The future role of the International Monetary Fund (IMF) is today in
doubt. Former Treasury Secretaries George Shultz and William Simon
have urged that it be closed. President Clinton wants the IMF to
devote more attention to preventing crises rather than responding to
them. Even the IMF criticized its own recent operations in Asia for
protecting foreign lenders at great cost to borrowing countries and
their citizens. Protestors in the Asian countries and elsewhere
complain that the IMF is a lackey of the United States, doing the
U.S. government's bidding to the detriment of local populations.

Before agreeing to provide more money for the IMF as part of the 1998
budget agreement, Congress insisted on greater transparency in
decision making and higher interest rates on IMF loans. These changes
are first steps toward reform of international lending institutions.
Still, more fundamental reforms are needed to reduce the risks of
destabilizing crises that have become more frequent and more costly.
Even the losses suffered by bank depositors during the Great
Depression pale by comparison to recent losses in Mexico, South
America and Asia.

Five factors go a long way toward explaining why there have been so
many large financial and foreign exchange crises in developing
countries during the current period of sustained growth and
development in the world economy: weak banks; government interference
and direction of lending (part of crony capitalism); a large volume
of misdirected bank lending; domestic government and IMF bailouts to
protect foreign lenders and domestic oligarchs; and fixed,
unsustainable exchange rates.

The proposal for reforming the IMF that we shall make in this article
seeks to restore international lending while avoiding the excessive
risk-taking that leads to financial bailouts and severe depressions,
as in Mexico, Thailand, Indonesia, Korea and Russia. The new IMF
would avoid both the problem of excessive risk-taking, followed by
collapse, and the risk of a protracted reduction in capital flows to
developing countries. The challenge is to reduce the costs of the
present system while retaining the benefits for economic development
of international lending and capital movements.

A Record of Failure

The IMF and the World Bank were created at the end of World War II to
foster long-term economic growth and stability in an environment of
weak international capital markets. The World Bank's role was to
boost capital flows to promote long-term growth. The IMF's role was
to provide short-term assistance to facilitate the maintenance of
fixed exchange rates. The presumptions underlying the creation of
these Bretton Woods institutions were that, first, countries would
maintain fixed exchange rates tied to gold and the dollar and,
second, that private international capital flows would be rather
modest.

Both presumptions proved to be wrong. The fixed exchange rate system
ended in 1971, when President Nixon devalued the dollar and closed
the gold window. All major currencies--the dollar, yen and
deutschemark--soon began a managed float. And instead of a dearth of
private lending, large-scale lending to developing countries has
coincided with all of the major financial crises of the past twenty
years.

Why have these institutions, intended to foster growth and promote
stability, failed so badly in the Eighties and Nineties? Many reasons
have been offered. Two aspects of private financial arrangements are
of central importance: the form of international capital flows and
the structure of domestic banking systems in emerging market
economies.

Under current arrangements, corporations and bankers in the
developing countries borrow from financial institutions and markets
abroad. The loans are made at fixed exchange rates and denominated in
dollars, marks or yen, so the lender is paid in his own currency and
the borrower therefore bears the full risk of devaluation.

At present, banking systems in many developing countries are poorly
capitalized and, therefore, unable to withstand heavy withdrawals.
Bank depositors and stockholders are protected against loss by local
governments. Banks often act either as agents of their government's
development plans, or as captive financial arms of local industrial
firms, lending at below-market rates to favored enterprises without
careful screening for credit worthiness. Unlike prudent lenders, they
do not diversify loans over borrowers in many different industries.

When problems arise in a developing country or the world economy as a
whole, the financial position of banks in emerging market economies
weakens. Because governments protect domestic banks from failure,
banking losses become a fiscal burden on government and, therefore,
on domestic taxpayers. To pay for these losses, governments borrow
more from domestic and foreign lenders. The additional foreign
borrowing strains their ability to repay, increasing the risk of
devaluation, default and a foreign exchange crisis. Instead of
renewing or increasing short-term loans, some foreign banks demand
repayment in their own currency, further draining the borrowing
country's reserves of dollars, marks and yen. Other lenders, seeing
the loss of reserves, also demand repayment at the fixed exchange
rate. As foreign reserves decline and the country can no longer honor
its commitment to repay foreign borrowers at a fixed exchange rate,
it must default, resulting in devaluation and a currency crisis. The
crisis deepens the insolvency of domestic banks that have borrowed
abroad in foreign currency and have assets priced in domestic
currency. The currency and the weak domestic banking system collapse.
The economy goes into recession, or deep depression, triggering
additional bankruptcies, inability to repay domestic banks, and thus
more bank failures and defaults on foreign loans.

To prevent such defaults, the IMF has taken on the role of lending to
governments of developing countries in times of crisis. Much of the
money that the IMF supplies is used to pay off foreign banks and to
maintain the appearance of solvency at domestic financial
institutions. Local taxpayers must repay these debts to the IMF in
the future, so the banks' rescue is also at taxpayers' expense. The
countries are often left in deep depression. Mexico in 1995 and
Thailand and Korea in 1998 are the clearest examples.

Banking system insolvencies and improper government policies, not
unwarranted speculative attacks on exchange rates, are among the
principal reasons behind currency instability and financial failure.
The IMF adds to the problem by fostering the belief that it will bail
out the banks, however imprudent or insolvent they may be. The
ultimate cost is then borne by local taxpayers.

The Orchestrator of Bailouts

Prior to 1987, the IMF would not lend until borrowers worked out
agreements with private foreign creditors. This forced debtors to
negotiate in good faith with their creditors. Since 1987, the IMF has
often been a lender not of last but of first resort, offering loans
before private debtors and creditors reach agreement.

Why has the IMF undertaken the role of bailout agency for
international lenders? First, the IMF is not independent of its
member governments. Rather, its decisions are the direct result of
votes by its member governments' representatives, and voting power is
concentrated in the hands of a few nations. Accounting for 18 percent
of these votes, the U.S. Treasury has used its considerable power to
push through some IMF programs even over the objections of senior
staff and country experts within the IMF. Thus, political objectives
of one or more powerful members, rather than sound economic
reasoning, often guide IMF intervention.

Second, to the extent that IMF staff are able to determine policy,
they do not represent a reliable source of independent judgment.
Indeed, many espouse the views of the borrowing governments they
monitor. One reason is that their performance and promotion are much
affected by the quality of their relations with foreign officials.
Unfriendly actions by IMF staff members restrict their access to
these officials. The finance minister or central bank president
becomes "unavailable." Foreign officials use such subtle pressures to
restrict criticism and avoid unwanted recommendations for reform.

One consequence of the IMF's new role as orchestrator of bailouts is
that its programs are now much larger than before. Since foreign
banks did not suffer losses in Mexico, they did not believe they were
taking big risks in Asia and Russia. Bankers reasoned that if Mexico
was important enough to the U.S. Treasury and the IMF that the banks
had to be spared, Korea and Russia were at least as important.

In December 1997, for example, a prominent emerging-market
investments newsletter told its clients that anticipated IMF
protection would promote continuing inflows of funds to Brazil,
despite its poor fundamentals:

"The combination of increased Japanese capital outflows over the
year, a 'dip-buying' investor psychology which is spreading from U.S.
retail to emerging market investors, and the massive
Asian-crisis-inspired injections of high-powered global money by the
IMF, will combine to ensure a market in which there is tremendous
technical support. Add in the clear moral hazard caused by the IMF
bail-outs--two investors last week told me that they were planning to
put on large Brazilian positions (even though they were very unhappy
with the currency regime) because they were convinced that a
Brazilian crisis would result in an immediate IMF bail-out--and it is
hard to see why fundamentals should matter [emphasis added]."

Many lenders believed that the 50 or 100 percent rate of interest on
loans to Russia prior to its crisis was close to a free ride because
the United States and the IMF would not let Russia default. We now
know these judgments were wrong in the case of Russia. International
lending has declined as lenders have lost money and become much more
cautious. Nevertheless, bailouts elsewhere, and their adverse
consequences for investor behavior, continue.

Pluses and Minuses of a Floating Exchange Rate

One way to reduce the risk of financial collapse is to let exchange
rates float. With fluctuating exchange rates, lenders would either
take the risk of a change in the exchange rates or hedge that risk in
the marketplace. Losses from exchange risk would occur gradually and
openly instead of in the large, one-time adjustments that have
devastated developing countries. The United States, Canada, the
European Union, Japan, Switzerland and many others have adopted this
approach.

Freely floating exchange rates are an entirely feasible and, some
would say, desirable policy. But with floating exchange rates, prices
of imported and exported goods and services change frequently, both
up and down. Many producers, importers and exporters do not want to
be exposed to frequent price changes, so governments often intervene
in exchange markets to smooth these changes. Further, countries
seeking to end inflationary monetary policies often fix their
exchange rate for a time to signal a change in regime. This is a
risky strategy unless the country can convince the market that the
new exchange rate will remain fixed. These intermediate policies,
neither floating nor credibly fixed, often increase country risk and
end in collapse. Economists oppose exchange rate intervention, citing
the long-term benefits for financial stability of a credible policy,
but unfortunately these exhortations have little impact.

What Is To Be Done

1) Rules for Stable Banking

The nation-specific political constraints that govern economic policy
will continue to give rise to two types of countries: those with
relatively sound, well-managed financial systems and those without
them. Rather than continuing to participate in financial bailouts,
the new IMF we envision should maintain incentives that avoid banking
crises in the first place, and thus the accompanying severe losses to
borrowing countries and their citizens. The role of the restructured
international lender that might replace the current IMF would be to
serve as lender of last resort to countries with sound financial systems,
while providing incentives for strengthening financial systems elsewhere.

We propose that a restructured IMF limit membership to countries that
meet certain standards of conduct in their banking systems. The
decision to adopt these standards would be left to each country, but
doing so would be a condition for joining the system.

The standards we have in mind would be simple and easily verifiable.
First, all domestic commercial banks would have to issue part of
their liabilities in credibly uninsured debt. The owners of the debt
would bear the risk of bank default, so they would have to be
non-government entities--preferably foreign banks and institutional
investors. Their function would be to monitor the banks' decisions
and share the risk of failure with owners of bank equity. Hence, they
would have incentives to demand prudent policies.

Current bank equity capital requirements are based on an
international agreement to maintain enough equity to protect
taxpayers (who insure bank deposits) from the consequences of bank
loan losses. In theory, the equity buffer forces owners to bear the
bulk of losses and discourages banks from taking unwarranted risks at
taxpayers' expense. But in practice, bank equity capital requirements
are an inadequate deterrent. Equity capital is not measured
accurately by bank supervisors because it is difficult to foresee
defaults, and supervisors lack economic or political incentives to
identify problems early. Failing to recognize losses and then
subtract them from the owners' capital means that capital is
overstated. Owners of banks that enjoy government deposit insurance
have much to gain and little to lose by increasing bank risk once
capital is impaired. Gains on the risky loans, or gambles in foreign
exchange markets, accrue to them; losses are borne by the government.

Existing bank equity requirements would be strengthened substantially
by requiring banks to finance a minimum proportion of their assets
with uninsured debt. To protect their positions and minimize the risk
to themselves, uninsured debt holders would discourage bank
risk-taking in the wake of losses. Thus market discipline from
uninsured debt would prevent banks from abusing government protection
of deposits more effectively than current equity capital standards.

Second, depositors would continue to be insured, as in fact they are
in almost all countries. Deposit insurance raises some problems, but
its absence raises the much larger problem of bank runs and
destruction of the payments system. Further, explicit deposit
insurance has several advantages, including the opportunity to charge
for the service and strengthen prudential regulation.

A third requirement for IMF membership would be that countries open
their financial markets to competition from abroad. Branches of
foreign banks domiciled in the country would provide competition,
improve standards of performance and train local personnel. In the
event of a problem or crisis, these foreign branches would be
protected by their home offices, so they would contribute to
stability and enhance safety. Further, their domestic loans would be
a small part of a diversified portfolio of loans to many countries.
Such diversification is an effective means of reducing risk.

The three elements of this plan for reforming banking systems are not
novel; some variant of each of them is now accepted practice in
several countries. Chile and Argentina have in place requirements for
uninsured debt finance. A broad consensus, including the Bankers
Roundtable, some Federal Reserve officials, and some members of
Congress, advocates a similar requirement for the United States. Many
Latin American countries have already opened their markets to foreign
banks. Approximately half of Argentine deposits are now held by such
banks, and foreign banks operate successfully in Mexico, Brazil and
elsewhere. Moreover, the World Trade Organization's financial
protocol requires free trade in financial services to be achieved
over the next decade. The implementation of our proposals would
strengthen countries' incentives to open their financial markets
sooner.

Many other rules could be added in the interest of promoting bank
solvency. Our aim is to have few, transparent and verifiable
conditions for membership in a new IMF. We rely on incentives and
competition to lead bankers toward more prudent behavior. Market
discipline provides that incentive and encourages banks and their
debt holders to improve transparency, adopt effective bankruptcy
codes and develop rules for contract enforcement.

Governments can accelerate the process of improvement by adopting
accounting standards that increase transparency and provide uniform
measures of profit and loss. But accounting rules and legal
restrictions are of little benefit if no one has an incentive to use
or enforce them. Market competition is a lever that raises standards
because prudent lenders are more secure and better able to service
their customers without interruption.

Historically, when banks have faced market discipline, they have been
far more resilient in the face of shocks. Banks have responded to
losses by reducing asset risk or raising capital. By increasing their
cash holdings and cutting dividends to stockholders, they have tried
to reassure depositors that bank losses would not result in depositor
losses. When discipline is absent, however, banks have opposite
incentives. Initial losses are followed by increases in bank
risk-taking. Failing U.S. savings and loans in the 1980s, for
example, increased risk-taking. Banks gambled to achieve high profits
but instead took large losses. Losses on risky investments were
shifted to taxpayers via the deposit insurance system. In Japan,
Korea, Thailand, Indonesia, Mexico and elsewhere, risks were
increased both in order to continue supplying credit to borrowers
favored by bankers or the government and to increase profits. In all
cases, failures were borne by the taxpayers, as they were in the
United States.

The taxation of ordinary citizens to pay for bank bailouts can wipe
out the savings of a generation. Losses in excess of 20 percent of
GDP are not uncommon in Asia today nor were they in Latin America
earlier. Japanese bank failures will cost taxpayers 20 to 30 percent
of GDP. Our proposal seeks to eliminate or reduce the size of bank
bailouts. Banking reform is one crucial step. Reform of international
lending is the other.

2) Rules for International Institutions

More than a century ago, a British economic journalist, Walter
Bagehot, set out the classical principles for a central bank acting
as lender of last resort: lend freely in a crisis at a penalty rate
against collateral. Adapted to international lending, Bagehot's rule
is the proper rule for a restructured, more effective IMF.

Adopting Bagehot's rule would require three major changes in IMF
practices. First, until this year the IMF lent at below-market rates
of interest, in effect subsidizing borrowers and encouraging delayed
repayment. We propose that lending be done at a penalty rate; that
is, a rate above the pre-crisis market rate on the borrower's
collateral. A penalty rate encourages the borrower to negotiate with
private creditors to seek (lower) market rates. The IMF would lend
only when there is a liquidity crisis--that is, when private lenders
are unwilling to lend. That is precisely the responsibility that a
lender of last resort should fulfill. If the system functions well,
the new IMF would lend infrequently.

Second, the IMF should start to require collateral to guarantee
repayment. By pledging collateral, the borrower shows that it has
valuable assets, and thereby is not insolvent or bankrupt. The
lender, meanwhile, gets a guarantee of repayment, and it can sell the
collateral in case of future default. Further, requiring collateral
for loans encourages countries to maintain liquid assets to be used
for this purpose in a crisis, thereby reducing the chance that a
banking crisis will occur. Some part of the collateral would consist
of negotiable foreign bonds, while the balance would consist of other
items. For example, to guaranteeits loan from the U.S. Treasury in
1995, the Mexican government pledged its receipts from oil sales.
These receipts were deposited at the New York Federal Reserve Bank
until the debt was repaid. Collateral could also include other
dollar-denominated assets owned by the borrowing country's central
bank.

Third, IMF lending would be restricted to member countries that adopt
the earlier prescribed banking standards.

If it followed these rules, the IMF would not bail out insolvent
banks or banking systems in the guise of protecting the liquidity of
member governments faced with a run on their currency. Unlike the
present system, the IMF would not impose conditions on the borrowing
country, other than membership rules and collateral requirements.
Countries would be free to adopt the economic policies of their
choice, not, as is frequently charged, policies imposed by the IMF
acting as the agent of the U.S. government. Private lenders, knowing
that they would not be bailed out without loss, would have an
incentive to scrutinize more carefully the policies of countries to
which they lend. Borrowing arrangements would be fixed in advance.
Countries would avoid the weeks or months of negotiation during which
the Mexican, Indonesian and Korean crises became more severe and more
costly to local populations.

To finance its lending, member governments would contribute
marketable bonds to the IMF. These bonds could be sold in the market
or to central banks in hard currency countries to fund IMF loans. The
IMF would be allowed to borrow, against collateral, from central
banks in countries with internationally accepted monies. The central
banks would lend risklessly at a market rate against collateral. They
would be free to offset the effect of the borrowing on their own
interest rates and economic activity to avoid any inflationary effect
or conflict with domestic policy.

All currency crises are not banking crises. Indeed, our plan
separates the two by requiring countries to develop and maintain
prudential standards for banking, and it increases reliance on
floating rates to avoid currency crises. As the number of countries
qualifying for the new IMF increases, banking crises would become
less frequent. Currency crises might continue in countries with fixed
exchange rates, but they would be less costly because banking systems
would be much more stable.

3) Other Foreign Assistance

Two other changes would be a necessary complement to our program:
first, Congress should abolish the Exchange Stabilization Fund, a
remnant of the 1930s; and second, the World Bank should be
restructured to concentrate on long-run assistance to spur economic
development. It should not participate in IMF loans or compete with
the IMF as a source of emergency lending, as it has often done.

Originally intended to support the dollar exchange rate after the
1934 dollar devaluation, the Exchange Stabilization Fund has become
an off-budget slush fund that the Treasury uses to make foreign
loans. The appeal of the Stabilization Fund to the Treasury and the
administration is that it enables them to avoid the congressional
appropriation process. They obtain funding, in part, by spending some
of the Stabilization Fund's $25-30 billion of foreign exchange
holdings and, in part, through a complex arrangement called
"warehousing", under which the Treasury borrows directly from the
Federal Reserve to augment the Fund. Closing the Exchange
Stabilization Fund would require the administration to use the normal
congressional budget process. Foreign assistance, like any other
expenditure, could only be proffered with congressional approval and
oversight.

One of the IMF's most costly mistakes was to accept responsibility
for lending to Russia. It had no previous experience and no special
expertise in restructuring a non-market economy. It was unable to
enforce the lending conditions it imposed. And, because it was
committed to "successful transformation", it was reluctant to
withhold its loans.

Transformation lending to Russia took the IMF, with the support and
encouragement of the G-7 governments, far beyond its mandate and experience.
The Russian default is a principal reason for recent world financial
turbulence and the large losses borne by banks and financial
institutions in many countries. A restructured IMF would have been
prohibited by its charter, and also by the conditions of membership,
from lending to Russia. This was foreign aid that should have
required approval by the individual G-7 parliaments. The IMF and
World Bank should not become the means of circumventing parliamentary
oversight and appropriations.

Political Impediments to Reform

Experience shows that, economically, these reforms are feasible. For
over thirty years prior to World War I, market discipline reigned in
banking, and government interventions were typically limited to
liquidity assistance through Bagehotian lenders of last resort. An
integrated global capital market successfully mobilized far more
resources relative to economic activity for use by then-emerging
market economies than today's markets do. Lenders of last resort
operated successfully to stem liquidity crises. Banking crises in
emerging economies were infrequent; banking insolvency was a much
smaller problem; and currency collapses were rare compared to today's
experience.

Politically, however, there are significant impediments to reform.
Four constituencies will likely oppose some or all of our proposed
reforms: banks in developed economies; oligarchs in emerging markets;
IMF bureaucrats; and U.S. Treasury officials (and their counterparts
in other G-7 countries). Each of these groups would lose power,
influence or subsidies.

A possible quid pro quo to secure the support of global bankers is
the removal of barriers to the entry into new markets (which our plan
would require), and the expansion of their powers domestically and
abroad in exchange for the acceptance on their part of new capital
requirements based on market discipline. In fact, that is the
scenario envisioned by the Bankers' Roundtable in their recent
statements of support for enhancing market discipline in U.S.
banking. The Roundtable realizes that the creation of credible market
discipline would clear the way for deregulation, because it would
eliminate any possibility of a "safety net subsidy for risk."
Avoiding that subsidy has been one of Federal Reserve Chairman Alan
Greenspan's main arguments in blocking some of the most dramatic
elements of U.S. bank deregulation.

In emerging market economies, placating vested interests is more
complicated, particularly in countries with weak banking systems
where it would be difficult for banks to accept discipline. There are
countervailing pressures, however. Members of the World Trade
Organization have agreed to open their financial markets to
competition. Our proposal would strengthen this agreement by
restricting membership in the IMF, and therefore access to IMF
resources, to countries that adopt sound banking policies. Long-term
World Bank loans could be used to strengthen banks' capital structure
and, thus, their ability to compete.

Persuading the Treasury to relinquish its power to use the Exchange
Stabilization Fund, the IMF and the World Bank as off-budget slush
funds will not be easy. However, a large part of the public opposed
the recent appropriation to increase the IMF's resources. Congress
delayed approval for many months, insisted on minor reforms and
established an independent commission to recommend deeper reforms of
international financial institutions. Congressional recalcitrance and
public opposition have gained the attention of Treasury and IMF
officials.

The U.S. government is the largest contributor to IMF and World Bank
funding. If Congress responds to public concern about the large sums
spent in Mexico, Russia, Asia and Brazil, reform could be speedy,
credible and deep.

The IMF has responded to recent financial crises and its own past
failures by proposing increased transparency, better, more timely
release of information, and better surveillance and supervision.
These suggestions are useful but not in themselves adequate to
correct the problems in international financial arrangements.
Government supervision and accounting standards do not prevent
failures, as regulators sometimes fail to use available information
or look the other way when violations of prudential standards occur.

The incentive structure of the IMF is counterproductive to reform, as
it rewards officials for making loans, not for insisting on
prudential policies. Corruption in Russia, Indonesia and elsewhere
was not a secret. IMF officials had no incentive to emphasize
problems of this kind or even to insist on enforcement of the
conditions agreed to when the loans were made. The reason is clear:
unlike uninsured private market creditors, government and
international supervisors lack both the incentives and the ability to
enforce prudential standards. Supervisory failures of the U.S.
savings and loan system and the banking systems of Mexico, Japan,
Thailand, Korea, Indonesia and many other countries resulted from
failure to use available information in a timely way.

Our proposal anticipates greater reliance on fluctuating exchange
rates. But even in countries with fixed exchange rates, market
processes, incentives, diversification and competition will improve
safety and soundness and reduce the risk of financial collapse
accompanying exchange rate devaluations. Better information and
supervision help these processes to work, but they are complements
to, not substitutes for, market discipline.

By eliminating IMF discretion over the circumstances under which
lending occurs, and over the conditions and terms of that lending,
our proposal would end discretionary interventions by the IMF to
distribute emergency foreign aid to insolvent governments and
financial institutions as part of a bailout plan.

It is easy to construct examples of potentially beneficial emergency foreign aid using complicated economic models to show that, in some circumstances, the bene fits exceed the costs. The IMF often justifies its actions as a means of preventing crises from spreading to other countries. This argument has some merit. A crisis in one country calls attention to unwise policies and weak financial systems elsewhere. Global lenders suffer losses, so they may restrict credit to solvent borrowers in other countries that they previously financed.

But the solution does not lie in rescuing foreign lenders. That encourages continued imprudent behavior. A better solution is to give countries incentives to reform their financial structures and improve their policies to make them less vulnerable to contagion. Markets may err for a time, unable promptly to distinguish the solvent from the insolvent borrowers and the more risky from the more secure loans. These errors do not persist for long.

History teaches that misaligned incentives, not inherent financial fragility, are the primary source of insolvency crises in the world today. Government safety nets and IMF bailouts are a major part of those incentive problems. It is possible to correct these core incentive problems by constructing a world financial system subject to market discipline, with fewer and smaller liquidity crises.

Although our proposal would prevent the IMF from giving ad hoc foreign assistance to insolvent financial systems, other mechanisms would be a useful supplement. Bankruptcy laws that delineate loss-sharing rules are the answer to insolvency problems in financially developed economies. Similar rules would have evolved much faster in underdeveloped economies if IMF-orchestrated bailouts had been absent over the past twenty years.

In April the IMF announced a new facility to lend to qualified countries in advance of a crisis. Although the proposal moves the IMF in a direction we recommend, the change is insufficient and incomplete. First, the program is an addition to, not a substitute for, current IMF programs. Lending to countries with insolvent banks could continue. Second, lending criteria are vague and subjective. Unless objective conditions are clear in advance, response to crises will continue to be delayed. Third, there are no collateral requirements to maintain lending standards and reduce risk. The new program, like previous programs, invites discretionary judgment and political influence.

We envision a new IMF, one providing elastic and immediate liquidity to member countries that share a commitment to sound financial practices, and one engendering a system based on rules that increase incentives for prudent behavior by lenders and borrowers, the private sector, the governments and the IMF.

What the IMF would lose is its power to demand policy changes as a condition of its loans. It could not act as the agent of any government; its staff would gather information and make interventions under predetermined rules; it could not exercise discretionary lending that could be prone to corruption. The rules would be the same for all borrowers, no different for those favored by the U.S. Treasury Department than for others. Countries could at their own risk choose other policies by opting out of membership in the IMF.

Such reforms as we propose would greatly reduce the frequency and size of financial crises to the benefit of all nations, and would convert the IMF into a less costly, more stabilizing institution that could offer incentives for prudent policies and market-based solutions.

Charles W. Calomiris is professor of finance and economics at Columbia University. Allan H. Meltzer is professor of political economy at Carnegie Mellon University. Both are visiting scholars at the American Enterprise Institute.