199
Myerson: Rethinking the Principles of Bank Regulation
lend their money to the bank at low interest
rates.
Notice, however, that this reassurance
was based on two distinct effects of equity.
For any given level of risk in the bank’s
investments, the probability of investment
losses large enough to affect the depositors
becomes smaller when the bank has more
equity. But more equity also means that the
owners who control the bank have more
incentive to avoid risks of such large losses.
That is, equity helps to solve moral hazard
in banking.
Of course, a large bank may have many
small shareholders who are not actively
involved in the bank’s management, but the
term equity denotes the equal proportionate
sharing of profits among all owners, includ-
ing large investors who can actively oversee
management. Plus, any systematic failure to
serve equity interests could open the possi-
bility of a take-over bid. So it is not unreason-
able to identify the interests of a bank with
the interests of its equity owners.
The waves of bank runs that brought on
the Great Depression in the 1930s led to
the creation of government deposit insur-
ance programs in America and elsewhere.
Deposit insurance effectively responded to
the possibility that widespread fears of bank
failure could create a self-fulfilling prophecy,
as depositors would run to withdraw their
deposits before the bank failed because of
an inability to quickly liquidate all its invest-
ments. With deposit insurance, however,
depositors became less concerned about
how much equity their bank had in its total
investments, and so banks could borrow at
low interest rates even with less equity. But
when creditors are publicly insured, bank
default risk becomes a public problem. Thus,
the requirement that banks should have ade-
quate equity has become a responsibility of
public regulators.
But now we come to the critical point of
confusion that Admati and Hellwig compare
to Hans Christian Anderson’s story of a
deceived emperor whose lack of clothes
went unacknowledged. When individual
depositors bore the risks of bank failure,
people could see clearly enough that they
should not deposit funds in a bank with too
little equity. But when the responsibility for
insuring depositors against bank failures has
been transferred to the government, then
equity requirements can be portrayed as a
technical regulatory subject that only experts
can understand, and banks can hire the most
experts to address this subject.
Persuasive voices in regulatory debates
have argued that equity is expensive for
banks, and that requiring more equity would
increase their costs of lending and cause a
decline of economic investment and growth.
With this logic, the levels of bank equity have
been allowed to decline over the twentieth
century until, in the run-up to the financial
crisis of 2008, many large banks in Europe
and America could finance their assets with
3 percent or less equity. When losses put this
thin layer of equity at risk, the result was a
global financial meltdown.
But like the boy who finally observed that
the emperor had no clothes, Admati and
Hellwig offer straightforward and convincing
testimony to contradict all these arguments
for minimizing bank equity. Complicated
arguments call for complicated counterar-
guments, but Admati and Hellwig use their
expertise in economic analysis to identify the
simplest rebuttal.
A key point in this discussion is the famous
insight of Modigliani and Miller (1958) that,
under some basic assumptions, the sum of
the total value of a corporation’s equity to its
owners plus the total value of its debt to its
creditors should be equal to the total value of
its income-earning assets.
1
This result tells us
1
Standard accounting estimates of equity value are cal-
culated to make this equation an identity, but Modigliani
and Miller’s argument applies to actual market values,
Journal of Economic Literature, Vol. LII (March 2014)
200
that the total cost of issuing debt and equity
to finance any given portfolio of investments
should not depend on how much is debt
and how much is equity. Under these basic
assumptions, then, changing equity require-
ments should not affect the total cost of bank
lending at all. Modigliani and Miller tell us
that, at least for banks that are organized
as corporations, profitable new investments
could be financed by selling new shares of
stock in the bank as well as by borrowing
more money.
A good economic theorist can find plenty
of exceptions to the simplifying assumptions
of this Modigliani–Miller theorem. But the
primary reason the cost of financing invest-
ments can sometimes depend on the mix
of debt and equity is that, at times, another
party bears the difference in costs. In par-
ticular, when the government is (implicitly
or explicitly) insuring a bank’s creditors,
then increasing the fraction of debt financ-
ing can increase the value of this insurance
from the government. In this case, increas-
ing debt may make the bank’s owners bet-
ter off at no cost to the creditors, but the
bank’s gains would be at the expense of the
tax-paying public, which is bearing risks that
private investors would not accept without
being paid a greater interest premium. Then,
if tighter equity requirements would prevent
the bank from making some investments,
these would only be investments that private
investors would have refused without subsi-
dized government insurance.
Thus, Admati and Hellwig argue, although
equity may indeed seem expensive to a bank,
it is only because public insurance enables
them to borrow at low rates that do not
properly respond to the greater risks that
their creditors must bear when the bank has
less equity. When the cost to the public of
showing that an arbitrage opportunity would exist if two
firms with the same prospective income streams had dif-
ferent total values of their equity and debt.
providing this insurance is properly taken
into account, the total social cost of the
investment is not increased by requiring
more of it to be financed by equity.
There are other exceptions to the
Modigliani–Miller theorem that are worth
mentioning. Myers and Majluf (1984)
argued that, when the managers of a firm
decide whether to finance new investments
by borrowing money or by selling new equity
shares at any given price in the stock market,
the managers will be more likely to choose to
sell new equity when they have adverse pri-
vate information that suggests that the mar-
ket has overvalued their stock. So the public
may rationally take a firm’s decision to issue
new equity as bad news about the firm, and
the price of its stock should be reduced to
take account of this adverse inference. This
Myers–Majluf effect can indeed make equity
financing more expensive than debt for the
current owners of the firm. But notice that
this adverse inference occurs only when a
firm’s decision to raise capital by selling new
equity shares depends on its managers’ pri-
vate information about the profitability of
their business. The Myers–Majluf adverse
inference would not apply when a regulator
requires the firm to sell new equity shares
based on information that is publicly avail-
able. Thus, a bank’s cost of raising new
equity may actually be less when new shares
are issued under a transparent regulatory
requirement than if the same shares were
issued by a discretionary decision of its man-
agement. This point by itself may be a good
reason for regulating banks’ equity based on
transparent public information.
Taxes can provide one other basic reason
why the profitability of investments could
depend on the mix of debt and equity that
are used to finance them, if the tax laws treat
interest payments to creditors differently
from dividend payments to shareholders.
Admati and Hellwig urge reform of current
tax rules that tend to subsidize debt over