Late in 2001, the new first deputy managing director of the International Monetary Fund, Anne Krueger, made a bold suggestion. Under certain conditions, she proposed, a government's international debt repayments should be temporarily suspended while negotiations take place on restructuring that debt. With her statement, the IMF officially endorsed one of the more radical suggestions for improving the international financial architecture to have come forth since the Mexican and Asian crises in the 1990s. If implemented properly, the Krueger proposal would represent some improvement over the IMF's current prescriptions for states facing financial collapse. But given the domestic origins of most financial crises, her plan cannot eliminate them altogether.

The problem that Krueger addressed is straightforward. When any debtor nation develops economic difficulties, its creditors worry about being repaid, so they move as quickly as they can to protect their positions. For example, they can decline to roll over (that is, extend) loans that have matured. They may even sell the loans before maturity, although that move, of course, requires willing buyers. The difficulty that some governments faced in rolling over maturing debt played an important role in several recent debt crises: Mexico (1994-95), Russia (1998), Brazil (1998-99), and Argentina (2001-2). Even those creditors willing to roll over their loans at a satisfactory interest rate may hesitate for fear of being alone -- and thus caught in a payments crisis.

This problem can arise for any debtor and its diverse creditors. Within national borders, such crises are dealt with through bankruptcy proceedings. When a debtor, such as a private firm, gets in trouble, it can "file Chapter 11" under U.S. bankruptcy law, thereby legally suspending payments to all creditors except the tax authorities. That suspension gives the ailing firm breathing space, in which several things can happen. The debtor can reorganize its finances to make resumption of debt service more likely. It can borrow additional funds, with repayment preference given to these new creditors. And it can negotiate with its old creditors to extend those loans' maturity and perhaps ease their terms. All this reorganization is done in the United States with the protection and guidance of a court of law. Only if the debtor cannot restore its financial health are its assets liquidated and the proceeds distributed to its creditors -- again, under the guidance of a court.

Another key feature of U.S. bankruptcy proceedings is that not every creditor needs to agree to the negotiated deal in order to be bound by it. Under the "cram down" provision of bankruptcy law, if a settlement is reached by creditors carrying two-thirds of each class of claims (that is, each group of loans ranked by the urgency with which they need to be repaid), all creditors are covered. A single small creditor is therefore powerless to hold up a deal in an attempt to get better treatment.

No such procedure, however, covers international claims on sovereign borrowers. If a government runs into problems with its payments, each creditor may move to protect its interests, risking a rush to the door in times of real or threatened difficulty. The new IMF proposal is designed to fill this gap and thus provide an international analogue to national bankruptcy proceedings when the borrower is a government.

Although the need for such a provision has been recognized since the early 1980s, it was recently given a strong push by the U.S. court case Elliott Associates v. Peru. In early 1996, Elliott Associates, an investment fund specializing in high-risk securities, purchased government-backed loans to Peruvian banks at a heavy discount. The purchase was made while Peru was negotiating a settlement with its other creditors. Several months later, by agreement with those creditors, Peru settled their claims by offering some cash and Brady bonds, which are notes backed by U.S. Treasury securities. Elliott then sued for full payment of principal and accrued interest. That claim was dismissed by a U.S. federal court but later reinstated on appeal. To assure payment, Elliott obtained restraining orders in late 2000 in both New York and Brussels to block Peru's payments to the holders of the Brady bonds -- i.e., the creditors who had agreed to the debt restructuring. To avoid a formal default, Peru settled out of court by paying Elliott $56 million, five times the amount that Elliott had paid for the loans less than five years earlier. This episode rewarded a holdout at the expense of creditors who had reached a negotiated settlement, and it underscored the difficulties surrounding government debt restructuring, at least with respect to private creditors. (When only public creditors are involved, governments negotiate debt restructuring in a group called the Paris Club, where the same governments meet repeatedly and the potential for reciprocity has generally moderated the stands taken.)

Applying the analogy of bankruptcy proceedings to governments, as Krueger has suggested, is therefore attractive. But like all analogies, it can be carried only so far. In bankruptcy, two business judgments must be made. The value of the firm's physical assets at the time of liquidation must be compared with the prospects of the firm's eventually paying back more if it is allowed to continue to operate. In contrast, governments service debt not out of operating earnings but out of tax revenues, which face competing social claims. Thus the parties involved must make political and ethical judgments about not only the feasibility of continuing to service the outstanding debt but the desirability of doing so. Similar calls must be made when a local government files for bankruptcy in the United States.

The IMF is probably the best judge of a debtor country's immediate fiscal prospects. It can also assess the government's technical competence and its integrity in carrying out its promises -- even though the fund does not advertise the fact. But it is not well suited to gauge the complicated and controversial ethical tradeoffs between violating a contractual promise to pay creditors and imposing social costs to carry out that promise. Such judgments can be made only by politically responsible officials.

But the IMF could provide a forum for discussions on the subject. So there is, in principle, a rationale for providing a legally sanctioned pause in debt servicing in the case of sovereign debt held by private parties, for discouraging holdouts, and for providing new credits where appropriate. Any such plan, though, would raise four crucial questions: What kinds of indebtedness should it cover? To which countries should it apply? Would the possibility of an enforced pause in debt servicing reduce international flows of capital, particularly to developing countries? And what steps need to be taken to put such an arrangement in place?


The obvious candidates for international Chapter 11-style coverage are the foreign holders of a government's bonds and the foreign banks that have lent to that government. But if a pause in payments were limited to debt owed to foreign creditors, foreign debt-holders could arrange to sell the maturing debt to local residents instead. This move may be especially difficult to control when the bonds in question are issued domestically rather than internationally. But foreign holdings of domestic bonds are often extensive -- as was the case in Mexico, Russia, Brazil, and most recently, Argentina. Foreign creditors, moreover, will not appreciate being held up while domestic creditors are repaid. Thus domestic holdings of government bonds must also be subject to the pause. And if domestic banks rely heavily on government bonds as a source of their liquidity, suspending payments on those bonds may also require introducing restrictions on bank-deposit withdrawals, unless arrangements are made with the central bank to liquefy them when necessary.

What about other domestic claims on the government, or situations in which foreign holdings of domestic bonds are negligible but international credits are covered by the payments pause? Again, foreign creditors are not likely to agree to overt discrimination, except possibly in cases in which the origin of the crisis is clearly external rather than budgetary. It would not be acceptable, for example, in terms of either efficacy or equity, to halt payments to foreigners in the name of providing temporary relief during debt renegotiation while allowing residents to be repaid and then freely export the proceeds. Residents would therefore have to be subject to effective controls on converting domestic money into foreign currency.

The last point seems obvious, yet implementing it in today's world could be highly problematic. Many countries have dismantled their controls on foreign exchange or greatly reduced them to cover only financial institutions. And those controls remaining are at the mercy of publics determined to export capital -- including through purchases of foreign currency such as U.S. dollars. In recent crises, none of the afflicted countries had really effective exchange controls, and Malaysia was widely criticized for tightening them as the Asian financial crisis unfolded in 1998. In short, there is no simple solution to the first question.

What about which countries to cover? Everyone talks of the "emerging markets," which have acquired greater access to the international capital market in recent years. But the composition of this category constantly changes. Should it include the low-income members of the Organization for Economic Cooperation and Development, such as Mexico or Portugal? Should it include Italy, whose future fiscal outlook is by no means rosy? Or Germany, which recently threatened to break through the budget-deficit barrier of Europe's monetary union? Or the United States, the world's biggest debtor nation? What would be the grounds for extending the provision to some members of the IMF but not to others? Again, there is no clear answer.

Then there is the question of whether including an automatic rollover provision in government debt would deter the flow of capital to developing countries. It is worth noting that corporate debt has long been subject to bankruptcy proceedings in the United States and many other countries, yet corporate bond markets at home and abroad have remained robust. Some creditors may have been deterred, but not enough to keep the corporate bond market from thriving. Moreover, deterring at least some private capital from heading to emerging markets could actually be a good thing. Research shows that the domestic benefits of foreign capital are surprisingly ambiguous. And in a world economy ridden with import protection and high taxes -- which produce incentives for tax evasion -- the benefits are mixed even on theoretical grounds alone. On balance, the evidence suggests that foreign capital does result in a rise in per capita income in debtor countries. But the evidence is much stronger on the benefits of foreign direct investment -- which involves importing management skills, marketing know-how, and technology as well as funds -- than it is for portfolio equity or interest-bearing debt. Indeed, if the economic losses associated with recent financial crises are attributed in large part to heavy foreign indebtedness and to the withdrawal of foreign funds (an argument this author finds dubious), the net effect of interest-bearing debt during the 1990s was probably negative rather than positive. After the fact, it is clear that too much international borrowing and lending took place. Any device that now introduces greater caution into such lending will thus be beneficial.

That brings us to the final question. If Krueger's rollover provision would indeed improve current arrangements, how could it be brought about? One suggestion, which the U.S. Treasury seems to favor, is that standstill clauses should be adopted voluntarily in debt contracts involving cross-border transactions, or at least those involving currencies other than that of the borrower. Debt contracts, whether bonds or loans, contain many conditions today, including the waiver of sovereign immunity when the borrower is a national government. So it would be simple in principle to add a clause permitting temporary extension of the debt's maturity, at the initiative of either the debtor alone or with approval of some third party, such as the IMF.

But any potential debtor eager to get access to capital markets would be reluctant to start the process, particularly since it could be done only for new debt. New creditors would thus be subordinated to existing creditors. It would take many years before all outstanding debt came to contain such clauses, even if adoptions to cover new debt were universal. But sufficient publicity on the advantages of such provisions, along with pressure by international financial institutions, might gradually bring it about. The process could be accelerated if major rich countries adopted the practice in their own public debt offerings.

A second possible route would be through legislation in the major creditor countries, as well as in those countries where debtors have significant assets that might be legally seized by a creditor. Such legislation would be designed to immunize debtors in distress from successful lawsuits, provided the agreed conditions were satisfied. Sales of claims on debtors in distress to residents of countries without such legislation would undoubtedly occur, but the courts in such countries would be unable to enforce any penalties.

A third possible route would be through a treaty, which would have to cover at least the same set of countries as the second route to be effective. One form would be formal amendment of the IMF's Articles of Agreement, which would be an especially desirable route if the IMF were asked to play a key role in the process. For example, it could certify the need for a payment standstill and make sure the debtor country meets the conditions for a settlement. This third route would be the cleanest and most coherent, even though the ratification process would be arduous.


How does a government run into serious debt problems in the first place? One explanation is exceptionally bad luck. An unforeseen sharp and durable drop in world demand for a country's principal export product, for example, could result in a dramatic reduction in government revenue. Another reason is shortsighted borrowing by a country's leaders. Those who lend to such governments, of course, also gamble, and they are paid higher interest for their risk. In short, the possibility of default should not be a surprise, especially when the borrowing is shortsighted. But without debt relief, the losers are the citizens of the debtor country. The government will have no choice but to squeeze its economy severely as a way to enlarge its trade surplus, which in turn will service the outstanding debt. (But squeezing the economy often worsens the government's budget deficit, of course, causing a double bind.)

For this reason, countries may sometimes need relief from debt servicing -- or at least the possibility of relief -- even when their long-term debt burden is bearable. Should a lot of its debt come due in the near future, creditors may worry about a country's ability to refinance in the short run. In the process, they make refinancing impossible -- a classic example of a self-fulfilling prophesy. Of course, borrowers would be well advised to avoid getting themselves into such a situation in the first place. But sometimes such "bunched" borrowing may be unavoidable because external events lie beyond the country's control. To head off the often harsh consequences of a creditor panic, therefore, the international community should be able to liquefy the troubled economy with enough foreign exchange in the short run to persuade foreign creditors and potential creditors that they will not be left holding an unserviceable claim. That reassuring role is played by a central bank within a national economy.

This issue arose in Mexico in 1995 and in South Korea in 1998. In both cases, large currency depreciations had just occurred, making exports much more competitive and prospectively reducing those countries' substantial trade deficits. But for each, a large amount of short-term debt was to mature in the coming months. Thus creditors were individually reluctant to extend their loans, even though the economic fundamentals of each country seemed to be satisfactory, or on the way to becoming satisfactory. In each case, the international community assembled a large official support package, centered on IMF loans but also augmented by funds from other institutions and governments. In the case of South Korea, the IMF also pressed creditor banks into extending their maturing loans. In the end, both efforts succeeded, but the economic costs were higher than they should have been. Furthermore, both attempts were ad hoc and fraught with suspense over whether the requisite support could be assembled and whether it would work. Large amounts of the assembled official funds had to be used to cover the privately held maturing debt, so creditor panic was partially neutralized rather than averted. (It is noteworthy that in both cases resident funds began to leave in volume before the currency devaluations, whereas withdrawals of foreign funds occurred mainly afterward.)

As a result of these and other experiences, the IMF has taken two important steps to deal better with such crises in the future. First, it has streamlined its procedures so that it can act much more quickly, even though its processes still take more time than may be available in an emerging financial crisis. Second, it has increased the resources for emergency lending through its New Arrangements to Borrow -- which, with the earlier General Arrangements, let the IMF call on 25 countries for up to $46 billion under appropriate conditions.

The IMF was originally set up to offer loans that would cover temporary imbalances in current account transactions. In the mid-1990s, the IMF extended its responsibilities to cover international capital transactions as well. But some recent international-debt problems have been primarily budgetary rather than external in origin. This fact was especially true of Russia in 1998, but also of Brazil in 1998 and Argentina in 2001, even though all three situations developed an international dimension. The basic problem was that those governments could finance their deficits only on increasingly onerous terms. Rolling over their debt at ever-higher interest rates aggravated the budgetary problem, and it also enlarged the imbalance in external payments because foreigners held government debt and declined to renew it.

The IMF lends to governments, so IMF loans provide temporary budgetary as well as balance-of-payments relief. But the IMF should not provide budgetary relief when there is no strong external dimension to the problem. If the IMF creates the presumption that it might do so, it will only encourage governments to get relatively cheap loans from the IMF at medium-run maturity -- better terms than they can usually get on their public debt under stressful conditions.

There are thus three quite different circumstances calling for IMF support. First, the fund can help finance a temporary deficit in international payments that might arise, for example, from a recession-induced fall in foreign demand for a country's exports, or to cover the period before a currency devaluation can improve the trade balance. This approach was the IMF's traditional purpose.

Second, the fund can help avert a creditor panic when a lot of external debt matures in a short period of time, in order to assure creditors that the country has enough liquidity to repay the debts. In that case, the ready availability of IMF support may make its actual use unnecessary, since bankers are always willing to lend to those whose credit rating is high.

Finally, the IMF can participate in consolidating and perhaps reducing an external debt that has become too heavy for a government or country to carry indefinitely. Here the IMF could provide funds to help persuade private creditors to accept some debt reduction in exchange for some cash up front, assuming that the alternative of default is even less attractive. The Krueger proposal concerns only this last case, without emphasizing IMF financing.

Of course, world affairs rarely fall cleanly into these categories. Elements of each may be present at the outset, and the first case may easily provoke the second, or even the prospect of the third. All the same, the IMF's ability to deal with financial crises and to forestall creditor panic would immeasurably improve if it could provide sufficient resources to cover even the worst contingency. Even simply having such resources would greatly reduce the probability of the worst contingency. This capacity could be provided by empowering the IMF to issue its own "currency," the Special Drawing Rights (SDRS), in such emergencies. At present, the IMF can issue SDRS (by 85 percent vote of its board) only to meet the long-run needs of the world economy for additional international liquidity. As a result, SDRS have been issued only twice, in 1970-72 and 1979-81, in amounts worth about $27 billion today, making up a paltry 1.4 percent of the world's international reserves. (A third, special allocation of $27 billion was agreed to in 1997 but awaits U.S. ratification.)

This approach would mean adding an additional purpose for the issuance of SDRS. The conditions for temporary issuance would have to be tightly drawn, and any SDRS actually issued would subsequently be withdrawn when the emergency had passed, just as the Federal Reserve Board first injected and then withdrew extensive credit following the collapse of Long-Term Capital Management in 1998. The classic conditions for action by a lender-of-last-resort to banks are that, in an emergency, it should stand ready to lend without limit, at a penalty interest rate, and against good collateral, and it should make this readiness known ahead of time. These actions assure adequate liquidity in a financial system; they are not a bailout of insolvent banks, which requires a different course. But dealing with insolvent institutions is much easier in a favorable financial environment than in a cash-strapped one.

With the ability to use SDRS, the IMF could play a surrogate role of lender-of-last-resort for international financial crises -- a role it cannot assume today because of the potential shortage of funds. It could then concentrate on the lending conditions instead of worrying about creating a coalition of willing new lenders in each case. Furthermore, IMF conditionality provides a substitute for collateral, although the pledging of collateral by states, as was done by Mexico against the U.S. Treasury loan of early 1995, should not be excluded.

Some critics argue that such an approach would increase moral hazard. But the IMF can reduce this risk by imposing stringent conditions and other measures. These include conditioning IMF lending on prior acceptance of internationally agreed standards, the presence of standstill clauses in international loan covenants, and a higher-than-normal rate of interest. Empowering the IMF this way would require an amendment to the IMF's articles, but that amendment would be a natural addition if the articles are to be modified anyway to provide for a standstill in debt repayment with IMF involvement, as Krueger suggests.


Financial crises are not new but the inevitable byproduct of economic development, a kind of adolescent growing pain for young economies. As countries evolve from poor, agricultural economies into high-income economies, a severe tension develops between the real economy and the financial structure necessary to sustain it. A key feature of development is to socialize savings, so that people move private assets such as jewelry into financial institutions. That way, savings can be mobilized for productive investment.

A key problem with this otherwise desirable process is that it gives bankers (or their backers) more money than they ever dreamed of. All kinds of alluring projects become financially possible. Some bankers do not seem restrained by the fact that they are dealing with other people's money. They start to invest on a large scale, which may create a real boom and a resulting sense of euphoria. Industrial production, profits, employment, and capital gains all rise. Things go so well that any initial caution is soon forgotten. But unless productivity rises along with everything else, such booms risk turning into Ponzi schemes that are unsustainable in the long run. Former Federal Reserve Chairman William McChesney Martin once famously said that the role of the central bank is to take away the punch bowl just as the party really gets going. But that oversight requires a prudent, independent, nonpartisan central bank with its eye on the long run -- something most developing countries do not have, at least not until they have been through a few serious financial crises.

Every country experiencing a financial crisis in the 1990s had an unsustainable domestic situation: namely, a mismatch between financial claims and the real economy's performance. But the details differed significantly. Sometimes private banks were involved, sometimes governments, often both. Several governments discovered the wonders of domestic and international financial markets, which let them rely on floating bonds to finance government expenditures without imposing unpopular taxes in the short run. Thus Russia, Brazil, and Argentina all found a way to avoid difficult choices -- but only for a while.

Americans have no reason to be smug. The United States had a serious financial crisis nearly every decade during its formative developmental period starting in the 1810s and culminating in the Great Depression of the 1930s. (European countries had similar experiences during the nineteenth century.) And the lessons learned then, which produced a raft of regulatory legislation, did not avert the U.S. savings-and-loan crisis of the 1980s, which was brought on through legislation aimed to help important constituents of key congressional leaders. It was fashionable (and sometimes partially correct) to blame these crises on foreigners, but they were overwhelmingly domestic in origin.

Theoretically, these crises might have been avoided. But doing so would have required Platonic monetary guardians who were detached, disinterested, and farsighted -- the antithesis of the modern politician. In practice, the capitalist system works by harnessing greed, not charitable inclination, to achieve economic progress. It has been smashingly successful during the past two centuries, especially the last 50 years, and its success is now spreading from Europe and North America around the world. But progress has not been smooth. Instead, it has been a process of trial and error, punctuated by financial and economic crises that eventually provoke improvements in the institutional structure and the legal incentives that channel greed in socially constructive directions. Each new generation of financial wizards and their lawyers will try to find lucrative loopholes in the rules put in place by the previous generation -- in response not to their foresight but to their own acknowledged mistakes.


The bottom line is that financial crises are an inevitable companion to the economic development of any country. Contrary to recent conventional wisdom, they are not the fault of the international economic system. Nor does their presence signify serious defects in that system. To be sure, the internationalization of financial markets makes more funds available and increases the number of potential investors, who are driven by euphoria that can tempt even hard-nosed Western bankers and money managers. Thus foreign capital can affect the magnitude and the detailed dynamic of financial crises. But history suggests that the prime mover is domestic -- and in all recent crises, domestic funds were implicated in precipitating the collapse.

As country after country has discovered, a high level of disclosure of information is required for well-regulated and well-functioning financial markets. Yet full disclosure is anathema to commercial and financial practice in many societies. A clear hierarchy of financial responsibility and risk must exist, ranging from shareholders to bondholders to depositors. It must be clear who bears the costs if something goes wrong. Yet governments in developing countries are often complicit in apparently private economic decisions, and politically influential shareholders are reluctant to accept losses. Indeed, they will struggle vigorously to avoid any losses. But a financial system cannot function well as long as this is the case. New laws and regulations alone will not eliminate such struggles.

Furthermore, depository banks play a vital role in development, as trusted institutions where financially unsophisticated citizens are willing to place their savings. They represent a big social improvement over private savings held in the form of commodities such as gold bracelets, for they permit savings to be mobilized for development. Workers must be confident that they can retrieve their savings when needed. That confidence, in turn, requires either that the banks invest savings carefully or that the state guarantees those savings.

But banks in many developing countries have engaged in a massive violation of the trust that has been placed in them, by directing savings into the operating expenses of loss-making corporations, dubious "national champion" investments, real-estate speculation, or loans to politically well-connected individuals. Bank shareholders resist the implication that they should lose their stakes, and the struggle continues over who shall bear the losses that have already been incurred in financial crises. Until a clear and generally accepted hierarchy of responsibility is established, national financial systems will remain unreformed and will fail to achieve their social potential.

The international system may not be responsible for financial crises. But foreigners, having contributed to the preceding euphoria, may aggravate a crisis and increase its economic costs. Moreover, crises can spread from one country to another through a variety of channels, commercial as well as financial. So the international community might be able to help mitigate the damage. Every proposal for reform needs to be assessed with this possibility in mind. But it also must take into account the possible harm it can do, either by significantly retarding the international flow of beneficial capital or by encouraging incautious lending on the assumption that any trouble will be covered by international action.

Krueger's proposal for an internationally sanctioned standstill on sovereign debt, if cleanly and comprehensively implemented, would represent a modest improvement on existing arrangements. A plan to empower the IMF to issue SDRS in a financial emergency, under stringent conditions, would also represent some progress. Both would require amending the IMF articles, which is no easy task. But neither approach will banish financial crises for good.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now
  • Richard N. Cooper is Maurits C. Boas Professor of International Economics at Harvard University and a regular book reviewer for Foreign Affairs.
  • More By Richard N. Cooper