Skip to main content

Rethinking IMF Rescues

The financial rescues arranged by the International Monetary Fund in Southeast Asia have generated much controversy, not only within the Congress but around the world. Defenders of the IMF argue that its hard currency loans have prevented the financial flu from reaching epidemic proportions, while its stringent loan conditions will help cure the disease and prevent it from recurring. The IMF’s critics disagree. They argue that the rescues invite reckless financial behavior by borrowers and lenders, creating what economists call a moral hazard. Others attack the IMF’s conditions as being inappropriate for Asia’s problems, excessively painful, unlikely to last for cultural reasons, and likely to trigger a political backlash against the West—or some combination of the four.

In early June, responding to Treasury Secretary Robert Rubin’s call for private sector ideas on modernizing the global financial architecture, Brookings convened a symposium of experts with widely diverse views to address what is perhaps the most vexing issue raised by the IMF’s actions: whether it has created a moral hazard, and if so, what if anything should be done about it. The diverse views, summarized below, illustrate the complexities of this issue, but show a strong consensus in some key areas.

The Moral Hazard

Moral hazard is a well known problem in the world of banking and insurance. Policymakers seek to contain it for banks—which benefit from deposit insurance and central banking services—by requiring some portion of their assets to be backed by capital, or shareholders’ funds, and by supervisory restraints on risk taking. Insurers seek to contain risk taking by requiring their insured to absorb some of the first dollars of their losses (deductibles) or by issuing insurance only up to some limit.

All of the speakers recognized that international financial rescues inevitably entail some moral hazard and all agreed that it is unlikely that international financial crises—marked by market panic and contagion—will disappear. A key challenge for policymakers is to balance the benefits of rescues when such crises occur against the costs associated with the distorted incentives that these rescues create.

In fact, the Asian crisis has highlighted two types of moral hazard: global and domestic. Global moral hazard arises when lending by the IMF or individual countries (such as U.S. lending to Mexico in 1994-95) encourages potential borrowing countries to pursue policies that make them more vulnerable to currency runs and other crises. Allan Meltzer argues that the global moral hazard created by IMF rescues has played a major role in the nearly 100 banking crises that have occurred in the developing world in the last 15 years. He says that without these rescues, banks would have behaved far more prudently. Other participants suggested, however, that most of these banking crises were not associated with the IMF, had purely domestic origins, and were often caused by poorly implemented financial liberalization policies.

Domestic moral hazard arises when actions and cultures within countries themselves lead to excessively risky behavior. So-called crony capitalism—implicit government guarantees of uneconomic bank-financed projects—is a prime example. Financial decisions stem from corruption and political expediency rather than from economic rationality. Virtually all of the participants agreed that domestic moral hazard is a much more important source of Asia’s problems than global moral hazard.

Meltzer said that domestic moral hazard flows not from the prospect of IMF guarantees but from policymakers taking risks by guaranteeing private sector debt that they know may result in financial disaster. He argued nevertheless that the prospect of international financial rescues can aggravate domestic moral hazard by making it easier for governments to provide private sector guarantees. However, others questioned whether the possibility of financial rescues really works in this fashion. After all, political leaders must take political heat for having to admit their failed policies and may even have to sacrifice their jobs.

That leaves the question of how much the prospect of the IMF’s rescue contributed to the Asian crisis. On this, participants were divided. Some argued that the IMF essentially played no role because no one believed the Asian miracle would end. Moreover, Asian countries were growing so rapidly prior to the crisis that they would naturally attract large inflows of foreign capital—which could run out much faster than it came in—regardless of the prospect of an IMF rescue.

Others sharply disagreed, asserting that investors would not have poured so much money into Asia—certainly not at such relatively low interest rates—if they had not expected some kind of IMF rescue if things turned sour. Some blamed the IMF’s bailout of Mexico in 1994-95 for leading to the Asian crisis of 1997-98, although most disagreed with that conclusion.

James Tobin offered a third view: while moral hazard is a serious and difficult problem, it was not the major cause of the Asian currency crisis. In Tobin’s view, fixed exchange rates were the main culprits. Had the Asian countries adopted floating exchange rates, the depreciation in their currencies—which was inevitable given their large current account deficits—would have occurred gradually, without the sharp and ultimately contagious plunges that actually occurred. Moreover, had their exchange rates floated, Asian banks and corporations would not have been so eager to borrow in steadily appreciating foreign currencies, which significantly aggravated the crisis.

What Should Be Done?

In finding solutions to the moral hazard problem, it is vital to distinguish whether a crisis is one of illiquidity (borrowers will be able to repay once any panic in the markets has dissipated) or of insolvency (where the economic fundamentals demonstrate that borrowers do not have the means to repay their debts).

In the case of illiquidity, temporary financing from a lender-of-last resort (domestic or international) is fully appropriate and should entail little moral hazard. When the problem is one of insolvency, however, lending to borrowers clearly can lead to moral hazard, encouraging countries to ignore serious and persistent structural problems. When borrowers are insolvent, their debts must be restructured and written down or exchanged for equity, or in a worst case their enterprises liquidated—preferably through an orderly bankruptcy process that can allow private borrowers and countries the breathing room to address their structural problems.

A difficult task confronting policymakers at the outset of any crisis, however, is to determine what kind of a crisis it is—illiquidity or insolvency. Although participants provided no clear answers, there was unanimity on the need for an international financial institution to act as a lender of last resort to provide liquidity and stability to the financial markets in the event of a serious international financial crisis. As one participant stated, if there were no IMF, one would have to be created. No individual country, including the United States, has the political clout or even the political will to impose IMF conditions for lending or to fund the size of rescue packages that have been required. Only one participant recommended the abolition of the IMF—Allan Meltzer—and even he argued for giving lender-of-last-resort functions to the Bank for International Settlements under stricter traditional borrowing standards for central banks.

There was much discussion on how to reduce the likelihood that future crises will require financial rescues, and if so, of the adverse impact these rescues might have on incentives by private actors to avoid excessive risk taking. These suggestions fall into two categories: market-like incentives to encourage more appropriate behavior by private and public actors and regulation designed to accomplish the same or complementary objectives.

Market-like Incentives

Speaking on behalf of the Shadow Financial Regulatory Committee—an independent group of experts in the U.S. banking and financial system—George Kaufman and Robert Litan outlined one incentive-based scheme. To be eligible for future IMF borrowing, or to avoid large penalty interest rates on any such borrowing, they argued that countries must adopt changes in their domestic law that implement automatic write-downs, or haircuts, on foreign currency denominated interbank loans that the banks do not roll over at maturity.

Ideally, the haircuts would be equivalent to the discount that markets impose on a country’s sovereign debt just prior to the IMF rescue. The haircut proposal is designed to remedy the fact that multinational banks largely escaped any significant pain because they essentially were repaid out of the proceeds of the IMF lending packages, unlike investors in Asian equities and bonds who took major losses during the recent Asian crisis. Kaufman and Litan argued that the possibility of future haircuts would curtail foreign currency bank lending and borrowing and, in any event, result in more accurate pricing.

The haircut proposal was criticized from several directions. Alice Rivlin suggested that the prospect of haircuts would precipitate crises by encouraging creditors to rush for the door. Several participants questioned whether the modestly higher interest rates on foreign currency lending generated by haircuts would be effective in discouraging excessive lending and borrowing given the low interest rate spreads on short term loans that represented the bulk of the problem in Asia.

William Niskanen wondered whether advance announcement of IMF lending conditions would be more credible if the IMF’s resources were augmented or kept where they are now.

Countries might not take seriously the intentions of a too flush IMF to act on its preannounced conditions for asking for assistance. At the other extreme, if the IMF’s resources were perceived to be inadequate, then its preannounced conditions would be irrelevant; the shortage of funds itself would provide the deterrent to excessively risky behavior at the cost of not having an effective rescue mechanism in place once a crisis occurs. Given the big uncertainties over the nature and kind of future international crises, it was not surprising that participants failed to reach consensus on the precise amount of IMF funding needed.

A number of participants, nonetheless, agreed on the need for specific mechanisms to ensure that key parties are not bailed out by IMF lending packages. For example, Morris Goldstein argued that IMF funds should not be used to protect creditors or shareholders of non-banks, such as finance companies, investment banks and non-financial corporations. Goldstein also urged the IMF to lend into arrears—that is, to continue lending to countries when borrowers are behind in repayments to creditors—as long as the country remains committed to a reform program. Such a policy would reduce the leverage of foreign creditors to demand debt guarantees by host country governments eager to obtain IMF assistance, which the creditors typically count on to help repay their loans. Along the same lines, Goldstein urged that standard borrowing contracts be revised to permit restructuring of debts of insolvent companies by majority vote of the creditors rather than requiring unanimity, a step that would also facilitate prompt workouts.

Several participants argued for simpler ways of discouraging countries from policies that might require them to ask for IMF assistance. Allan Meltzer urged that international lenders—his preference being the BIS—should follow classical lender-of-last-resort rules by requiring borrowing countries to provide high quality collateral and pay penalty interest rates as conditions for funding. Michael Mussa responded that the IMF already imposes penalty interest rates in its new lending facility. Indeed, Jeffrey Shafer pointed out that the IMF’s Asian loans carry progressively higher interest rates the longer they remain unpaid.

Alice Rivlin offered yet another incentive-based way of addressing the moral hazard problem through a new certification system by an independent international organization. It would rate the quality of bank supervision, bankruptcy administration, and the conditions of financial institutions in various countries to signal to market participants which countries carry especially high risks. Countries with low grades from the new agency would have to pay more for credit until their ratings improved.

It was broadly agreed, however, that the IMF should not play this role, for at least two reasons: the Fund might be tempted to temper its assessment of a country’s performance to avoid endangering the country’s ability to repay its IMF borrowings or to support a previous IMF assessment; and an institution like the IMF, which is staffed largely by economists, may not have the skills to evaluate effectively credit risk, a task performed in the private sector by trained specialists.

Additionally, it was suggested that as developing and emerging market countries develop broader and deeper capital markets, they will rely less heavily on banks to provide financial intermediation and thus diminish the moral hazard associated with potential international bank rescues.

Finally, there was broad support for addressing creditor moral hazard by requiring greater loan diversification by commercial banks to minimize future financial crises and the impact of problem loans on the safety and soundness of internationally active banks. Markets should encourage this approach by lowering the value of bank stocks if banks, as expected, are forced to write down corporate loans in some Asian countries. However, the market has proven itself an imperfect device for encouraging risk diversification so this approach may fall more appropriately into the category of a regulatory proposal. Diversification of risk is increasingly being required by bank supervisory authorities.

The Regulatory Approach

In response to the widely held view that the heart of the Asian financial crisis resulted from weaknesses in country banking systems, Morris Goldstein offered a number of proposals for reforming the banking supervisory systems in developing and emerging market countries (drawing on his well-received book, The Case for an International Banking Standard, Institute for International Economics, 1997). Among other things, Goldstein urged the IMF to require these countries to implement and enforce higher capital standards for their banks, to adopt the Basle Committee’s Core Principles for Effective Banking Supervision, to introduce an effective bankruptcy system, and to establish deposit insurance systems that cover only, or give priority to, small depositors. Others suggested that countries require banks to back some portion of their assets with subordinated debt—uninsured instruments that cannot be redeemed until they mature, and thus cannot be the subject of runs, holders of which are typically risk averse.

Although Goldstein’s proposals were broadly supported, there were skeptics. Niskanen questioned the wisdom of international standards generally, claiming the Basle risk-based bank capital standards have inappropriately encouraged too much bank lending to governments and too little to the private sector. Barry Bosworth, among others, wondered whether too much attention was being paid to Asia’s banking problems as major contributors to currency runs, when exchange rate regimes are more important.

However, a majority of participants supported the IMF’s expanded role in imposing microeconomic structural and regulatory conditions on financial systems in the countries to which it has extended credit. These conditions help provide a stable foundation for renewed growth and signal to markets that the practices that contributed to the crisis are being abandoned.

Henry Kaufman offered the most far-reaching regulatory proposal for addressing moral hazard: the creation of a new international institution to oversee and set standards for bank capital; accounting; and training, reporting and disclosure. In addition, his proposed Board of Overseers of Major Markets and Financial Institutions would monitor the performance of financial institutions and markets; rate the credit quality of market participants; require higher capital standards on loans to banks in countries that remain outside the new system; and limit the eligibility of non-members to sell new securities and equities, bonds, and money market instruments.

While Kaufman’s proposal was viewed as much too extensive to be feasible, there was strong support for having internationally-accepted minimum capital and supervisory standards for banks. Additionally, there was broad support for a deregulatory option: the removal of restrictions countries maintain on foreign ownership of their banks and other financial institutions. The entry of multinational banks in developing and emerging market countries would better insulate those markets against bank failures because their lending and deposit-taking is more diversified than in purely domestic banks.


Financial crises are essentially inevitable, and therefore the need for international financial rescues in limited circumstances under the auspices of the IMF cannot be avoided. Moreover, the moral hazard that accompanies such rescues cannot be eliminated completely, but policymakers can take steps to reduce its contribution to the frequency and severity of financial crises. The Brookings symposium concluded that the IMF serves a necessary function in addressing international financial crises and outlined a number of significant proposals that can and should be implemented to diminish moral hazard and thus address significant concerns raised in the United States and in global financial circles.

Get daily updates from Brookings