205
Myerson: Rethinking the Principles of Bank Regulation
such questions. Admati and Hellwig offer
a shocking perspective on how the bal-
ance sheet of JPMorgan Chase Bank on
December 31, 2011, would be affected by
a change of accounting standards from the
generally accepted accounting principles
(GAAP) used in the United States to the
international financial reporting standards
(IFRS) used in the European Union. Under
both accounting standards, the bank’s net
equity value could be reported as $184 bil-
lion. But under the IFRS standards, the bank
would have to report $1.79 trillion more of
assets and liabilities from swaps and other
derivative trades that could be netted out
under the GAAP standards. This discrep-
ancy is almost ten times larger than the total
equity of the bank. Under GAAP standards,
JPMorgan Chase Bank could claim that its
equity value was about 8 percent of assets,
but under IFRS standards this fraction
would shrink to a mere 4.5 percent.
When such technical details of account-
ing can make such a great difference to the
representation of a well-known bank, one
may be tempted to give up and concede
that it is all too complex for the layman
to hope to understand. This has to be the
wrong answer. Banks are in the business of
issuing deposits that people can count on as
safe. The process of assuring depositors and
taxpayers (if they offer deposit insurance)
that these deposits are safe must involve
some form of communication that the gen-
eral public can understand. Of course, the
average depositor or taxpayer wants profes-
sional regulators to do the work of moni-
toring our banks’ balance sheets, and we
want our elected officials of government to
make sure that the professional regulators
do their jobs. But with so much money in
the banking system, our hope that regula-
tors and officials will not be induced to
falsely certify unsafe banks must depend
on confidence that a failure of appropriate
regulation could be discovered, reported
in the press, and understood by voters well
enough to cause a ruinous scandal for the
responsible officials.
We want our country to have a banking
system that can reliably channel millions of
people’s savings into credit for investments
in our growing economy. But this requires
that the banking system be able to main-
tain broad public trust, so we need a bank
regulation system that operates according
to reliable principles of accounting that can
give people some sense of the risks that their
banks have undertaken. If there are abstruse
financial transactions that generate risks
which cannot be adequately represented
in standard public accounting statements,
then perhaps such transactions should be
off limits for banks that are in the business
of issuing reliably safe deposits. Otherwise,
there is a danger that regulated banks may
be attracted to some transactions precisely
because they enable the banks to hide risks
off of their balance sheets. If society needs
marketmakers for such abstruse transactions,
then perhaps we should let other specialized
corporations fill that role.
Many forms of creative accounting have
been devised to keep assets and liabilities
off of balance sheets. But one advantage of
basing capital requirements on liabilities
rather than assets is that it can be easier to
motivate full reporting of liabilities on the
balance sheet when regulators have the right
to restrict a distressed bank’s payouts on any
unreported liabilities. Off-balance-sheet
transactions only work for the bank if it can
use its good credit to reassure the parties
who hold the bank’s liabilities. By accepting
the transaction in a form that the bank can
keep off its regulated balance sheet, however,
those parties are undermining the reliability
of the regulatory system for others, even
though they want to rely on it themselves.
It may be appropriate, then, to deter such
investors from funding
off-balance-sheet
liabilities by a threat that a regulator could
Journal of Economic Literature, Vol. LII (March 2014)
206
prevent them from getting their funds when
they want them.
4
5. Dangers of Risk-Weighting Assets in
Capital Adequacy Formulas
In the work of the Basel Committee on
Banking Supervision to coordinate interna-
tional standards for bank regulation, a major
theme was the effort to make equity capital
requirements more finely calibrated to the
risks of the bank’s investments. The effort
to define the minimal equity requirement
as precisely as possible would make sense
if there were some substantial social cost to
requiring banks to have more equity than
they need. But Admati and Hellwig have
cogently demonstrated that, from the per-
spective of society as a whole, bank equity
is not a particularly costly source of funds
for economic investments. Indeed, Admati
and Hellwig lead us to ask whether the
entire Basel-led attempt to fine-tune capi-
tal requirements may have done more harm
than good.
The first Basel standards from 1988 called
for a bank’s equity capital to be at least 8
percent of its total risk-weighted assets.
Risk weights were introduced into this cal-
culation to reduce the required equity for
a bank that made safer investments. Assets
that were considered safe were given small
risk weights of less than one. Then the
required equity under this system would
be computed by multiplying the value of
each asset by its appropriate risk weight,
summing these weighted asset values, and
4
Such a provision for regulators to block payments on
a distressed bank’s unreported liabilities could require a
more conservative form of regulatory accounting for deriv-
ative contracts, each with a liability equal to the maximal
potential cost of the contract for the bank in a worst-case
scenario, and with a corresponding asset value equal to the
surplus of the contract’s current expected (mark-to-mar-
ket) value for the bank over this worst-case scenario. The
reported maximal liability would have to be clearly stated
also to the counterparty in the contract.
then requiring equity to be worth at least 8
percent of the total. At the low end, loans
to sovereign governments of high-income
nations could even get risk weights of zero,
and so this system would allow a bank to
make arbitrarily large investments in sov-
ereign debt financed entirely by borrowing.
In later Basel standards, the concept of risk-
weights was refined by more sophisticated
formulas that could be based on the bank’s
own models of risk.
Presumably, the original idea behind the
risk-weighted capital adequacy formulas was
to encourage banks to make safer invest-
ments by offering to relax equity require-
ments when a bank made investments that
seemed safe. But this idea should have been
considered a very questionable way to pro-
vide such an incentive, as the basic reason for
requiring some minimal fraction of equity in
a bank is to guarantee that the bank’s own-
ers have a substantial incentive to invest
prudently. Worse (as critically observed by
Haldane and Madouros 2012), we should
recognize that any such attempt to codify
what kinds of investments should be con-
sidered safe by regulators can itself create
serious systemic risk for the entire financial
system when a class of assets turns out to
have been incorrectly categorized as “safe.”
We may assume that regulators, being
sophisticated professionals, are often right in
their evaluations of what is safe, but they will
be wrong occasionally. When that happens,
normal investors’ concerns about risk may
make them less willing to hold these assets
that have been wrongly categorized as “safe,”
so the expected rates of return on these assets
may begin to increase. Then banks through-
out the financial system will be tempted to
borrow at low rates of interest and earn sur-
plus profits from the higher rates of return on
these officially “safe” assets. The temptation
to do so increases when the amount invested
becomes large enough that it could cause the
bank to fail if the officially unrecognized risk