201
Myerson: Rethinking the Principles of Bank Regulation
equity by treating interest payments more
favorably than dividend payments. Such
provisions in the tax code increase the fra-
gility of the financial system by encouraging
corporations to use more debt instead of
equity in financing investments. This effect
is especially perverse when applied to bank-
ing, where capital regulation then must act
to reverse it.
When equity is too small, the possibility of
default can begin to distort investment deci-
sions in the bank. When evaluating a poten-
tial new investment, the bank’s owners would
put no value on any prospective returns from
the investment that would accrue after the
event of the bank’s failure, as any returns
in this event would only benefit the bank’s
creditors. Thus, the possibility of default
can reduce the bank’s willingness to make
productive investments from retained earn-
ings. Conversely, the bank’s willingness to
take risks with borrowed funds may increase.
Once there is some serious possibility of the
bank’s equity being wiped out by losses in its
investments, any additional losses will not
affect the bank owners, since they cannot
get less than zero from their equity shares
in the bank. Additional losses would then be
borne by the creditors (or by the taxpayers
who guarantee the creditors), but the owners
who control the bank have no incentive to
take these downside risks into account. Thus,
a bank with insufficient equity has distorted
incentives that can encourage its owners to
take excessive risks or gamble for resurrec-
tion (as in “heads our equity value rises, tails
the creditors lose more”).
Admati and Hellwig call these distorted
incentives for underinvestment and exces-
sive risk taking the “dark side of borrowing.”
Notice that the problem of equity owners
undervaluing prospective returns in the
event of default does not depend on deposit
insurance; it can arise in any corporation with
a given burden of debt that is large enough
to create a substantial possibility of default
(see Myers 1977). Deposit insurance can
exacerbate the problem, however, by making
the cost of new borrowing insensitive to pro-
spective returns in the event of default.
Unfortunately, when the time comes that
investment losses begin to deplete the bank’s
equity, the bank may not have the right
incentives to raise new equity. If the bank
were to increase its capital by selling new
equity shares, the result would be to transfer
to the owners some of the downside risks that
are being borne by the creditors, so the net
value of the existing owners’ shares would
be decreased. Thus, as Admati and Hellwig
emphasize, borrowing can become addic-
tive, as the primary advantages of increas-
ing the ratio of equity to debt would accrue
to the creditors, but not to the owners who
control the bank. The bank’s long-term abil-
ity to attract deposits may depend on a com-
mitment to force the bank to increase equity
when it becomes too small. Thus, for long-
term sustainability, a bank needs regulation.
3. Capital, Liquidity, and Lenders of
Last Resort
Before the 1980s, bank regulation focused
more on liquidity reserves than on equity
capital. Gorton (2012) has expressed some
concern about the focus on capital require-
ments, rather than liquidity requirements.
But Admati and Hellwig, while criticizing
the parameters of current capital regulation,
support the focus on equity capital itself. To
see the logic of this position, we must con-
sider the basic function of a central bank in
the economy. The basic idea is that a central
bank should be able to solve general liquid-
ity problems in the banking system as long
as banks have adequate capital or equity.
Let me try, in this section, to enlarge on this
point so as to indicate the fundamental infor-
mational nature of the problem.
The primary liquidity requirement for
banks has been that they should hold some
Journal of Economic Literature, Vol. LII (March 2014)
202
fraction of cash or safe government bonds
against their deposit liabilities. A bank’s cash
reserve does not have any logical relationship
with its equity value. On the balance sheet,
cash is just one asset in which a bank can
invest—one that typically has a low expected
return, as it generally pays little or no inter-
est. As legal tender, however, cash is the
most liquid store of value, the one that can
be immediately used to pay out on the bank’s
obligations. Other investments with higher
expected rates of return must typically be
held over some period of time to yield their
returns. But if a bank holds good long-term
investments, then it should generally be able
to sell some portion of these investments to
other investors when it needs more cash to
meet its obligations. To understand why illi-
quidity could sometimes be a problem, then,
we need to understand why a bank might
sometimes be unable to quickly liquidate
some investments when it needs cash.
This illiquidity problem could occur if
the bank has special expertise in evaluat-
ing its long-term investments. Then, when
it tries to sell long-term investments early,
other investors might infer that the bank
got bad news about the prospective future
returns from these investments, so the bank
with special expertise may find that it can
only sell its investments early at a value
much lower than what they would expect to
earn at maturity. That is, liquidity problems
may be fundamentally informational, but
informational problems are why we need
a banking system in the first place. People
lend their savings to banks and other finan-
cial institutions for investment because the
banks have better information about how to
find good investments.
Thus, liquidity for long-term economic
investments can depend on the existence
of other financial institutions that have the
ability to evaluate the investments’ qual-
ity and can bid competitively to buy them.
The critical macroeconomic problem is the
possibility of a general panic in which there is
a loss of confidence in such financial institu-
tions. In case of such a panic, there would be
a general surplus of good quality investments
offered for sale at a loss. But then, a “lender
of last resort” should be able to buy up these
discounted assets and expect some positive
long-term profit in providing short-term
liquidity to the beleaguered banks.
This is the basic logic of central bank-
ing, where the central bank is the financial
system’s lender of last resort. The classic
prescription of Bagehot (1873) is that, in a
financial crisis when banks are generally
short of cash, the central bank should stand
ready to lend unlimited amounts (at a high
interest rate), but only to solvent banks that
have good collateral worth more than their
liabilities. To do its job, then, the central
bank needs more than just a deep ability to
issue more money when necessary; it also
needs some basic expertise in evaluating
financial investments, so that it can identify
when a bank’s investments can be taken as
good collateral.
2
In this sense, a lender of last
resort may be better understood as a moni-
tor of last resort, and its bold lending then
serves as public signal that the borrowing
banks have been found creditworthy by the
experts at the central bank. Thus, the cen-
tral bank should be able to solve the liquid-
ity problem for banks, but only when they
are solvent, where solvency means that the
value of the bank’s equity is positive. From
this perspective, we see that capital regula-
tion is fundamental to liquidity, but also that
a central bank’s ability to apply stress tests to
evaluate the solvency of banks is as essential
to the central bank’s function as its ability to
print money.
2
This expertise must be complemented (perhaps in
another agency) by reliable expertise in the resolution of
insolvent banks; that is, in selling or reorganizing their
assets and operations to achieve the greatest possible value.