability
of banks to profit at the expense of taxpayers, bank shareholders
would in fact be discouraged from excess leverage and correlated
risk taking in the first place. Similarly, creditors would monitor the
bank because additional bank capital does not buffer them against
losses in times when rescues are not systemwide. Thus, the purpose
of the special capital account is to give banks and their creditors
the right incentives—rather than the mechanical role of buffering
against future losses, which is the current approach to using bank
capital under the Basel capital requirements.16
In this respect, our proposal differs from the use of contingent capital, which is a new instrument with built-in features to
convert debt into equity once the bank or the system is in a state
of crisis. Contingent capital is largely an ex post mechanism that
attempts to deal with a paucity of equity capital in midcrisis (see,
for example, Flannery [2009]), whereas our proposal is an ex ante
incentive device, intended to diminish the probability of entering
a crisis state in the first place. Furthermore, calibration of the two
capital requirements can be made more robust by relying on multiple ways to assess the risk of bank assets (including the use of
historical data, market data, regulatory stress tests, and systemic
risk assessments)—thereby maintaining a buffer in the special
capital account.

Would Higher Capital Requirements Hurt a Bank’s Value?
The capital regulation framework described in Acharya, Mehran,
and Thakor (2010) is intended to inject more capital into banking
and steer banks to actions that would reduce the likelihood of
crises, without diluting the monitoring incentives of uninsured
creditors. Opponents of higher capital requirements might object
on two grounds, however.

15 If

the trigger is based on regulatory discretion, it will convey information to the
market that the regulator knows that something is wrong. This will cause creditors
to withdraw funding to the bank, precipitating the very crisis the regulator had
wished to avoid.
16 Of

course, in practice, regulatory design of required leverage ratios may not
fluctuate on a frequent or perfect basis, resulting in actual contributions to
the special capital account—an issue that would require a certain amount
of regulatory calibration over time.

First, they would argue that equity capital is very expensive for
banks in the sense that bank shareholders demand a very high
return on their investment. So, asking banks to post more capital
will force them to reduce the size of their balance sheets because
they will be unable to find investments with sufficient rates of
return to cover those demanded by shareholders on the additional
equity. This reduced lending by banks, in turn, will lead to less
credit creation and thus lower growth and hurt global GDP.
We can show the weakness in this argument by discussing
a corollary to it: Higher leverage is preferred because it leads to
a higher return on equity (ROE). Some bankers put forth this
reasoning to suggest that higher capital requirements will reduce
shareholder value in banking, but of course it is not so simple.
True, a bank’s ROE will decrease with a decline in leverage, but so

Higher bank capital improves the incentives of
banks to monitor their own borrowers and to
develop stronger long-term relationships, an
outcome that in turn generates economic value.
too will its cost of equity capital (that is, the minimum expected
rate of return demanded by shareholders to compensate for the
decline in risk). So, changes in leverage would have no impact
on bank value or on the size of bank balance sheets.
Moreover, this decline in leverage increases the loss-absorption
buffer to debtholders who can afford to reduce their required
yield, lowering the bank’s cost of funds. Needless to say, with
taxes, an increase in leverage causes ROE to rise faster than the
bank’s equity cost of capital, so shareholder value goes up, ignoring agency costs and other frictions associated with leverage. But
this is nothing more than the debt-tax-shield argument, which
should also apply to nonfinancial firms. The point is that if banks
put more equity capital on their balance sheets, the rate of return
their shareholders require will decrease, and equity will not seem
nearly as expensive.
Opponents of higher capital requirements would also argue
that banks will simply be worth less to their owners if those
owners are forced to post more capital. After all, if deposits cost
3 percent and equity costs 20 percent, would the owners of the
bank not be worse off if they were forced to fund at the margin
with equity rather than deposits? Mehran and Thakor (2011)
expose the theoretical fallacy of this logic, but one may argue that
this is ultimately an empirical question. The empirical evidence
in Mehran and Thakor shows that bank capital and bank value
are actually positively correlated in the cross section of banks.17
17

Mehran and Thakor also examine the endogeneity of bank capital, and suggest
that the association from bank capital to bank value is causal.

Chart 2

Mean Book Equity Ratios for U.S. Banks: 1893-2010
Percent
25
20
15
10
5
0
1893 1903 1913 1923 1933 1943 1953 1963 1973 1983 1993 2003
Sources: Data through 2001 are from Flannery and Rangan (2008); data since 2001
are from the Federal Reserve’s Y-9C reports.
Note: Data since 2001 are average quarterly capital ratios for the 100 largest bank
holding companies.

That is, banks with more equity capital:
• generate higher net present value for their shareholders (the
value created for shareholders over and above what they
invested in the bank is higher when the shareholders invest
more capital in the bank),
• are acquired at higher prices in mergers,
• are paid more in goodwill in the acquisition price, and
• experience higher total (enterprise) values (debt plus equity).
Mehran and Thakor argue that higher bank capital is good
not only for greater safety and soundness of the banking system,
but also for the banks themselves. Higher bank capital improves
the incentives of banks to monitor their own borrowers and to
develop stronger long-term relationships, an outcome that in turn
generates economic value.

Calibration of Capital Requirements
The answer to the question of how much capital banks should
hold is invariably tied to the return distribution of the bank’s
assets, both on balance sheet (actual) as well as off balance sheet
(contingent). To determine how much capital is required, one has
to know how risky the assets are. Since bank balance sheets are
relatively opaque (Morgan 2002), banks are especially susceptible
to the “asset-substitution” problem.
Just how opaque and full of surprises bank balance sheets
can be was highlighted during the recent financial crisis, with the
rather slow recognition of subprime risk hidden in the plethora
of complex structured credit products. The resulting uncertainty
about the precise distribution of returns and contingent assets

www.newyorkfed.org/research/current_issues

CURRENT ISSUES IN ECONOMICS AND FINANCE v Volume 18, Number 4

and liabilities introduces significant “model risk” into any regulatory calibration of bank capital requirements.18
Further, opaqueness of bank balance sheets, combined with
the structural incentives for banks to benefit strategically from
the opaqueness, can create more tail risk for bank asset return
distributions.19 That is, these distributions are both more
complex (less normal) as well as harder for outsiders to estimate,
making it more difficult for debt to perform its monitoring and
disciplining role. At any rate, banks are thinly capitalized compared with other industries, so the margin of error around capital
adequacy needs to be quite small.
Given these considerations, which are only exacerbated by
distortions introduced through access to the safety net (deposit insurance and lender of last resort), our view is that a sensible policy
path is to put a premium on robustness of calibrations along two
dimensions.
First, one should develop and apply several different estimates
of capital adequacy and develop appropriate loss-absorption
mechanisms to help address the distortions. Assessments of
capital adequacy can be based on different ways of estimating
asset quality and risk, such as a set of regulatory risk-weighting
schemes along the lines of Basel III, plus stress tests along the
lines of the SCAP, as well as market measures of systemic risk
based on credit default swap spreads, equity returns, and volatility (for instance, as proposed by Acharya, Pedersen, Philippon,
and Richardson [2010] and Brownlees and Engle [2010]).20
This is the “belt and suspenders” approach, which calls for some
redundancy in the number of ways in which capital adequacy is
assessed.
The special capital account could provide the second margin
of safety in the calculation of capital adequacy—a buffer for the
regulator’s own model risk. This margin is necessary because
opaque balance sheets, contingent exposures that are off balance
sheet, and fat-tailed asset-return distributions all make it likely
that calculations of the needed capital buffer will be imprecise.
Moreover, the possibility of contagion and the thin capital cushions of banks make this buffer more of an imperative.
Capital ratios for banks have been increasing since their recovery from the shock of the financial crisis. Prior to the introduction of deposit insurance, bank capital ratios were quite volatile
and at times very high, fluctuating between 10 percent and more
18 A risk-sensitive capital regime presumes a model of

riskiness of the banks’
assets and activities; these are called “risk weights.” Though carefully chosen,
these weights could turn out to be wrong—meaning the model of bank asset
riskiness could be wrong.
19 See Rajan (2006) and Acharya, Cooley, Richardson, and Walter (2010).
20 To be sure, only stress tests have the potential for taking systemic risks into

account based on granular asset-level data. Current regulatory risk weights on
assets and internal risk-weighting models do not account for systemic risk.

than 20 percent between the end of the nineteenth century and
the Great Depression (Chart 2). The period after World War II
saw capital ratios hovering steadily at around 6 percent, increasing only after the introduction of the Federal Deposit Insurance
Corporation Improvement Act in 1991 (fully implemented in
1993). Currently, the debate continues over what the regulatory
minimum capital ratio will be.

Conclusion
This article examines the important issue of banks’ choices of
privately optimal capital structures, the circumstances under
which regulators would be imprudent to rely on those choices,
and the optimal design of capital regulation. To that end, we also
propose a novel capital framework for banks, based on Acharya,
Mehran, and Thakor (2010), comprising two types of capital
requirements. The first is a regular tier-one capital requirement
that would help deter excessive risk-taking incentives. The second
is a special capital account that would also limit risk taking, but
would ensure that creditors’ disciplining incentives are preserved.
In particular, the special capital account would belong to the
bank’s shareholders when the bank is solvent, but would go to the
regulators—rather than the bank’s creditors—if the bank fails.
The capital requirement proposed here is robust in the sense
that it could simultaneously accomplish four goals. The first is
to bring more capital into banking and hence contribute to the
safety and soundness of the financial sector—without necessarily requiring banks to issue new equity. The second is to improve
bank incentives to reduce the probability of a crisis, rather than
focusing on what to do when a crisis occurs. The third goal is
to accomplish all of this without diluting the market discipline
provided by uninsured debt. In this respect, our proposal differs
from those that would simply infuse banks with more equity
through higher minimum capital requirements. And the fourth
is to do this in the simplest manner possible, using well-known
instruments (equity and retained earnings to build up equity)
rather than new instruments whose pricing characteristics and
market impact may be hard to gauge.

References
Acharya, Viral V. 2009. “A Theory of Systemic Risk and Design of Prudential Bank
Regulation.” Journal of Financial Stability 5, no. 3 (September): 224-55.
Acharya, Viral V., Thomas Cooley, Matthew Richardson, and Ingo Walter. 2010.
“Manufacturing Tail Risk: A Perspective on the Financial Crisis of 2007-09.”
Foundations and Trends in Finance 4, no. 4: 247-325.
Acharya, Viral V., Itamar Drechsler, and Philipp Schnabl. 2011. “A Pyrrhic
Victory? Bank Bailouts and Sovereign Credit Risk.” Unpublished paper, New York
University, August.
Acharya, Viral V., Hamid Mehran, and Anjan Thakor. 2010. “Caught between
Scylla and Charybdis? Regulating Bank Leverage When There Is Rent Seeking
and Risk Shifting.” Federal Reserve Bank of New York Staff Reports, no. 469,
September; revised April 2011.

Acharya, Viral V., Lasse Pedersen, Thomas Philippon, and Matthew Richardson.
2010. “Measuring Systemic Risk.” Unpublished paper, New York University, May.
Acharya, Viral V., João Santos, and Tanju Yorulmazer. 2010. “Systemic Risk and
Deposit Insurance Premiums.” Federal Reserve Bank of New York Economic
Policy Review 16, no. 1 (August): 89-99.
Acharya, Viral V., and Anjan Thakor. 2010. “The Dark Side of Liquidity Creation:
Leverage-Induced Systemic Risk and the Lender of Last Resort.” Unpublished
paper, New York University.
Acharya, Viral V., and Tanju Yorulmazer. 2007. “Too Many to Fail: An Analysis of
Time-Inconsistency in Bank Closure Policies.” Journal of Financial Intermedia
tion 16, no. 1 (January): 1-31.
Admati, Anat R., Peter M. DeMarzo, Martin F. Hellwig, and Paul Pfleiderer. 2010.
“Fallacies, Irrelevant Facts, and Myths in the Discussion of Capital Regulation:
Why Bank Equity Is Not Expensive.” Stanford University Working Paper no. 86,
August. Available at http://www.gsb.stanford.edu/news/research/admati.etal.html.

Jensen, Michael C., and William H. Meckling. 1976. “Theory of the Firm:
Managerial Behavior, Agency Costs, and Ownership Structure.” Journal of
Financial Economics 3, no. 4 (October): 305-60.
Kuritzkes, Andrew, Til Schuermann, and Scott M. Weiner. 2005. “Deposit
Insurance and Risk Management of the U.S. Banking System: What Is the
Loss Distribution Faced by the FDIC?” Journal of Financial Services
Research 27, no. 3 (September): 217-43.
Mehran, Hamid, and Anjan V. Thakor. 2011. “Bank Capital and Value in the
Cross-Section.” Review of Financial Studies 24, no. 4 (April): 1019-67.
Merton, Robert C. 1977. “An Analytic Derivation of the Cost of Deposit Insurance
and Loan Guarantees: An Application of Modern Option Pricing Theory.” Journal
of Banking and Finance 1, no. 1 (June): 3-11.
Modigliani, Franco, and Merton H. Miller. 1958. “The Cost of Capital, Corporation
Finance, and the Theory of Investment.” American Economic Review 48, no. 3
(June): 261-97.

Akerlof, George A., and Paul M. Romer. 1993. “Looting: The Economic Underworld
of Bankruptcy for Profit.” Brookings Papers on Economic Activity, no. 2: 1-74.

Morgan, Donald P. 2002. “Rating Banks: Risk and Uncertainty in an Opaque
Industry.” American Economic Review 92, no. 4 (September): 874-88.

Brownlees, Christian, and Robert Engle. 2010. “Volatility, Correlation, and Tails for
Systemic Risk Measurement.” Unpublished paper, New York University, May.

Myers, Stewart, and Nicholas Majluf. 1984. “Corporate Financing and Investment
Decisions When Firms Have Information That Investors Do Not Have.” Journal
of Financial Economics 13, no. 2 (June): 187-221.

Calomiris, Charles, and Charles Kahn. 1991. “The Role of Demandable Debt
in Structuring Optimal Banking Arrangements.” American Economic Review 81,
no. 3 (June): 497-513.

Myers, Stewart, and Raghuram G. Rajan. 1998. “The Paradox of Liquidity.”
Quarterly Journal of Economics 113, no. 3 (August): 733-71.

Diamond, Douglas W., and Raghuram G. Rajan. 2000. “A Theory of Bank Capital.”
Journal of Finance 55, no. 6 (December): 2431-65.

Rajan, Raghuram G. 2006. “Has Finance Made the World Riskier?” European
Financial Management 12, no. 4 (September): 499-533.

———. 2001. “Liquidity Risk, Liquidity Creation, and Financial Fragility:
A Theory of Banking.” Journal of Political Economy 109, no. 2 (April): 287-327.

Schuermann, Til, and Kevin J. Stiroh. 2006. “Visible and Hidden Risk Factors
for Banks.” Federal Reserve Bank of New York Staff Reports, no. 252, May.

Farhi, Emmanuel, and Jean Tirole. 2009. “Leverage and the Central Banker’s
Put.” American Economic Review 99, no. 2 (May): 589-93.

Song, Fenghua, and Anjan V. Thakor. 2007. “Relationship Banking, Fragility, and
the Asset-Liability Matching Problem.” Review of Financial Studies 20, no. 6
(November): 2129-77.

Flannery, Mark J., and Kasturip P. Rangan. 2008. “What Caused the Bank Capital
Build-up of the 1990s?” Review of Finance 12, no. 2: 391-429.

About the Authors
Viral Acharya is the C. V. Starr Professor of Economics at New York University; Hamid Mehran is an assistant vice president at the
Federal Reserve Bank of New York; Til Schuermann is a partner at Oliver, Wyman & Company; Anjan Thakor is the John E. Simon
Professor of Finance at Washington University in St. Louis.
Current Issues in Economics and Finance is published by the Research and Statistics Group of the Federal Reserve Bank of New York.
Linda Goldberg, Erica L. Groshen, and Thomas Klitgaard are the editors.
Editorial Staff: Valerie LaPorte, Mike De Mott, Michelle Bailer, Karen Carter, Anna Snider
Production: Carol Perlmutter, David Rosenberg, Jane Urry
Subscriptions to Current Issues are free. Send an e-mail to Research.Publications@ny.frb.org or write to the Publications Function,
Federal Reserve Bank of New York, 33 Liberty Street, New York, N.Y. 10045-0001. Back issues of Current Issues are available
at http://www.newyorkfed.org/research/current_issues/.

The views expressed in this article are those of the authors and do not necessarily reflect the position
of the Federal Reserve Bank of New York or the Federal Reserve System.

CURRENT ISSUES IN ECONOMICS AND FINANCE v Volume 18, Number 4

Follow Us on Twitter!
The Research Group recently launched a Twitter feed, designed to offer the first word on developments the Group,
such as:
• new publications and blog posts,
• updates on economists’ work and speaking engagements,
• postings of key indexes and data,
• media coverage of the Group’s work, and
• other news of interest to website visitors.

Full text of Current Issues in Economics and Finance : Robust Capital Regulation, Volume 18, Issue 4

FRASER