Reforming the International Monetary and Financial System

9 Key Issues in International Monetary and Financial Reform: A Personal Record of the Conference

Alexander Swoboda, and Peter Kenen
Published Date:
December 2000
  • ShareShare
Show Summary Details
Alexander K. Swoboda and Jeromin Zettelmeyer* 

The conference on Key Issues in the Reform of the International Monetary and Financial System had a twofold purpose: (1) to broaden the debate on international financial architecture to more general issues of international and financial reform and (2) to allow experts outside the usual policy forums, notably from academia, to contribute to that debate.

Many of the issues that underlie the agenda for strengthening the architecture of the international financial system are not new but arise in new guises with changing circumstances. A striking change in the environment over the last two decades has been the technological revolution in telecommunications and information systems that has underpinned and stimulated financial market integration and capital mobility as well as domestic and international financial liberalization. As a result, markets for goods, services, and assets are becoming ever more unified and developing economies have increasingly been drawn into these globalized markets. Private portfolio capital flows are playing a dominant role in the financing of current account imbalances in advanced countries and an ever-increasing one in financing, and sometimes causing the current account imbalances of developing economies. At the same time, policy is still made predominantly at the national level even though markets have become global.

The conference program (see appendix) was designed to examine the policy and institutional responses to the accelerating sequence of crises in the 1990s that are appropriate, at both the national and the international community level, in today’s world of increasing capital mobility. The first day was devoted to the theme of mitigating instability under high capital mobility or, if you prefer, under global finance. The first of four sessions was devoted to exchange rate volatility among the major currencies, a major manifestation of instability in today’s system; the second to another major source of instability, the boom-bust character of capital flows to emerging market economies; the consequences of the volatility of capital flows for the balance between adjustment and financing was taken up in the third session. The fourth session focused on one of the more contentious issues in the reform debate, the involvement of the private sector in crisis prevention and crisis resolution.

The second day was devoted more specifically to the actual and desirable role of the IMF in the changing international monetary and financial system. Two sessions dealt with this topic and examined the IMF’s role within the system as (1) a provider of financing and (2) a provider of advice. Rather than providing a linear summary account of the papers and discussion we instead offer below a somewhat personal summary of the discussion as it relates to the main issues that were taken up during the two days of the conference, irrespective of the order in which the discussion actually took place—although the organization of this chapter is not entirely dissimilar to that of the conference program.

Exchange Rate Regime Among Major Currencies

Although discussions of international financial architecture have generally focused on preventing and resolving emerging market crises, changes in the exchange rate regime among major currencies, and the issues they raise for national and international policy, have recently come to the fore in discussions of international monetary reform, notably with the creation of the euro. Briefly stated, there are three interrelated questions that were the topic of the first session of the conference and of the paper by Jean Pisani-Ferry and Benoit Cœuré: Are fluctuations among the major currencies, specifically the euro, the yen, and the U.S. dollar, likely to increase or decrease in the future; what are the costs of such fluctuations; what, if anything, can and should be done to moderate them?

On the first question, Pisani-Ferry and Cœuré saw little reason to believe that either short-term volatility or some medium-term “excessive” movements, or misalignments, in the exchange rates of major currencies would not continue, and may even increase in the future. The euro can be expected to be fairly volatile for at least three reasons: a system with multiple key-currencies is likely to be more volatile than one dominated by a hegemonic currency; with the introduction of a new major currency, portfolio shifts are likely to occur; and the euro zone is less open than its individual components and hence more inclined toward benign neglect in its exchange rate policy. Participants in the conference, somewhat surprisingly, appeared to agree with this diagnosis of continuing exchange rate volatility. What was less surprising was much less agreement on what the costs were and whether and what could and should be done about it.

Among the costs, of course, are those that will have to be borne by the issuers of the major currencies themselves. These seemed not to worry most participants much—in particular, they felt that short-term volatility in the exchange rate of major currencies could relatively easily be hedged against by the residents of these countries. That there maybe serious costs to major and lasting misalignments in terms of misallocation of resources was generally acknowledged, although there was fairly sharp disagreement on how serious such misalignments and their undoing were from a systemic point of view. Fred Bergsten argued that “very large misalignments cause huge, substantive problems and economic distortions and set you up for subsequent financial crises as the exchange rates reverse, as they always do—witness 1985–87—and cause an upsurge of protectionism that can profoundly damage the trading system.” A lively debate ensued with Rudiger Dornbusch, who claimed that, though the 1980s appreciation of the dollar appeared in the charts “as the most glorious bubble I have ever seen,” its unwinding did not result in the catastrophic outcome described by Bergsten. In fact, he argued, the real economy of the United States appears not to have been harmed all that much by the behavior of the dollar in the 1980s.

Two other costs of substantial misalignments were stressed by the authors of the paper. The first concerns the strains such misalignments can put on political institutions, in the European Union notably. However, the cost on which Pisani-Ferry and Cœuré as well as other participants put the most emphasis is the cost to third countries, notably emerging market and developing countries. As Pisani-Ferry and Cœuré put it, “the euro-dollar-yen rate has the character of a public-good for the world economy; this was obviously exemplified by the Asian crisis of 1997.” Indeed, this public-good character of the exchange rate regime among major currencies, and of international monetary arrangements more generally, was at the center of the evening address to the conference of the IMF’s Managing Director, Michel Camdessus. These fluctuations do pose a major problem for third countries, which will be taken up further below.

The central question is, of course, what can and should be done to moderate the volatility of major currencies’ exchange rates. The central idea of the Pisani-Ferry and Cœuré paper is to take what they call a two-handed approach to the problem which de-links policy coordination from exchange rate targets, as they find the Plaza and Louvre attempt to stabilizing exchange rates and coordinating policies through attainment of target exchange rates seriously wanting. In Pisani-Ferry’s words at the conference: “the link that was established between exchange rate targets and coordination did not withstand the test of time and, therefore, the two should be dissociated.” The paper sets out a modest approach to better coordination of macroeconomic policies among the G-3, which would essentially try to build a consensus on principles of macroeconomic policy response to various shocks. This would be, in Pisani-Ferry and Cœuré’s view, easier in the case of symmetric shocks affecting all countries in the coordination zone simultaneously, but some principles could also be developed for asymmetric shocks. The idea would be “to let different participants implement the same principles in different ways according to their own domestic policy-setting institutions.” The second element of the two-handed approach would be the monitoring of the foreign exchange market. More specifically, Pisani-Ferry and Cœuré would improve the provision of information to, and the processing of that information by, the foreign exchange market, for instance, by the collection and publication of some aggregated measure of foreign exchange positions and the reporting of abnormal risk exposure situations to the G-7 by the Chairman of the Financial Stability Forum. The IMF, in turn, would compute fundamental equilibrium exchange rates for the major currencies as a way to structure the dialogue at the G-7 and would make its estimates available publicly to provide a guide for the market’s expectations.

Discussants on the whole shared the skepticism of the authors of the paper as to the possibility and the desirability of attempts at exchange rate stabilization within the G-3 or G-7, including several variants of the target zone approach to policy coordination (although the main proponents of the target zone approach argued that the latter had been misrepresented by both authors and discussants; see below). Thus the possibility of targets as such was regarded as unrealistic. They generally welcomed the separation of policy-coordination issues and exchange rate monitoring stabilization. Dale Henderson congratulated the authors for identifying the main lesson we have learned in fifteen years: that “we cannot and should not view the coordination process through the exchange rate—we need to discuss macroeconomic policy.” Participants, however, were not very optimistic about the prospects for effective policy coordination among the major countries; nor did everybody feel that it would be appropriate for the IMF to provide estimates of fundamental equilibrium exchange rates.

Among the participants who spoke on the subject, Alan Blinder and Horst Siebert were perhaps the most skeptical about exchange rate stabilization attempts. Alan Blinder began by questioning the premise that fixing the nominal exchange rate is something desirable in and of itself. He argued that nominal exchange rate flexibility was necessary to allow sufficient real exchange rate flexibility to cope with international productivity growth differentials, terms of trade shocks, and lack of cyclical synchronization, a task that domestic wage and price movements cannot accomplish without large real output and employment costs. In addition, nominal flexibility is needed to accommodate inflation differentials at a constant real exchange rate. One problem is that such a floating system may indeed produce excessive short-run volatility. Blinder put it this way: “Exchange rates nowadays are much more like the price of IBM stock than they are like the price of IBM computers. They are going to bounce around a great deal because the asset tail completely wags the trade dog…. What we would like of course, if we could find one, is a remedy that makes exchange rates behave more like IBM stock and less like Yahoo! stock.” Blinder thought that the transparency and information measures suggested by the paper’s authors were on the whole useful, but he doubted that they would contribute very substantially to avoiding excessive short-term volatility in exchange rates. He expressed similar doubts with regard to devising principles of policy coordination that would be robust enough to prove durably effective. As for the public-good aspect of fluctuations in the dollar-yen-euro exchange rate, the best solution would be to let third countries float lest they fall victim to the “foreign-exchange bubble” syndrome.

Horst Siebert reached similar conclusions by a somewhat different route. He argued that the international coordination of policies required by stabilization of exchange rates along the lines of target zone schemes was just not achievable under present circumstances, and misguided at that. Far better, if one wants exchange rate stability, to let it be the outcome of price stability within each of the major countries. This would have the added advantage of putting responsibility clearly where it belongs. By emphasizing national responsibility for domestic price levels, with appropriate national policy institutions and procedures, “the price level remains stable in each country; but you cannot have two nominal anchors for monetary policy, and the price level should be the relevant anchor.”

A number of participants, however, argued for some form of stabilization of the major exchange rates. Ronald McKinnon focused on the yen-dollar rate, arguing that the liquidity trap to which Japan seems to be subjected today is the outcome of a decrease in interest rates in the United States coupled with an obdurate long-run expectation of appreciation of the yen that keep Japanese interest rates very low. In addition, the threat of protectionism makes it impossible to depreciate the yen. The result, according to McKinnon, is that “the Japanese economy is trapped both in terms of the interest rate being at zero and the exchange rate being about as low as tolerable.” The solution would be to stabilize not so much the current exchange rate but the future expected one so as to break the “ever appreciating yen” psychology. The staunchest defenders of stabilization of the exchange value of the major currencies were, not surprisingly, Fred Bergsten and John Williamson. Bergsten argued strongly that disillusion with the Plaza and Louvre arrangements, or the ERM crisis of 1992–93, does not imply that target zones have been tried and failed. These were not target zone experiments, but various forms of pegging attempts. Critics do not understand that “the target zone system intends to be a system of flexible exchange rates, managed by a different method than the ad hoc intervention that prevails today, in a world where nobody is going to float truly freely.” The aim is not so much to reduce short-term volatility, though the system may anchor expectations and thus help reduce volatility, but to avoid very large and exceedingly costly misalignments.

The immediate next question is what exchange rate regime is appropriate for third countries, emerging market economies in particular, given that volatility in the exchange value of major currencies will continue. This question is taken up below, with an emphasis on crisis prevention and management. Here it is discussed briefly in the more general context of the evolution of the system toward the extremes of the spectrum ranging from pure floating to hard pegs of the dollarization on currency board variety. Ricardo Hausmann indicated that the Pisani-Ferry and Cœuré paper had convinced him that nothing much would be done to stabilize major currencies, which would continue to exhibit a high degree of volatility. In addition, prospects for coordination of G-7 policies, along the lines suggested by the authors of the paper, did not appear very likely to him because information supplied to markets on foreign exchange positions would be hard to interpret, the appeal to “fundamentals” is more of an admission of ignorance on the part of economic theory than an effective guide to policy coordination, and principles for coordination would be difficult to establish and might serve chiefly to define how meetings would take place. Instability in currencies would continue with attendant swings in commodity prices and terms of trade, and one must ask why such fluctuations in major currencies should contribute to leading to very large output collapses, in the order of 10 to 15 percent of GDP in crisis countries. One reason is that these countries cannot borrow abroad in their own currency and cannot borrow for the longer term domestically, ineluctably leading to currency and/or maturity mismatches. There are two solutions, Hausmann argued: do away with mismatches through capital controls or adopt currency boards or dollarization. The latter does away with mismatches at the cost of real exchange rate fluctuations, but one may hope that the latter is less disruptive than the financial catastrophes that follow if one tries to maintain an autonomous currency. This will tend to lead to a world of fewer currencies in which the IMF may have to rethink its role as members of these enlarging currency zones (dollar, euro, and possibly some Asian currency zones) will turn increasingly inward for coordination purposes. Against this point of view, one may mention, as Tom Willett did, that capital mobility, though substantial, is not so perfect as to defeat all attempts at managed floats.

Characterizing Emerging Market Crises

Two conference papers—one by Mussa and others, and the other by Calvo and Reinhart—characterized recent swings in emerging market capital flows empirically and contrasted them with crises in other periods (the debt crisis of the 1980s) or countries (the ERM crisis). According to Mussa, Swoboda, Zettelmeyer, and Jeanne (henceforth Mussa and others), it is surprisingly hard to argue that the 1990s were a period of unique volatility of capital flows, largely because of the debt crises of the 1980s, which lasted longer and led to large cumulative losses. Recent crises in emerging market economies, however, elicited responses from the international community that, in selected cases, involved unusually large quantities of official financial support. Thus, there is little indication that, even with the relatively large official support packages provided in some recent crises, it is possible to avoid substantial short-run costs when crises do occur; but some suggest that by helping to avoid a prolonged disruption of normal international financial relations, recovery may be aided and longer-run costs may be reduced. This raises the question of whether the traditional trade-off between financing and adjustment needs to be rethought in the light of the experience of the 1990s (see Box 1).

According to Calvo and Reinhart, the emerging market crises of the 1990s were superficially similar to the ERM crisis in that both involved exchange rate collapses and (in the case of the European crisis, involving the Scandinavian group of countries) financial fragility. The emerging market crises were entirely different in terms of their consequences, however, in large collapses in output and, at least temporarily, losses in capital market access. Calvo and Reinhart attribute the severity of the output collapse to a “Fisherian channel,” which is complicated by the presence of “liability dollarization” (i.e., private and/or public foreign-currency-denominated debt). Characteristic of emerging market crises is a sharp capital flow reversal, which forces a sharp real depreciation. This can either be channeled through a collapse in nontradeables prices, including real estate, implying a sharp rise in ex post real interest rates that leads to a rise in nonperforming loans and possibly a banking crisis. Or it can be channeled through a nominal depreciation, which may mitigate the former effect, but then dollar liabilities become a problem, resulting in widespread balance sheet deterioration and a collapse of the domestic financial sector. Calvo and Reinhart present evidence on the severity of emerging market crises in the 1990s as measured by the extent of the capital flow reversal and by a combined index of real exchange depreciation and reserves loss, and conclude that the severity of the Asian crises of the 1990s even surpassed that of their Latin American counterparts and is a significant departure from their historical norm. In addition, they show that capital flow reversals that involved banking crises tended to be both more protracted and more severe in terms of cumulative GDP loss than those that did not, as one might expect in the presence of a “Fisherian channel.”

Box 1.The “Financing versus Adjustment” Framework in Light of Recent Crises

Montek Ahluwalia and Jeffrey Frankel both pointed out that “financing versus adjustment” traditionally describes the choice facing policymakers in an open economy affected by an adverse external or real shock (for example, a deterioration in the terms of trade). When faced with such shocks, maintaining consumption and the current account at unchanged levels (in other words, not adjusting) requires borrowing abroad (financing). Thus a trade-off exists between financing and adjustment. However, the nature of this trade-off critically depends on whether the shock is transitory or permanent. For permanent shocks, borrowing does not lessen the need for adjustment, because without any adjustment loans could never be repaid. In contrast, if shocks are transitory, it is optimal to finance them, that is, to only marginally adjust consumption at the time of the shock, spreading the adjustment over a long term.

How does the experience of the crises of the 1990s fit into the “financing versus adjustment” framework? To the extent that external or real shocks contributed to the onset of the crises (which was not always the case), these shocks were small relative to the huge capital flow reversals that followed. Except for official borrowing and reserve holdings, “financing” was thus not available to smooth the shock. On the contrary, capital outflows served to magnify it. In this setting, if reserves and official financing are small relative to the volume of potential private outflows, the choice between adjustment and financing is not available, in the sense that “financing is not an option, and the only way of adjusting is through a recession” (Frankel). A further complication, noted by Peter Kenen, was that the fall in income required to achieve a given current account adjustment was probably larger than in the past, as dollar-denominated liabilities in the corporate sector implied that the exchange rate depreciation had a contractionary impact, in addition to its traditional expenditure-switching effect. As a result, adjustment in the Asian crisis was “dysfunctional” in the sense that it occurred mainly through expenditure reduction rather than expenditure switching.

Even in the new world of balance sheet effects and volatile private capital flows, however, “financing” may still play its traditional role, provided that the volume of potential official financing is large and private outflows take the form of a financial run, as noted by Montek Ahluwalia, In this case, adjustment will only be required to the extent that there are fundamental weaknesses in addition to liquidity problems. Thus, in the context of a liquidity crisis, the familiar alternative between adjustment and financing reemerges, except that the latter must now be understood as official financing only. The critical distinction is no longer between transitory and permanent shocks but rather between liquidity and solvency as the cause of the crisis. While these distinctions are closely related—they both hinge on whether the underlying problem is self-correcting or not—there is a complication in that a liquidity problem may be self-correcting if and only if there is enough official financing, unlike a transitory shock, which is exogenous to the amount of financing available.

Finally, Jeffrey Frankel pointed out that the traditional trade-off between adjustment and financing (taken to include private borrowing) might still be relevant during the time frame in which signs of weaknesses begin to emerge, but capital flows have not yet reversed. In this period, policymakers may have the choice to ignore these weaknesses and borrow (or draw down reserves), or address them by adjusting, including (as Yoon Je Cho remarked) through structural adjustment. However, exploiting this trade-off in the direction of excessive financing could be damaging because it “improvidently postpones adjustment” (Kenen), and when the crisis comes, the situation is much worse.

A related topic that received substantial attention was the relative roles of “fundamentals” and “systemic” weaknesses in causing the crises of the 1990s, with the former including domestic policy and regulatory failures, and the latter including a tendency of advanced integrated capital markets to overshoot, both on the inflow and on the outflow side. Two papers—Mussa and others, and in less detail, Lipton—described the crises of the 1990s in these terms,1 and took the position that there was some merit in both explanations. One way of integrating the two views was suggested by Rudiger Dornbusch in his comments on Mussa and others. According to Dornbusch, one should distinguish between financial crises resulting from unsustainable policies (such as growing macroeconomic imbalances, which in the past has been typical for Latin America) and situations of vulnerability “that can last forever, but which also can all of a sudden be brought into question.” The latter, exemplified by the Asian crises, are driven by firms’ and bankers’ balance sheets rather than the aggregate current account imbalances. Both types of crises involve fundamental weaknesses, but crises of the balance sheet type require a liquidity crisis as a trigger, and excessive inflows play a role in facilitating the buildup of balance sheet vulnerabilities in the first place. Thus—at least for the crises that involved financial fragility rather than, or in addition to, macroeconomic imbalances—fundamental and “systemic” weaknesses are hard to disentangle.

While almost all speakers on this issue embraced the view that the crises of the 1990s contained both a fundamental and an important systemic component, there was wide divergence of opinions on the relative importance of these components. In Martin Wolf’s view, recent financial crises ought to lead to “a recognition of very substantial [systemic] failure…. I think a system that forces countries to make current account adjustments of close to 20 percent of GDP in two years, a system in which net lending by banks can shift by well over 10 percent of GDP in a year or two, a system that brings fears of worldwide recession only a few months ago out of cloudless sky, is guilty of serious malfunction.” Similarly, Yung Chul Park professed to belong to the “financial panic” (i.e., systemic) school of thought. David Folkerts-Landau, on the other hand, took the position that the crises of the 1990s were prominently driven by a “massive failure of national policies in many key emerging countries,” in addition to “financial linkages across the emerging markets themselves,” such as Korean merchant banks taking positions in Brazil and Russia “on the back of official guarantees.” Even Folkerts-Landau, however, acknowledged in his spoken comments that the international financial system, while not the primary cause of the crisis, “is not blameless.” In this context, he mentioned both international bank lending behavior, which had played a significant role in the liquidity crises, and technical factors, such as investor behavior dictated by risk management techniques.

In the discussion, both Guillermo Calvo and Michael Mussa took issue with Folkerts-Landau’s emphasis on the preeminence of policy mistakes in emerging market countries in causing the crisis. Calvo emphasized that capital flows to emerging markets had a tendency to amplify the consequences of policy mistakes, adding market failure to policy failure. Mussa pointed out that the run-up to the crisis, in which interest rate spreads underwent large declines just as capital inflows peaked, was clearly driven by the suppliers rather than “demanders” of capital. Similarly, the massive panic when the crisis hit was a very important part of the crisis, which deserved greater recognition than it had received so far in the financial architecture debate.

Exchange Rate Regime for Emerging Market Economies

The question of which exchange rate regime is best for emerging market economies provoked a lively debate at the conference, inspired by the Lipton and Calvo and Reinhart papers, which presented widely divergent views on the subject. David Lipton argued the classic case in favor of flexible exchange rates as a way of reducing the volatility of capital flows to emerging market economies and avoiding costly financial crises. He started from the observation that increasing integration of emerging market economies into global finance has both made the failure of pegged regimes more costly and their defense more difficult. The latter results from the sheer magnitude of swings in capital flows that emerging markets can now be exposed to. The former arises from the fact that the ensuing overshooting of the exchange rate—reflecting a loss in confidence after the regime has failed—transforms the currency crisis into a national bankruptcy crisis by raising the local currency burden of foreign debt.

In this setting, a floating regime has two important advantages. First, faced with sharp outflows, a country will be less likely to squander reserves in an attempt to maintain the original parity. From the perspective of the international community, this is good, since it means that countries will be less likely to ask for bailouts. Second, and perhaps more important, flexible exchange rate regimes are likely to reduce the likelihood that countries are attacked in the first place, for several reasons. In the absence of an official exchange rate guarantee, and faced with day-to-day volatility of exchange rates, borrowers will have an incentive to hedge their foreign exchange exposures. This prevents balance sheet vulnerabilities of the type that led to financial crises in Asia and elsewhere. In addition, economic shocks would lead to a continuous adjustment of the exchange rate, and thus avoid catastrophic “lumpy” adjustments in the form of currency crises.

Lipton’s position on exchange rate regimes was broadly shared by several other speakers. Peter Kenen emphasized that one problem with pegged exchange rates was that they imparted an incentive to ask for international assistance only after it was too late, that is, after a currency attack had been successful and the country had squandered most of its reserves. Jack Boorman, while agreeing with Lipton that more exchange rate flexibility was desirable, qualified the case for floating regimes in two ways: first, by reminding participants that there had been instances where pegged regimes successfully imparted discipline on policies (such as in the context of stabilizing from high inflation), and second, by arguing that “markets can go to sleep” even in the presence of floating regimes, and as such, a floating regime does not fully protect countries from a massive swing in confidence.

A contrarian view was presented by Guillermo Calvo based on his paper with Carmen Reinhart. The essence of Calvo’s argument is that the classic case in favor of floating regimes—while logically correct—is irrelevant in practice, because most countries with floating regimes face incentives that induce them to stabilize their exchange rates, that is, to not actually let their currencies float in the spirit of the term. The main reason for this unwillingness to let the currency float is that dollarization—in particular, liability dollarization, or foreign-denominated debt—is widespread in emerging markets regardless of the exchange rate regime, implying that swings in the exchange rates impose large risks on balance sheets. Another reason is reliance on imported raw materials. Either way, the result is a “fear of floating,” which in some cases—El Salvador, the Philippines and Venezuela—has made “floating” regimes essentially equivalent to soft pegs. This, in turn, creates further incentives for liability dollarization, by the same logic emphasized by critics of explicit pegs.

Calvo stressed that in the presence of liability dollarization, a country will be vulnerable to an attack regardless of whether the exchange rate regime is a peg or a float. The only way of removing this risk is full dollarization, which turns dollar-denominated liabilities into domestic-denominated liabilities. However, Calvo recognized that there were a number of arguments against full dollarization (part of which coincide with arguments against pegged regimes), which he discussed as follows:

  • Loss of the exchange rate as a relative price. Calvo played this down on the grounds that “in a realistic economy, there are several distinct goods, each with a distinct labor market,” while the exchange rate constitutes only one relative price: “devaluation makes no group totally happy.” In addition, devaluation could be substituted by other instruments, such as fiscal policy or a wage subsidy.
  • Loss of a policy instrument. According to Calvo, the monetary and exchange rate policy is an instrument that most emerging markets should be happy to lose, since in the presence of “fear of floating” it turned out to be mainly a procyclical one (an adverse shock leads to tighter monetary policy to prevent exchange rate weakening).
  • “The Fed will be thy Lord.” Calvo cited evidence that the Fed policies already have substantial impact on emerging market economies, including those with supposedly floating regimes. “Well, the Fed is our Lord whether we like it or not.”
  • Use of foreign money entails loss of seigniorage. “The two countries involved could share the seigniorage.”
  • Dollarization leaves a country without a lender of last resort. This would hold true only to the extent that the lender of last resort has to rely on money issuance. “However, under dollarization and seigniorage sharing a large portion of international reserves could be used to provide lender-of-last-resort services.”

While several speakers (see above) felt that Calvo and Reinhart had overstated the case against more flexible regimes and the ease with which the role of the exchange rate could be replaced by other policy instruments, others (including Jeffrey Frankel and Ronald McKinnon) expressed sympathy with their views. McKinnon particularly agreed with the notion that the scope for free floating is severely limited in practice, both for trade and for balance sheet reasons.

Finally, some speakers, including Takatoshi Ito in his paper on IMF advice, and John Williamson in his comments on Ito’s paper, defended “intermediate” exchange rate arrangements, including exchange rate bands, against the proponents of either flexibility or full dollarization. Like Boorman, Ito’s paper reminded the reader of the usefulness of temporary pegs to stabilize from high inflation. Williamson expressed the view that the exchange rate band had worked well in Indonesia, and took issue with what he viewed as a tendency to assess exchange rate regimes solely from the vantage point of avoiding crises: “There are other things in the world, like having a high rate of growth. And I find it difficult to believe that the East Asian countries would have had 25 years of miracle growth with floating exchange rates.”

The Role of Capital Controls

The macroeconomic dilemma which may lead policymakers to resort to capital inflow controls was compellingly described by Oldrich Dedek and Roberto Zahler, and later echoed in remarks by Yung Chul Park. According to Zahler, in the presence of open capital accounts, countries with sound macroeconomic management can easily become victims of their own success. As perceived country risk declines, countries with low levels of foreign debt and high domestic rates of return will increasingly be viewed as attractive. The ensuing debt and equity portfolio inflows lead to rising consumption and investment, rising prices, and a growing current account deficit. As external liabilities accumulate and the real exchange rate appreciates, expected dollar returns fall and perceived country risk begins to rise again. Ideally, this would lead to a gradual undoing of the previous boom cycle. But, more often than not, the reversal in capital inflows is sudden, triggered by bad news, or a shift in market sentiment following a realization that the country has become vulnerable to a liquidity crisis.

As Dedek and Zahler both emphasized, forestalling this sudden shift poses a major challenge for macroeconomic policy during the boom phase. Monetary policy is caught between the conflicting objectives of dampening domestic demand by increasing interest rates and fending off inflows by lowering them. Allowing more exchange rate flexibility is hardly a panacea: its first-order effect goes in the wrong direction if the objective is to fend off excessive real appreciation, while the effect of higher exchange rate volatility as a deterrent to short-term inflows (a mechanism that was also pointed to by other speakers, including Guillermo Ortiz) may come at a substantial economic cost.2 Sterilized interventions are costly for the central bank and effective only in the short term. Fiscal policy tends to be too rigid, and the reduction in expenditures required to offset a large inflow can easily go beyond what is politically and socially tolerable and hurt long-term growth to the extent that it involves public investment. In essence, this leaves taxes on inflows or other controls as a residual option. Dedek and Zahler expressed very different opinions on the usefulness of this option, however. According to Dedek, the Czech experience with capital controls had not been very encouraging, as markets were quick to work their way around them. His view was echoed by J. de Beaufort Wijnholds in a later session, citing the European, and particularly German, experience in the early 1970s. However, according to Zahler (who was later supported on this point by Barry Eichengreen), inflow controls can be effective if the authorities are careful to integrate them into a sufficiently broad strategy of regulating inflows. In Chile, this strategy had rested on three pillars: first, a one-year time requirement for the repatriation of foreign direct investment; second, regulatory requirements which reduced the speed at which Chilean firms could obtain financing in foreign markets (inter alia, related to the firms’ credit ratings); and finally, a noninterest-bearing reserve requirement covering the first year that foreign funds were held in the country, regardless of their effective duration.

Implicit in Zahler’s comments was the notion that to the extent that inflows regulation in Chile was effective, this was related to its integration in a process of gradual capital account liberalization. In other words, the point of departure was one where the capital account had been fairly closed, and the inflow taxes and other regulatory requirements introduced thereafter represented a less restrictive, and more market-friendly, means of controlling inflows. This point was echoed in a later discussion by Montek Ahluwalia: “Once you get rid of capital controls, it is probably not a good idea to stick them on again. But the impression is that if you have a few capital controls, it is not necessary that you should be dismantling them very quickly.”

The potential role of capital controls in mitigating the volatility of international capital flows was also discussed in Calvo and Reinhart’s paper from the perspective of crisis prevention and by Mussa and others from the perspective of both prevention and resolution. Calvo and Reinhart cited evidence from a paper by Montiel and Reinhart (1999) that indicates that capital controls had no statistically significant effect on the total volume of inflows, but altered the composition of inflows away from short term and portfolio inflows and toward FDI. At the same time, however, Carmen Reinhart mentioned several caveats. The apparent effect of inflows on the maturity composition of debt could, to some extent, be an artifact of reclassification. More important, capital control measures might simply have led to a substitution of domestic short-term debt for foreign short-term debt. To the extent that domestic short-term debt is also an implicit claim on the reserves of the central bank, such a substitution would not ameliorate the liquidity problem after a sudden capital flow reversal. It follows that a tax on all short-term borrowing may be a preferable strategy to just taxing foreign short-term borrowing. “Thus, governments that attempt to pursue capital controls will likely be driven to cast a wide net which covers all financial intermediaries, and even nonfinancial corporations, since the latter participate in the sizable interenterprise credit market. This is an enormous task. Moreover, countries that succeed in this task may find themselves deeply immersed in central planning.”

Mussa and others made two general points on inflow taxes as a preventive device. First, Chilean-style inflow taxes involve a similar trade-off as that associated with the provision of any public good. In this case, a reduced risk of liquidity crises is paid for by distortionary taxation. Depending on the country and situation, the cost may or may not be worth the benefit. Second, reliance on short-term credit may be an equilibrium response by both borrowers and lenders to asymmetric information (in particular, lack of transparency on the side of borrowers). Intuitively, short-term lending could be regarded as the market’s way of providing finance while at the same time keeping the borrower “on a short leash,” that is, imposing market discipline.3 In this situation, taxing short-term inflows may not make sense, as short-term financing may be the only way for emerging markets to obtain finance for certain uses, including worthwhile ones. Instead, reliance on short-term finance should be reduced by addressing its underlying causes. This can include the adoption of standards for greater transparency and sound management of public debts, and strengthened regulation and supervision of private borrowers.

Outflow controls as a way of resolving crises, according to Mussa and others, involve an entirely different set of arguments. In a liquidity crisis, imposition of controls on foreign currency payments on the principal of outstanding loans (and possibly other capital outflows) might serve a useful purpose by protecting the exchange rate against massive depreciation that would force many debtors with substantial foreign-currency liabilities into effective insolvency. Conceivably, even foreign creditors might see some advantage in this solution, provided that interest continues to be paid and there was the reasonable expectation that principal repayments would not be long interrupted. However, there are a number of important limitations and drawbacks. First, capital controls will do little to resolve a crisis if borrowers are insolvent even in terms of their capacity to meet obligations in domestic currency, or at the long-run equilibrium exchange rate. If the underlying problems are deep and persistent, then the imposition of controls on capital outflows will provide only temporary respite. Second, the imposition of controls to forestall a situation of potential default (in contrast with the maintenance of controls already in place) may be regarded by creditors as much the same thing as an outright default, leading to the same problems of prolonged disruption of access to international credit markets. On the side of creditors, this may be a rational response if liquidity and solvency problems cannot be neatly differentiated, that is, if liquidity problems are taken to be a signal that the debtor may have more persistent difficulties in meeting payment obligations. Finally, according to Mussa and others, “in an environment of increasing capital market integration, controls on outflows must be drastic to be effective, and for that reason require a comprehensive apparatus of restrictions ready to be activated in the event of a large capital flow reversal. It is hard to imagine that countries that consistently pursue a long-run objective of integration would wish to pay the reputational costs associated with the maintenance of such an apparatus on a permanent basis.”

Except for the general rationale for outflow controls put forward (but immediately and strongly qualified) by Mussa and others, no conference participant attempted a defense of outflow controls. A cautionary note was sounded by Roberto Zahler, however, who argued that although “as a norm, capital outflows should be free, prudential considerations suggest a strategy of gradual and selective opening up” for some financial institutions, namely those that “operate with a de jure or de facto government guarantee.”

Involving the Private Sector in Crisis Prevention and Resolution

Along with the closely related issue of international moral hazard and the role of official crisis lending (see below), the extent to which the private sector should be involved in crisis prevention and resolution proved to be the most controversial and intensively discussed topic of the conference. Barry Eichengreen’s paper was entirely devoted to the topic; two other papers, by David Lipton and Michael Mussa and others, discuss it in less detail and in a broader context; four panel statements made it their main focus; and several others raised it more briefly.

Two main approaches emerged in the course of the debate, basically divided over automatic versus voluntary bail-in for the private sector. The first, roughly in line with statements by Barry Eichengreen, David Lipton, and J. de Beaufort Wijnholds, can be broadly defined by three ideas: (1) greater involvement of the private sector in crisis prevention and resolution is desirable and necessary to reduce the reliance on large-scale official crisis assistance; (2) if appropriately designed and implemented, certain bail-in measures—in particular, orderly workout procedures and innovations in bond contracts—have the potential to significantly reduce this reliance; and (3) some prodding—or at least a lead role—of the international official sector in getting the private sector and emerging market borrowers to adopt such procedures is justifiable and probably unavoidable to move beyond the status quo.

In contrast, a second approach, which roughly describes the position taken by William Cline and Pablo Guidotti, argued that (1) while greater private sector involvement may be desirable in some cases—particularly from the perspective of more efficient crisis resolution and to accelerate the resumption of capital flows after the crisis—the international community’s best hope in crisis prevention lies elsewhere (namely, in better policies and risk management on the side of countries and creditors); (2) bail-in measures will not, in general, obviate the need for official crisis assistance, except perhaps if they take the form of commercial credit lines in combination with a much improved domestic debt structure (as emphasized by Guidotti); and (3) private sector involvement should be “as voluntary as possible” (Cline) and avoid official prodding. Any official prodding, such as through IMF conditionality, should be directed at emerging market country policies directly rather than at the terms of bond contracts (Guidotti).

A number of speakers expressed views that were not encompassed by either approach, but even here the two “philosophies” outlined above provide useful reference points. Jack Boorman and Michael Mussa were close to the Eichengreen-Lipton-Wijnholds school in taking the view that some prodding of creditors and debtors into greater private sector involvement may indeed be necessary, but they were more skeptical of the potential of private sector involvement, desirable as it may be, in obviating the need for official intervention (see below). Other speakers took more extreme views on both sides of the spectrum. Compared to Eichengreen, Lipton, and Wijnholds, Rudiger Dornbusch and Charles Calomiris were possibly even more concerned about moral hazard, and less concerned about the adverse consequences of debt workouts. In addition, while Eichengreen, Lipton, and Wijnholds all came down in favor of a case-by-case approach to private sector involvement in crisis resolution (although Lipton also thought that laying down general principles for conducting workouts would be desirable), Calomiris stressed the importance of clear ex ante rules: “Ambiguity is an invitation to political manipulation in the guise of political coordination.” On the other side of the debate spectrum, David Folkerts-Landau went further than either Cline or Guidotti in rejecting changes in sovereign bond contracts as unhelpful to stabilizing emerging market capital flows (see below).

The differences in basic outlook were reflected in participants’ views about specific areas in which bail-in mechanisms have been proposed. They are briefly summarized as follows:

  • Most speakers—including all those who subscribed either to the two main approaches outlined above or some combination of the two—were opposed to involuntary bail-in mechanisms such as “mandatory haircuts,” to put options in interbank loan contracts, and (with somewhat less force) to Willem Buiter and Anne Siebert’s universal debt-rollover option (UDROP).4 The main concern, developed in both Eichengreen and Mussa and others was that while these mechanisms might help reduce moral hazard on the side of creditors and—in the case of put and rollover options—prevent pure liquidity crises, “application of these mechanisms on a regular basis may increase the incentive for creditors to flee from a country at the first signs of trouble…. This logic does not apply just to extreme proposals such as ‘mandatory haircuts,’ but to any scheme that increases the expectation that private investors will have to bear a larger share of losses in the event of a crisis” (Mussa and others). Eichengreen pointed out that this problem extended to Buiter and Siebert’s UDROP, as borrowers would generally have incentives to invoke the rollover option not just in a liquidity squeeze but in the event of solvency problems, in order to gamble for redemption. In this case, “the costs of creditors who are late to exit are likely to be larger in the presence of UDROPS than in their absence,” suggesting that “UDROPS may aggravate the risk of crises for countries with weak fundamentals.”
  • While no participant opposed the idea of commercial contingent credit lines outright, several speakers were skeptical of the ability of such arrangements to provide substantial net liquidity to a country in times of crisis, on the grounds that private banks, “at the same time they provide additional credits, can draw down their exposure to the country or sell short government bills and bonds” (Eichengreen). In contrast, those who tended to be more skeptical of bail-in proposals expressed “considerable enthusiasm” (William Cline) toward commercial contingent credit lines, mainly because of their voluntary nature. Pablo Guidotti suggested that the net liquidity provided by commercial contingent credit lines could be quite sufficient in conjunction with a sound debt and liquidity management structure on the side of the country, to keep the country’s short-term debt low and avoid the need for official financial assistance.

Perhaps not surprisingly, the most controversial class of proposals regarding private sector involvement turned out to be innovations in bond contracts such as majority voting, sharing, nonacceleration, and collective representation clauses. On one side of the debate, Barry Eichengreen, generally supported by J. de Beaufort Wijnholds, viewed “the addition of such clauses to bond contracts as the only practical way of creating an environment conducive to flexible restructuring negotiations” as an alternative to “ever-bigger bailouts.” Moreover, because any individual country would be reluctant to take the first step in adopting such contracts for fear of creating a bad precedent, it was essential that the introduction of such clauses be coordinated and encouraged (say, via IMF conditionality) by the international official community. On the other side of the debate, David Folkerts-Landau argued that the key to reducing the volatility of capital flows to emerging markets in the first place was to create a “stable investor base” that did not exist at present, with many investors routinely exiting (and reentering) the market entirely. To the extent that changes in bond covenants reduced the rights of individual investors, they would be outright counterproductive: “What is badly needed is to create a stable [investor] base for emerging market debt—not debt with special features that curtail investors’ rights and expose them to additional liabilities.” William Cline advocated a somewhat milder version of this argument, welcoming “experimental” changes in bond clauses, but worried that mandating such changes might have disruptive consequences on capital flows. Finally, Mussa and others took an intermediate position, arguing that procedures that facilitated workouts could play a useful role ex ante in mitigating moral hazard, but expressing doubts whether better bond contracts really had the potential of paving a “third way” for crisis resolution between outright default on the one hand and large-scale official support on the other: “Arguably, bond contracts that specify qualified majorities for approval of restructurings and other similar reforms might make the resolution of future national defaults somewhat easier. But, in situations where creditors are asked to accept substantial losses or where there are many creditors with differing claims against many debtors, resolution of national defaults is unlikely to be quick or easy.”

In terms of the underlying rationale, the controversy on innovations in bond contracts can be traced back to disagreements on two fundamental issues:

First, what is the importance of sustaining a high level of capital flows to emerging market economies relative to reducing capital flow volatility? William Cline and Pablo Guidotti felt that the former should be a paramount objective. In Cline’s view, the central challenge in international finance remains the problem of creating “effective collateral substitutes” in the absence of either global bankruptcy courts or sovereign collateral, so as to overcome the financial system’s inherent bias toward international flows that are too low. It follows that proven “collateral substitutes” such as the creditor protection embodied in existing bond contracts should better not be tampered with. In contrast, Eichengreen, Mussa, Wolf, and many other conference participants viewed the central challenge of financial architecture as how to mitigate excess inflows as well as excess outflows. Thus, a reduction in the average volume of international borrowing may be acceptable if it reduces volatility and contributes to avoiding crises. As provocatively put by Martin Wolf: “Because there is so much difficulty in handling sovereign foreign-currency indebtedness, there should be much less of it.” Mussa and others suggests furthermore that the primary “collateral substitute” at work today is the ability of capital markets to deny a debtor access following debt restructuring (i.e., an adverse reputation effect), rather than the legal terms of a bond contract as such. However, the likelihood of an adverse capital market reaction to a country’s failure to keep its original commitments need not be tangibly affected by the nature of the contracts, that is, whether or not these commitments are renegotiated in an orderly way. This explains why the issue of changes in bond contracts is treated as a sideshow by Mussa and others, in contrast to both Eichengreen and Cline.

Second, are innovations in bond contracts of the type advocated by Eichengreen likely to raise the cost of borrowing (and thus reduce the volume of capital flows)? For Pablo Guidotti (and, implicitly, William Cline) the answer is a clear “yes”; and as implicit from the previous point, both Cline and Guidotti would view this increased cost as a bad thing. The paper by Mussa and others is noncommittal, but takes the view that if such mechanisms do raise the ex ante cost of borrowing, this need not be bad but could be just what was intended (namely, to have the cost of borrowing better reflect its risks). Finally, Eichengreen argued that there may not be any increase in the cost of borrowing at all. From a theoretical point of view, there were two opposing effects: while the probability of some form of workout may go up, the magnitude of any eventful disaster, and consequently the overall losses that must be distributed, might well go down. “As The Economist put it in a recent leader, ‘the prospect of an orderly renegotiation rather than a messy default might actually make some bonds more attractive.’” As an empirical matter, Eichengreen observed that some British-style bonds, which already incorporate some of the desired clauses, do not seem to command higher spreads than their American equivalents, although he acknowledged some methodological difficulties in coming up with a firm conclusion.5 Thus, in Eichengreen’s view, facilitating workouts by appropriately modifying bond contracts may very well both reduce the cost of capital and reduce moral hazard by making large-scale rescues less likely.

A final contribution to the debate on private sector involvement was made by Martin Wolf, who observed that to the extent that emerging market crises increasingly involved problems of private rather than sovereign debt, a natural forum in which to “bail-in” private creditors was national bankruptcy courts, adding urgency to the desirability of developing bankruptcy procedures in emerging market countries. He also noted, however, that bankruptcies in the private financial sector were very often quasi-public, limiting this approach to smaller institutions or the corporate sector.

International Official Assistance and the Role of the IMF

Virtually all speakers agreed that the role of the IMF should comprise at least two components: (1) “catalytic” lending—that is, noncrisis lending in small amounts to serve as a stamp of approval and to encourage private capital flows to developing countries; and (2) liquidity support to stop crises from spreading to “innocent” countries—that is, to prevent contagion. In this context, several speakers welcomed the new contingent credit line (CCL) as a “potential tool for crisis prevention” (Lipton), but some had doubts about whether it was up to the task in its current form. Charles Calomiris criticized the CCL on two grounds—first, access was not automatic following qualification but depended on an additional “subjective ex post evaluation,” and second, the criteria for qualification were not clearly specified in terms of easily verifiable standards of behavior. David Lipton, in addition, criticized the lack of hard access limits, which made the facility vulnerable to “a bargaining process in which the private sector attempts to pressure the international community into upping its ante in the prevention sweepstakes.” While recognizing the logic of this argument, Jack Boorman doubted that rigid access limits would be believed by markets.

Apart from this discussion, the controversy on the role of official financial assistance in general and of the IMF in particular concentrated on two points (see Box 2 for an overview of tasks that speakers believed should be within the IMF’s responsibility):

  • First, should the IMF, and the international community more generally, continue to provide emergency financing to countries who may not be innocent in allowing a crisis situation to arise but are likely to regain solvency in the medium term and/or are sufficiently large to pose a systemic threat, as it did in the case of the Mexican and Asian crises?
  • Second, should the international community make an effort to create new facilities or institutions to thwart “truly systemic” threats, i.e., a large pool of resources to be used in the event that the international financial system as a whole is endangered?

Finally, a portion of the conference was devoted to discussing the quality of IMF policy advice in the past, with an emphasis on IMF programs in Asia. In what follows, the discussion on the above three topics is summarized in turn.

Box 2.What Should the IMF Do?

The following list presents options put forth in the conference regarding possible roles for the IMF (Points 2 and 4 are elaborated on in the body of the text):

  • 1. certification of good policies/catalytic lending,
  • 2. conditional crisis lending to individual countries,
  • 3. contingent lending to innocent victims of contagion,
  • 4. global last-resort lending,
  • 5. elaboration of broad guidelines for conducting international debt workouts,
  • 6. facilitation of swap arrangements among central banks (as a means of pooling official liquidity),
  • 7. design/administration of a system of controls on short-term capital movements,
  • 8. warning markets about policy developments,
  • 9. warning policymakers about markets developments, and
  • 10. enhanced surveillance, with the capacity to take “remedial action.”

Large-Scale Crisis Assistance to Emerging Market Countries

The case for discontinuing Mexico- and Asia-style crisis lending was made by David Lipton in the context of a broad strategy for reform of the international financial system. In Lipton’s view, the central role that the international official community has assumed in the last several years to stabilize emerging market crises is ultimately undesirable, particularly because of its moral hazard implications. Importantly, Lipton took the position that in the cases of Mexico and Asia official crisis lending was justified, given the exchange rate regimes that prevailed at the time (prior to the crises) and the absence of orderly workout procedures to better involve the private sector. (In this, he disagreed with Charles Calomiris and Rudiger Dornbusch, who stated that debt restructuring would have been preferable to large-scale financial assistance even in the case of Mexico.) In the future, however, “the international community should end IMF large-scale bailouts for member countries with balance of payments problems.” In Lipton’s view, this is feasible if “the burden for coping with economic problems is shifted away from the IMF and to the markets and private sector participants,” in particular, by encouraging the adoption of flexible exchange rate regimes. At the same time, general principles—but no fixed rules—should be defined for private sector workouts that would allow a private-sector-led crisis resolution when crises do occur.

Lipton’s view—or at least its main conclusion, that is, that large-scale country emergency financing could and should be eliminated entirely—was questioned by Jack Boorman, Montek Ahluwalia, and William Cline in comments on Lipton’s paper, and also contradicted by some of the arguments and conclusions developed by Michael Mussa in his presentation. The views of these critics can be summarized in four points:

  • (1) Crisis prevention, including greater reliance on flexible regimes, will not reduce the risk of crises as far as Lipton hopes. In Jack Boorman’s words, “David [Lipton] puts more faith in the capacity of floating rates to limit the buildup of debt problems than I think we can conclude from history.”
  • (2) Once a crisis erupts in a large emerging market country, it can pose a substantial systemic risk. One cannot presume that the presence of the CCL will entirely shield economies that are linked to the crisis country through trade and financial channels from adverse spillovers. Thus, the distinction between intervention for “national” and “purely systemic” purposes, which is central to the Lipton paper, cannot, in practice, neatly be made. In the words of Montek Ahluwalia: “The most successful case of intervening in a systemic crisis is one where the crisis doesn’t actually become systemic. So I think this whole issue of whether lending to a particular country amounts to simply intervening in a national crisis or becomes a case of preventing a systemic crisis is something which, in practical terms, is not always easy to determine.”6
  • (3) Even ignoring systemic spillovers, there may be cases in which international crisis lending is clearly welfare-enhancing when compared to the alternatives (capital controls or debt restructuring). Martin Wolf put this in simple and somewhat provocative terms: “If a country like Korea can be given a very large amount of money up front at probably a penal interest rate and you have a very high chance—in fact, in my view, a virtual certainty—that a few years later it can all be repaid, and this meets the immediate run on Korea … why on earth should we not do it?”
  • (4) To the opponents of large-scale financial intervention, the answer to this question may be obvious: “because it creates moral hazard.” However, to the defenders of traditional intervention, moral hazard may be a serious matter (see Box 3), but not one that necessarily outweighs the benefits of preventing a systemic crisis or a national default. As Michael Mussa put it: “If we hadn’t rescued the 800 people we rescued from the Titanic, we would have taught an even more valuable lesson about people being careful about getting on ocean liners. But there is a trade-off here: if every time national default threatens, we say, “let’s force it,” then we are not only going to get the creditors; we are also going to do a lot more damage to a lot of innocent victims.”

In summary, participants’ attitudes to large-scale IMF crisis lending largely hinged on how they assessed the evil of international moral hazard (Box 3) relative to the evil of hurting innocent bystanders—within the crisis country and internationally—and to what extent they believed that either of these problems could be ameliorated by other means, such as private sector workouts, exchange rate flexibility, the CCL, and improvements in domestic policies. Those who viewed moral hazard as the most critical issue concluded that “the IMF must foreswear such activities institutionally” (Calomiris). To those most concerned about mitigating the disruptions caused by financial crises, “a role remains for the old-fashioned medicine” (Boorman), in addition to a greater emphasis on prevention and on involving the private sector.

Large-Scale Facility for Preventing Global Liquidity Crises?

According to David Lipton, “while there should be no more bailouts for countries, the IMF should be equipped to respond if the health and integrity of the international financial system is endangered. That means creating a large pool of resources as a last line of defense, and creating a governance structure for that pool that ensures a very high threshold of systemic threat for activation.” Lipton motivates this proposal with the threat that arose to the international financial system in the wake of the Russian crisis and the Long-Term Capital Management (LTCM) debacle: “For a month or so, there seemed to be a potential for crisis to spread to a number of large developing countries and to a have a significant and dangerous impact on major markets as well.” Eventually, the crisis dissipated following an easing of major currency interest rates and approval of the IMF program to Brazil; however, “had the crisis intensified and spread, and had the international community wanted to boost reserves in a large number of large countries, it might have required hundreds of billions of dollars to do so convincingly.” While Lipton concedes that “monetary policy management in the advanced countries is potent and may prove decisive in combating dangerous conditions,” this does not, in his view, obviate the need to “have a large pool of resources on the ready” in case “capital markets judge that international reserves in a wide range of countries are inadequate.”

Box 3.Moral Hazard Consequences of International Crisis Lending

The subject of international moral hazard, and its practical importance as a factor affecting capital flows in the last several years, was a subject of considerable controversy throughout the conference. Mussa and others argue that to the extent that international financial assistance raises capital inflows or lowers spreads by reducing either the true risk of a liquidity crisis or the total economic losses suffered in a crisis—rather than just shifting the distribution of such losses away from investors—it is incorrect to speak of moral hazard, as larger inflows and higher risk-taking on the side of investors might actually be welfare enhancing in such a case. Moreover, to the extent that international rescue operations do create moral hazard, this “cannot be understood as a simple transposition to the international context of moral hazard that typically arises through domestic policies such as implicit or explicit government guarantees.” The latter operate through an “expectation that the costs of excessively risky behavior by borrowers and lenders will ultimately be borne by a third party, namely the average domestic taxpayer. In contrast, the subsidy element in international support packages tends to be small,” as loans and interest to international official lenders are almost always repaid. Consequently, “the direct generation of moral hazard because of expectations that the international community will absorb losses that should be borne by others cannot be a substantial problem* This said, international financial assistance might contribute to moral hazard indirectly by magnifying the shifting of losses inside the country in ways that encourage imprudent risk-taking. For example, the anticipation of international financial assistance might allow domestic agents to borrow more from international investors than they would have done otherwise, raising the ultimate cost of a bailout for the domestic taxpayer.”

Empirically, Mussa and others take the view that excess inflows to Asia had little to do with international moral hazard, since the possibility of a large-scale crisis that might elicit international support was simply not an issue at the time of those inflows. “Russia, in contrast, is the one clear case where moral hazard arising from expectations of international support plausibly did influence private capital flows significantly prior to the crisis.” william Cline cited research suggesting that “post-Mexico, the increase in the markets and decrease in the spreads was driven more by global liquidity as measured by [the price on] high-yield bonds in the United States rather than by moral hazard.” While stressing that moral hazard ought to be taken seriously, Barry Eichengreen also reminded participants that “most investors in Mexico and Asia were not in fact let off the hook,” and observed that the surge in foreign investment to Asia was “lending by banks, not lending through bond markets, and bank creditors were not rescued in Mexico.” In contrast, Charles Calomiris argued that, in addition to the case of Russia, moral hazard had been a relevant factor both in Asia and recently in Latin America. As early as April 1997, The Economist had warned of the dangers of an imminent crisis in Asia. Thus, the continued downward trend in Asian spreads until July could have been driven only by international moral hazard. Moreover, Mexican spreads had been lower than Argentine spreads for the last three years. “Now, when you ask rating agencies and investors why that’s the case, they basically tell you that Mexico is, to overstate the point, the Fannie Mae of Latin America. That is, it enjoys an implicit guarantee of the United States.” Michael Mussa retorted that in his view, the difference between Argentine and Mexican spreads reflected the difference in their exchange rate arrangements and the fact that “everybody was concerned about Brazil and everyone knew that if Brazil got hit, Argentina was going to take a hard knock.”

Lipton saw two possible mechanisms for creating such a pool. One is a general allocation of SDRs, which could be decided by an 85 percent vote of the IMF Board and “thus could in principle be decided quickly in crisis conditions.” However, this would prevent a “concentration of support for reserve accumulation in accordance with the needs that might exist in a crisis,” since SDRs are allocated in accordance with quotas. “An alternative approach would be a two-step procedure combining a large general allocation of SDRs with a pooling of those SDRs by a select group of large countries into a trust fund that could be used for crisis defense.” For example, this group could consist of the General Arrangements to Borrow (GAB) or New Arrangements to Borrow (NAB) participants. They would need to “devise a governance structure that permits activation to defend the financial system and prevents activation to help individual, favored countries.”

Lipton’s proposal was criticized from several angles. At least two speakers—Charles Calomiris and Horst Siebert—questioned the premise that the creation of a large pool of liquidity was necessary and/or desirable given the scope for major central banks to diffuse a global liquidity crisis, and given the potential for such a large pool of resources to be misused. “If there is a really systemic crisis, wouldn’t the role of the lender of last resort be that of the main central banks in a concerted action?” (Siebert). “I would argue that pursuing such a thing in practice would do more harm than good for political economy reasons” (Calomiris). This skepticism was echoed by Kwesi Botchwey: “Given the balance of political forces in the world today and the way that they impact on Fund policy, what guarantee is there that this option will always be exercised in a principled way?” Yung Chul Park was even more explicit, opposing the trust fund idea on the grounds that it would tilt the balance of power in the international financial system even more toward the main sponsors of such a trust fund, namely the G-7 countries, and “provide incentives for other countries in other parts of the world to create their own regional arrangements.” Finally, Jack Boorman took the view that Lipton’s idea merited further study, but at the same time raised a host of practical difficulties, including: how exactly the trust fund would be activated, whether it would require new legislation at the time of crisis, whether the general IMF membership would be willing to see the decision-making authority of the IMF tied up in the hands of a select group of countries at a time of crisis management, and finally, whether the resources of the trust fund would be available immediately in a crisis event or only as a “second line of defense.”

Quality of IMF Advice

What had been intended as a general discussion of the quality of IMF advice, both in the normal exercise of surveillance and as part of conditionality, inevitably concentrated on recent crisis episodes—in particular, the Asian crisis on which the paper by Takatoshi Ito focused. In his presentation Ito focused on the criticism that has been leveled at the IMF programs in Thailand, Indonesia, and Korea and tried to evaluate those programs’ success by providing a before and after type of “event” study. One main aim of the programs was to restore market confidence and to prevent a collapse of these countries’ currency or, at least, to restore exchange rates to a less depreciated level. Looking at the behavior of the exchange rates of these three countries before and after the announcement of an IMF program and rescue package, Ito concluded that these packages, and hence the underlying policies on which they were conditional, had failed to convince the markets that they would work. In his comments, Mohsin Khan (seconded on this point by Mario Draghi) raised doubts about the validity of the type of before/after analysis used by Ito in crisis situations. He pointed out that recent studies that try to define the counterfactual more carefully show that IMF programs have, on the whole, been successful in meeting their aim. One implication of Khan’s comments is that to criticize an existing program effectively one should propose an alternative program and try to evaluate its differential impact analytically and empirically.

Three main dimensions of IMF advice during the Asian crisis were subsequently discussed: tighter fiscal policy, tight monetary policy, and structural measures, in particular with respect to reforming the financial sector.

On the first point, with regard to fiscal policy, it is widely acknowledged that the initial tightening incorporated in the programs was too stringent, but that element of the programs was fairly rapidly relaxed. The IMF advice on monetary policy is much more contentious. Ito stated that the critics had two fundamental objections to the raising of interest rates to prevent further depreciation and its inflationary consequences. The first is that it squeezes the corporate sector and produces an output collapse; the second is that, at times of crises, it will not work to prevent the collapse of the exchange rate. Ito argued that in the case of Thailand, the advice on interest rates would have worked only if the official rescue package had been much larger; the $17.2 billion that were committed did not even suffice to cover the $23.4 billion forward commitment of the central bank. In addition, interest rates were held high for too long a period, an indication again that the policy package and the conditionality were not credible enough to restore confidence in a sufficiently short time.

John Williamson also felt that the advice on high interest rates may have been misguided. While the thesis that high interest rates were needed to prevent a collapse of the exchange rate and the attendant solvency problems the latter would cause in view of existing currency mismatches is coherent, it may not work for reasons touched on by Ito and recently put forward by the World Bank’s Joe Stiglitz. Stiglitz argues that high interest rates will have their usual depressing effects on the economy but that their effect on the exchange rate is ambiguous because, as Williamson put it, “what interests an investor is the product of the coupon rate of return and of the probability of getting that rate of return; while raising the interest rate increases the first of these terms, the second decreases as the probability that the debtor will go into bankruptcy increases.” Williamson thought that the effect of interest rate increases would be the traditional one in situations where what is needed is to increase the rate from, say, 5 percent to, say, 10 or 20 percent. But when the interest rate rises so much that you begin to question the solvency of large parts of the corporate sector, or even of the government itself, Williamson said, the effect on the exchange rate may well be perverse. This is a phenomenon that should not just be dismissed but should be studied empirically.

Other speakers staunchly defended the high-interest rate policy advice in times of crisis. Kiettisak Meecharoen felt that market sentiment about Thailand at the time of the crisis was so negative that strong measures, including high interest rates, were indispensable as was the shift to a floating exchange rate. Mario Draghi gave four reasons why it was crucial to tighten monetary policy drastically. First, avoiding a collapse of the exchange rate was crucial in view of the huge foreign exchange exposure of both firms and banks. Second, many of the negative effects of high interest rates through their impact on the net worth of debtors or the incentive they provide to undertake high-risk projects are felt in the medium term, whereas the effects of a rapidly depreciating exchange rate on the balance sheet and income of banks and firms is felt immediately. Third, the trade links prevailing among Southeast Asian countries make contagion difficult to avoid when one exchange rate collapses. And, fourth, if there is a severe collapse, the positive effect of a depreciation on exports is greatly diminished.

As for interest rates staying high too long, much depends on the determination to correct underlying policy problems and its implementation. Guillermo Ortiz, when comparing the Asian and Mexican experience, attributed the rapid recovery of the Mexican economy and its early return to capital market access partly to the speed with which a program was negotiated with the IMF and the determination with which it was implemented. He and other participants felt that one mistake in the Asian crises was that interest rates were raised too late and in some cases initially too little, prolonging the period during which they had to stay high.

The third area in which the IMF has been much criticized is in its insistence on structural reform as part of programs for crisis countries. All participants who expressed themselves on the subject felt, with some nuances, that bank restructuring was of the utmost importance in crisis countries and that there should be no hesitation in closing insolvent banks while providing a clear safety net for depositors to avoid a banking panic. There was much less agreement on whether it is appropriate to include other aspects of structural reform in IMF conditionality when dealing with crisis situations. Proponents of such structural measures feel that crises provide a window of opportunity for introducing reforms; opponents feel that they blur priorities that should go to macroeconomic policy, the exchange rate regime, and bank restructuring and may give the impression that the crisis will last for a very long time and will not be over unless structural reforms have been completed. This does not help build confidence. In this context, capital account liberalization was one of the contentious structural reforms that was discussed. This issue was discussed above. In the present discussion, the question was whether capital account liberalization should be included in conditionality in crisis situations. Mario Draghi thought that the suggestion should not be dismissed out of hand, although one should proceed cautiously. The one firm advocate of capital account liberalization in crisis situations was Jacob Frenkel, who argued that, coupled with floating exchange rates, capital account liberalization would allow the authorities to aim interest rate policy at the rate of inflation, its primary function. Moreover, if the long-term goal is integration into the world’s goods and asset markets, a constant push for capital account liberalization is the proper strategy lest controls never be removed. This point of view was not generally shared. Stanley Fischer, as he put it in his concluding remarks, found himself “for once” in disagreement with Frenkel; Fischer thought that capital account liberalization should indeed proceed over time to reap the many benefits of trade in assets as well as in goods, but that it should proceed prudently and not necessarily in times of crises.

IMF advice at other times, in so-called normal times and as part of “regular” programs, conditionality, and surveillance, was not discussed during the conference’s last session. Two remarks may be in order here. Regular IMF advice receives higher marks in Ito’s paper than crisis advice, a view shared by John Williamson (and evidence cited by Mohsin Khan does support the conclusion that IMF programs have been, on the whole, successful). Second, if IMF medicine is to be, as most participants wish, preventive rather than curative, such “regular” programs deserve careful attention and could benefit from incorporating some of the lessons learned during the recent crises.


In closing remarks, Stanley Fischer focused on three topics: the rationale and benefits of capital movements; the issue of moral hazard; and prospects for continuing reform of the international financial system.

There is first, the issue of the benefits of capital movements and, more broadly, of asset market integration. In answer to the skepticism voiced, notably by Martin Wolf, as to the desirability or need for encouraging flows of capital to developing and emerging market economies in view of recent crisis experience, Fischer argued that there was a serious and analytically coherent case for freedom of capital account transactions. In the first place, the same analytical apparatus that economists use to justify free trade in goods at a point in time applies to trade in assets over time (intertemporal trade, if you prefer). Second, though net capital flows may be small in relation to either domestic saving or domestic investment, they are not necessarily small relative to net investment and can thus make a significant difference to growth. Third, capital account transactions allow for better risk-sharing for both lenders and borrowers. Fourth, international capitals flows, and direct investment in particular, make for healthy competition for domestic financial institutions and are often accompanied by significant technology transfers. Finally, though this is difficult to model precisely, there seems to be an association between tight controls on capital movements and generally inward-looking, anticompetitive national economic structures.

This is not to deny that there are a number of serious problems associated with capital flows to developing and emerging market economies, including excessive volatility in response to changing market sentiment. Moreover, risk often appears not to be priced correctly, as indicated in the behavior of spreads on emerging market debt, signaling the existence of systemic problems. On balance, however, one observes that governments have not pulled out of international capital transactions in spite of the many shortcomings of international financial markets. This is a significant indication that the net balance of benefits and costs is, on the whole, perceived to be on the side of remaining open rather than retreating into autarky. This is more reason to make the world more stable and to lessen the dangers countries are subjected to when opening up their capital accounts. In this context, Fischer stated that there is, in some circumstances, a case for market-based controls on short-term capital inflows and, certainly, for prudential regulations. Another crucial measure to avoid excessive instability is the adoption of proper and coherent exchange rate regimes. In this respect, increasing capital market integration and large potential capital flows make fixed exchange rates increasingly difficult to maintain unless the peg is very hard indeed. “I think the move to floating exchange rates for countries that decide to integrate into the international capital market, albeit with appropriate prudential controls and possibly controls on short-term inflows until their system is strong enough, will continue, while others like Argentina will peg very hard, and those that are not yet open may continue with crawling pegs or other forms of pegging,” Fischer stated.

Turning to moral hazard, Fischer argued that the one case where it played a major role was the Russian crisis. In contrast, available evidence strongly indicates that moral hazard did not play a major role in capital flows to Asia. This, however, does not mean that moral hazard should be dismissed as an unimportant issue. If there is a safety net—and it is one of the functions of the IMF as set up in the international architecture created in 1945 to “give comfort to members” and to provide such a net—moral hazard will be present in equilibrium and the task is how best to deal with it. Thus, the issue is how to obtain the right trade-off between that hazard and efficiency in a sustainable equilibrium. In this context, Fischer stated, “it is not clear that the way we have handled recent crises is sustainable.” Whatever system will prevail, whether of the Lipton or of another variety, moral hazard is an issue that will have to be dealt with. This will involve better policies on the part of the creditor as well as of the debtor countries, and methods of involving the private sector. In this context, changing bond and other debt contracts is an obvious and important first step, and it is difficult to understand the reluctance of a number of governments in advanced countries, notably within the G-10, to take the initiative.

Finally, Fischer addressed the concern that the impetus toward reform of the system was faltering and that nothing much was and would be achieved. He suggested that “we are making more progress than meets the eye” and listed six areas in which significant improvements were under way. Contingent credit lines, in spite of the criticisms to which they have been subjected, have great potential in the arsenal for crisis prevention. Involvement of the private sector is a second area in which progress is being made even though it is more difficult to achieve than it was in the debt crisis of the 1980s. Third, the Financial Stability Forum has made an encouraging start. Fourth, steps to achieve greater transparency of the IMF are proceeding apace, providing not only information for responsible action by investors but also discipline on the IMF itself. Fifth, the setting up of international standards is one area in which much is being done and where work is proceeding very fast; their implementation represents one of the most important and worthwhile challenges for the international community. Finally, the changes that are taking place in the exchange rate regimes of individual countries are changing the nature of the international system in a positive direction. In the end, however, for all these steps to have the desired effects, better, more stable private sector behavior is essential, and this is why measures aimed at bond and other contracts and, more generally, at better involvement of the private sector in crisis prevention and resolution are crucial. Fischer concluded that it is reasonable to be hopeful that the system would behave in a better, more stable fashion in the future, but that “it is also guaranteed that we will have future crises that come from causes of which we will say later, ‘How did we manage not to foresee them at the time?’”

From this summary accounting of the proceedings, it is apparent that the discussion at the conference raised more questions than it provided answers. That, however, is the mark of all good conferences. Hopefully, the conference raised interesting questions and posed them in a way that points toward the right answers.


    JeanneOlivier2000“Debt Maturity and the Global Financial Architecture,”CEPR Discussion Paper No. 2520 (London: Centre for Economic Policy Research).

    • Search Google Scholar
    • Export Citation
*The authors are grateful to Manmoham S. Kumar for helpful comments and suggestions.
1Lipton’s paper refers to the school that emphasizes fundamental weaknesses as the “IMF camp.” Ironically, the other paper, written by four IMF staff members, actually places somewhat greater emphasis on systemic weaknesses.
2In Zahler’s view, the extent of exchange rate volatility required to make flexible exchange rates an effective deterrent was excessively costly for a country with a large tradables sector. This is related to Calvo’s observation that most emerging market countries display a “fear of floating”; see below. According to Dedek, deliberately increasing exchange rate risk could repel not just desirable but also undesirable components of capital inflows.
3This point is developed in Jeanne (2000).
4This refers to including a debt-rollover option in all foreign-currency-denominated debt instruments (both private and foreign, bonds and loans). At his or her discretion, the borrower would have the option of extending maturing debt for a fixed period (say, three months), presumably at a penalty rate (the precise terms could be negotiated by the borrower and lender themselves). The option could be invoked only once.
5Mainly, to control for other relevant differences across issuers.
6In the same vein, Jack Boorman asked: “Were the Asian crises systemic? They sure had some of those characteristics ex post. Was Brazil systemic? Technically it must have been, because there is a requirement that NAB-GAB resources only be used in such cases and they were used in the context of Brazil.”

    Other Resources Citing This Publication