I. Introduction

Two major views dominate policy discussions of the role of international capital flows in the global political economy. While both believe that high capital mobility is eroding national sovereignty, one sees this as a positive step, that constrains governments’ tendencies to follow overexpansionary macroeconomic policies for domestic political gain and promotes convergence toward low rates of inflation. Advocates of this viewpoint would typically agree with Haggard and Maxfield (1996) that “Increased financial integration holds governments hostage to foreign exchange and capital markets, forcing greater fiscal and monetary discipline than they might otherwise choose” (p. 36). ₁ This view is implied by most of the currently most popular economic models. The second viewpoint sees international capital markets as capricious followers of fads and fashions that pose serious challenges to domestic financial stability. The statements of Dr. Mahathir Mohamad, Prime Minister of Malaysia are typical of this view.

This paper suggests a reality that is quite different. It rejects both the idealized view that financial markets and governments consistently act on the basis of farsighted, well-informed expectations, which is so popular in many economic models, ₂ and the contrasting extreme that investors and speculators are the primary causes of international financial instability through the generation of irrational speculative bandwagons. Neither of these views can account for important aspects of the major international financial crises that hit the European Monetary System in 1992 and 1993, Mexico in 1994, and a number of Asian countries in 1997. Rather than being too skittish, the greater problem with international financial markets is that they have been too gullible: they have given excessive credibility to public pronouncements that pegged exchange rates will be preserved. Thus financial markets have frequently failed to provide early warning signals of the need for increased financial prudence. When crises have hit, it appears that financial markets have often overreacted—at least in terms of the benchmark economic model of well-informed, farsighted rational speculators. But they were belated messengers, not the causes of the crises. This view of the operation of financial markets I call the “Too Much, Too Late” hypothesis.

The latest generation of speculative crisis models suggests that there may be zones of vulnerability inside which a country is not certain to face a crisis, but in which the fundamentals are not strong enough to offer firms protection against medium-sized shocks and swings in market sentiment. ₃ Within these zones countries are vulnerable to self-fulfilling speculative attacks and contagious effects form crises hitting their trade partners. Thus a crisis in Italy or Thailand can be the shock which brings down the British pound or the Indonesian rupiah. In such cases, the speculative attacks are not divorced from the fundamentals, but are not inevitable. In the view presented here, these zone of indeterminancy, which may be fairly broad, interact with the relatively short time horizon of many market participants to reduce the effectiveness of international capital flows as a source of discipline over domestic macroeconomic policies. This view stresses the roles of individual incentives, of limited information—the public choice concept of rational ignorance—and/or short time horizons in both political and financial markets. Moral hazard distortions due to prospects of government bailouts play a role in this process, but are far from the whole story. These political and financial forces interact to provide insufficient international financial discipline—especially where countries have adopted pegged exchange rates—to prevent governments from pursuing policies which substantially increase the risk of generating international financial crisis over the longer term. Mistakes based on incorrect economic views contributed to most of the crises examined here, but it was not mistakes and bad luck alone that generated the series of international financial crisis in the 1990s. More fundamental political economy processes were at work.

Financial markets may not always operate according to the “Too Much, Too Late” hypothesis. And governments may not always succumb to shortsighted political pressures to adopt policies that will lead to crises. But the evidence reviewed in the plausibility probe offered in this paper suggests that the tendencies for such behavior are sufficiently strong that both scholars and policymakers should give the hypothesis serious attention. The hypothesis has important implications for the effectiveness of the highly popular idea that pegging exchange rates is a potent source of anti-inflationary discipline, suggesting that where adjustably pegged rather than fixed exchange rates are adopted, they may as easily reduce as increase discipline in the short run. ₄ This strengthens the view that for many countries efforts to promote financial discipline should focus primarily on domestic rather than international institutional arrangements. ₅

The hypothesis has important implications for the operation of the International Monetary Fund as well. Some have questioned whether there is still an appropriate role for the IMF in a world of high international capital mobility. ₆ To the extent that the “Too Much, Too Late” hypothesis is correct, it suggests that the IMF’s surveillance role should be strengthened in order to reduce the frequency of crises and its capacity for conditional lending once crises have hit be increased in order to minimize the adverse economic consequences when crises do occur. ₇ How to do this without increasing the moral hazard incentives for the generation of future crises is one of the most important issues currently facing the international monetary system.

In the following section, the basic political economy elements of the discipline hypothesis are laid out. In section III, assumptions about the behavior of international financial markets are analyzed and it is argued that the popular extremes of fully efficient farsighted markets and wildly irrational ones are both inconsistent with the evidence. In section IV, one possible intermediate view of the behavior of financial markets is laid out and its implications discussed. In section V, recent crises are examined in light of the “Too Much, Too Late” hypothesis developed in section IV. In section VI, the responses of governments to the crises are reviewed. Section VII concludes by briefly discussing some of the policy issues raised by the analysis.

II. The Discipline Hypothesis

Much has been made of the roles of high international capital mobility and pegged exchange rates in disciplining domestic macroeconomic policymakers and contributing to disinflation in both industrial and developing counties. ₈ While this hypothesis may help explain the convergence of inflation rates of the industrial countries to low levels, how can we explain the series of exchange rate crisis around the world in the 1990s starting in the EMS in 1992-93, then Mexico in 1994-95, and Asia in 1997-98? ₉

According to standard rational choice theory government actors should anticipate the reactions of the private market and thus avoid following courses of action that will promote crises. One version of the pegged-rates-as-discipline argument assumes that exchange rates are genuinely fixed, so that in the long run countries have no option but to adjust. ₁₀ According to this hypothesis, governments will find it in their interest to avoid pursuing policies that would lead to serious balance of payments difficulties, and in the face of exogenous shocks they will begin adjustment promptly. Just as during the days of the Bretton Woods system, some of the support for pegged exchange rates has come from falsely assuming that adjustably pegged rates were genuinely fixed rates.

A more accurate version recognizes that exchange rates can be adjusted, but plausibly assumes that there will be high political costs to such adjustment, giving governments an incentive to avoid policies that will lead to balance of payments crises. Thus, where the private market operates according to rational speculative expectations, not capricious speculative bandwagons, the anticipation of future crises that would be generated by the continued pursuit of policies inconsistent with balance of payments equilibrium at the current exchange rate (or the current rate of crawl in the case of higher inflation countries using a crawling peg) would provide continuous discipline over current macroeconomic policies. More generally, the greater is capital mobility and the more farsighted and informationally efficient are financial markets, the stronger would be these disciplining effects on national monetary policy; and the more farsighted the markets, the more quickly these effects would occur. ₁₁

Crucial elements in this argument are the following:

1. the strength of political biases toward inflation,
2. the magnitude of the costs which governments anticipate would be imposed on them by future crises and devaluation,
3. the magnitude of the costs imposed by following macroeconomic policies to avoid crises,
4. the rate of time discount of current versus future costs and benefits,
5. the degree of uncertainty about these costs and likely future developments, and
6. the behavior of the private sector, especially the degrees of rationality and farsightedness.

Typically changes in macroeconomic policies tend to influence output and employment more rapidly than prices. Thus with expansionary policies most of the benefits come early and the costs later. While contractionary policies suffer from their costs proceeding their benefits, where politicians and the general public have short time horizons, this creates an inflationary bias as discussed in the literature on political business cycles and the time inconsistency of optimal monetary policies. ₁₂ There is some controversy whether pegged or flexible rates will help more to overcome this bias. While Willett and Mullen (1982) argue that by steepening the short-term inflation-unemployment tradeoff, flexible exchange rates would help reduce this bias, more popular in the policy community is the view that since the costs of devaluation are high, pegged rates will be a more effective source of discipline. ₁₃ The argument is that expansion today would increase the probability of devaluation in the future and that the prospect of this expected future cost would be sufficient to more than offset the short-run incentives to expand. Clearly this discipline effect is more likely to work the smaller are the perceived current benefits from expansion, the higher are the perceived future costs of devaluation, and the stronger is the perceived link between current expansion and the probability of future devaluation.

The higher is the degree of international capital mobility the tighter will be this linkage and the more quickly will the expansion generate a sufficiently large balance of payments deficit to provoke a speculative crisis. While the effects of the expansion might take several years to generate a crisis on this score alone, the outflow of capital due to a combination of easy money and expectations of a worsening of the underlying balance of payments would speed the onset and increase the magnitude of a speculative crisis. This would both increase the expected future costs and shift them closer in time.

Ideally, as the likelihood increased that current policy was becoming too expansionary to be consistent with maintaining the fixed exchange rate or limited rate of downward crawl, incipient speculative capital outflows would put upward pressure on domestic interest rates. This rise in domestic interest rates would serve as an early warning system to investors that risk was increasing and to governments that they needed to put their house in order. Optimizing governments would respond to these incentives by adopting corrective policies; thus the financial market’s behavior would provide a healthy discipline over government policies without actual crisis being generated: the threat would be sufficient.

The frequency of crises, however, suggests that this mechanism often does not work. One problem is that even where governments are following prudent policies, a shock to the balance of payments might require that a government would not only have to avoid inflation, but have to follow deflationary policies in order to maintain the exchange rate. The costs of actual deflation would typically be much greater than merely giving up the short-term benefits of expansionary policy (especially where wage and price flexibility, labor mobility, and a country’s marginal propensity to import are low, the short-run costs of deflation could be quite high). ₁₄

It could quite plausibly be the case that the expected future costs of devaluation would be sufficient to induce governments to avoid inflationary policies, but not to adopt inflationary policies in the face of negative balance of payments shocks. Such expectations may well explain part of the process through which high interest rates in Germany following reunification induced a speculative attack on the French franc even though France had been following quite prudent monetary and fiscal policies. While many officials pointed to this case as an example of destabilizing speculation, it can be argued that this speculative attack was quite rational, albeit perhaps self-fulfilling. ₁₅ Even if rational, however, it shows one of the disadvantages of exchange-rate based discipline strategies. Clearly France was disciplined at a time when discipline was not required.

Perhaps more common, however, are problems with the discipline of the financial markets being insufficient rather than excessive. This may occur through two channels. If governments place a very high discount on the future, say because an election is near, then even substantial expected future costs may be insufficient to offset the expected short-term benefits from current expansion. As will be discussed below, such electoral considerations were one of the causes of the 1994 Mexican crisis. Furthermore, the greater is the uncertainty about whether the exchange rate is becoming seriously overvalued, the more likely is a government to more heavily discount possible future costs in relation to current benefits.

The less farsighted is the private market and the more convinced it is by government promises to maintain the peg, the less effective will it be as a source of discipline. On the other hand, even if the government has a very short time horizon, a farsighted private financial market could so speed up the adverse financial effects of an expansion that they would occur as soon as or perhaps even sooner than the favorable output effects.

In between the ideal farsighted market and the capricious herd of irrational speculators who destabilize a perfectly satisfactory situation, there is a conception of markets which behave with considerable aspects of rationality but which are neither terribly well informed nor farsighted. As will be discussed below, such markets may provide a useful disciplinary role once a crisis has started, but they will be insufficient to keep a crisis from developing. Even with such a shortsighted market, the substantial costs of most crises could be sufficient to discipline a farsighted government. Indeed, farsightedness by either the government or the market could be sufficient to produce discipline. Farsightedness by both could be redundant. But shortsightedness by both may be a powerful part of the explanation for the recent crises.

Such market and national government “failures” would provide a case for IMF conditional financing. This could be a double-edged sword, however, since the prospect of IMF bailouts could reduce the incentives for the private market to monitor their loans, and the availability of subsidized loans could reduce the incentives for national governments to avoid crises. Thus some have argued that the Asian crisis was an inevitable consequence of the Mexican bailout.

There is clearly some truth to the concerns expressed about the moral hazard problems generated by IMF bailouts. ₁₆ A strong case can be made that this has reduced the incentives for the major international banks to appropriately evaluate the risks of lending. A similar argument cannot be made for equity investors, however. Such investors in Mexico and Asia took tremendous hits as a result of the crises. Nor does the availability of cheap IMF credit appear to have been a major cause of financial laxity on the part of national governments. The economic and political costs to national governments in the Mexican and Asian crises have been enormous. The value of the subsidy in credit from the IMF has been only a tiny fraction of these costs. Nor do the IMF credits come without tough strings attached. IMF conditionality is of course far from fully efficacious, but it is not at all clear that the positive effect of the conditionality associated with IMF programs is weaker than the negative effects of the below-crisis level market rates of interest charged on these loans. Indeed, in my judgment the reverse is most often the case.

III. The Need for an Intermediate View of Rational but Imperfect Investor Behavior

Economists understandably tend to be suspicious of arguments that markets behave irrationally. Certainly, we would grant that some irrationality exists, but we tend to treat such behavior as an error term in our models, not as a determinant of systematic behavior that can be explained. A typical view of middle-of-the-road economists is given by Kindleberger in his classic study of Manias, Panics, and Crashes: “It is much easier to agree that most markets behave rationally most of the time than that all markets do so all the time.” Furthermore, particular market actors may face incentive structures which make their rational actions socially inefficient. Thus Kindleberger concludes “that despite the general usefulness of the assumption of rationality, markets can on occasions—infrequent occasions, let me emphasize, act in destabilizing ways that are irrational overall, even when each participant in the market is acting rationally” (p. 41).

The implications of the “hard core” assumption of economics and rational choice models are often misunderstood, however. Many academic economists in recent years have tended to write as if the assumption of rationality carries with it the assumption of high degrees of information, long time horizons, and risk neutrality. It is quite understandable that the dictates of mathematical tractability make such ancillary assumptions common in formal modeling, but this seems to have generated a tendency to therefore assume that they are true. This has led economists sometimes to confuse high with perfect capital mobility and to assume that traditional business cycle models must be based on irrationality. Casual observation clearly demonstrates that capital mobility is high, but the empirical evidence against the assumption of perfect capital mobility so often assumed in current international monetary models is overwhelming. ₁₇ As Guttentag and Herring (1987a:169) point out: “Largely because of the inadequacy of information, market discipline is often crude and unfocused. Risk premiums on uninsured deposits seem to be determined largely by the size of the bank rather than by a careful assessment of its financial condition. Tiering of deposit rates in the international interbank market is based less on careful assessments of the creditworthiness of individual banks than on perceptions of their access to government support and the ability of particular governments to provide it” (p. 169). ₁₈ This does not mean that models which assume perfect capital mobility are not quite useful for many purposes, but as we will discuss below, it does imply that for some purposes they can be highly misleading.

It became commonplace for new classical rational expectations economists to dismiss standard political business cycle arguments for relying on assumptions of irrationality. Even one of the major contributors to the traditional literature, William Nordhaus (1989), wrongly conceded this point while defending the empirical applicability of the approach. The traditional political business cycle approach does not require irrationality, however. The public choice concept of rational ignorance can put traditional political business cycle models on a sound theoretical footing. ₁₉ The rational expectations critiques have made valuable contributions in spelling out new versions of rational political business cycle games and stressing the role which learning should play in limiting the effectiveness of repetitive political business cycle games. As a massive amount of empirical research documents, traditional political business cycles games are not played at every election, but they are often played; and contrary to simple rational expectations theory, there is overwhelming evidence that current economic conditions have strong effects on popularity polls and voting behavior. ₂₀

In both the areas of public choice and financial market behavior important lines of research are bridging the gap between the extreme assumptions of idealized rational behavior on the one hand and irrational behavior on the other. Starting with the assumption that information is costly to obtain and focusing on private incentives to be well informed, such research shows that important aspects (albeit not all) of the behavior viewed as irrational in idealized models of rationality can be explained in rational terms when more realistic assumptions are made about the nature of constraints facing decision makers. Drawing on concepts such as rational ignorance from public choice analysis, Herbert Simon’s bounded rationality, Kahneman and Tversky’s emphasis on the framing of decisions, and the incentives facing individuals in organizations emphasized in the new international economics of Oliver Williamson and others, ₂₁ we are beginning to be able to give “reasonable” explanations for many types of behavior that from an idealized rational perspective would appear irrational. Such theoretical developments are helping economists and rational choice modelers to recognize the importance of many facts discovered by empirical researchers that would otherwise tend to be heavily discounted. This combination of theoretical and empirical research has given legitimacy within the academic profession to findings over the past decade or so of the absence of full speculative efficiency in the operation of the foreign exchange markets and “anomalies” in the behavior of domestic financial markets. ₂₂ The so-called behavioral finance has become an important new topic of research.

Public choice analysis easily explains why the knowledge and farsightedness of voters is often quite low. As a recent survey by Robert Schiller (1997) makes clear, even in the advanced industrial countries the average citizen typically has little understanding of the causes and effects of inflation and the operation of the macroeconomy. This is so, not because they are dumb, but because they have little direct incentive to be informed about such issues. The behavior of international investors and speculators is not so well explained on these grounds. These actors should have strong incentives to be well informed and to base their actions on relatively long time horizons. Yet their behavior is often difficult to square with these assumptions. Indeed, the two most popular views of speculators are diametrically opposed. On the one hand there is the view of the well-informed farsighted speculators envisioned by Milton Friedman and rational expectations theorists. In this view exchange rate volatility and speculative runs are the results of efficient speculative responses to current and expected future developments in the fundamentals. Speculators are the messengers, not the causes, of instability.

On the other hand is the view promulgated by Ragnar Nurkse (1947) in his influential study of the interwar international monetary chaos and adopted by many popular commentators (and a few economists) today. This sees financial markets as inherently unstable, dominated by short time horizons, and whipsawed by unfounded bouts of excessive optimum and pessimism. ₂₃ Consequently, speculators are commonly an independent source of instability and provide at best a very erratic source of discipline.

As noted above, the empirical evidence against both of these extreme popular views is overwhelming. To understand the behavior of international financial markets and their roles in providing discipline and provoking crisis, we must adopt a intermediate view; but the exact nature of the intermediate view that best fits the facts is far from clear. This has only recently become a subject of serious study by economists. For example, Froot, Scharfstein and Stein (1990) distinguish between two aspects of informational efficiency—rational pricing and efficient allocation of research. They show that short time horizons may lead speculators to allocate their research in a socially inefficient manner that results in some degree of herding behavior. They conclude that “In equilibrium speculators herd: they acquire ‘too much’ of some types of information and ‘too little’... about fundamentals” (p. 1479). “The herding equilibria also suggest that traders may focus on different variables at different times” (p. 1481). DeLong et al. (1990a, 1990b) consider how the absence of long time horizons on the part of rational speculators increases the amount of market behavior generated by noise and positive feedback or momentum traders. Short time horizons also play a critical role in models of rational speculative bubbles (see, for example, Blanchard and Watson 1982). Valuable models of the speculative behavior of the foreign exchange market in the presence of both noise or technical traders and those who concentrate on the fundamentals have been presented by Frankel and Froot (1987, 1989), DeGrauwe (1996) and Miller (1998). ₂₄

Caplin and Leaky (1994) present a model in which investors’ behavior is dominated by routines which the investors view as costly to adjust. In this model, “Information must change significantly before an individual is willing to bear the cost of adjustment.” Under imperfect information, agents view changes in the behavior of others as providing information about unobserved fundamentals. Consequently, even a market dominated by rational agents may suddenly display “a radical change in market sentiment.” There would be a tendency to label such behavior as irrational on the commonsense grounds that “a significant change in market sentiment ought to have an equally significant cause...[but] the absence of an immediate cause is not a reflection of irrational behavior on the part of market participants. The news that triggers such a crash is analogous to the straw that broke the camel’s back...The market’s reaction to this change in behavior then magnifies and propagates the initial disturbance”: (p. 548). This model also helps explain why markets may not react promptly to early warning signals. “Again such criticism does not imply market irrationality. Rather, the criticism is itself a manifestation of wisdom after the fact. Although much pertinent information is indeed known to individuals prior to the crash, routine behavior prevents agents from putting the pieces of the puzzle together. The crash is precisely the process by which the market aggregates the dispersed information” (pp. 548-49.) There appears to be a surprisingly strong tendency to ignore early warning signals that one’s view may be wrong. As Vertzberger (1990) conveniently documents, studies of decision makers typically find that “the search for and attention to information is biased toward information that is congruent with a priori expectations and predictions, and the interpretation of ambiguous events, toward their being consistent with expectations” (p. 113).

Such behavior forms a basis for Guttentag and Herring’s analysis of “Disaster Myopia in International Banking” (1986). They argue that the inherent human difficulty of evaluating accurately low probability events will be compounded by competitive pressures. “An institution that attempts to charge an appropriate premium to develop a reserve against a low-probability shock is likely to lose business to competitors who are willing to disregard the shock” (p. 3). They also note that the conditions that encourage disaster myopia also reduce the willingness of firms to invest in the information need to convert uncertainty into risk” (p. 5). They go on to argue that “compensation systems for managers that emphasize short-term performance can likewise discourage investment in information regarding low frequency shocks.... Even if shock probabilities are perceived without bias, the personal interests of decision makers may cause them to subject their firms to excessive insolvency exposure” (p. 5). “Since the compensation and power of loan officers are tied to current revenues from loan expansion, they have no incentive to invest in information that might counter disaster myopia. Their personal interest is in ignoring hazards that may not surface for a long time” (p. 12). “High job mobility reinforces this attitude, since the loan officer is likely to be in a different job, perhaps with a different bank, before trouble occurs” (pp. 12-13). Ironically they conclude that “underinvestment in information is likely if decision makers believe they can reduce their exposure quickly and cheaply should shock probabilities suddenly rise” (p. 5), but also suggest that biases operate which may make managers slow to react to early warning signals. “The reluctance of decision makers to react to evidence that shock probabilities have risen once their exposures have become very high is a reflection of “cognitive dissonance’.... Cognitive dissonance is likely to be resolved by ignoring the new information, rejecting it, or accommodating it by changing other beliefs in ways that serve to justify past decisions” (p. 15).

In the recent models of noise trading which can give rise to herd behavior, some economists assume that this reflects the behavior of irrational agents, but believe that we should accept that this is how the world sometimes works (e.g., De Long et al. 1990) while others argue that such behavior “may be viewed as rational agents trading for liquidity or hedging motives, consistent with the rational or efficient-markets perspective” (Dow and Gorton 1997:1025). Dow and Gorton go on to present “another motive for rational, uninformed agents to trade” which “stems from a contracting problem between professional traders and their clients or principals” (p. 1028). The problem is that the client cannot easily tell whether a lack of activity by the agent is due to a reasoned strategy or to lack of attention, thus agents have incentives to use trading as a signal that they are working even when they have no other basis for the trading. As a recent study by the World Bank (1997:126) notes, “investor herding is generally attributed to asymmetric information...It has been suggested that ...professional fund managers [are] particularly susceptible to herding behavior. The essence of this argument is that fund managers will follow the investment decisions of other fund managers in order to show clients that they know what they’re doing. If they follow other fund managers’ decisions and the investment turns out to be unprofitable, they are more likely to be thought of as unlucky than as unskilled, since other fund managers will have made the same mistake.”

Similarly, in large financial organizations, such as the major banks, imperfections in internal incentive structures can lead to aggregate inefficiency. Those who are pushing loans or investments quite naturally do not want to hear analysts bearing tales of caution and many organizations have tended to give much more clout to the former than the latter. Nor do these internal management problems appear to have fully disappeared over time. As Henry Kaufman has noted [(1995:3), quoted in Goldstein and Calvo (1996:253)], “Every firm wants to make deals in emerging markets when margins are high. Analysts become part of the dealmaking and are reluctant to speak up about problems that might derail a deal....The research function is not in the upper echelon of the firm....It is tough to move bad news up the chain of command.” Likewise, David Hale (1996: 134) has argued that “Wall Street further encouraged diversification into emerging-market debt and equity by investing heavily in new research, trading, and investment banking departments targeted on Latin America and East Asia. Despite the obvious deterioration in Mexico’s economic position in 1994, some of these firms also tried to downplay the issue of peso currency risk because of concern that it would jeopardize the transactions flow required to support their expensive overhead at a time when Wall Street’s domestic business was in recession.” ₂₅ One suspects that such problems, as well as a poor sense of history, contributed importantly to the excessive bank lending to developing countries which culminated in the international debt crisis of the 1980s. ₂₆ Furthermore, by the time the crisis hit, many of the managers making the bad decisions had already been promoted up and escaped bearing substantial personal costs. It is not only the IMF and national governments that lend other people’s money. There was clearly some long-term learning behavior and the major commercial banks in the industrial countries were not nearly as heavily exposed to the Mexican and Asian crises as they had been in the 1980s, but this learning mechanism was far from perfect.

From the standpoint of policy, the debate among economists about whether such types of behavior can or cannot be explained on rational grounds is irrelevant. Whatever the explanation, the brute fact is that the financial markets are least sometimes fails to operate in the farsighted efficient manner assumed in many economic models, and as a consequence their ability to serve as effective early warning systems is sometimes seriously compromised. ₂₇

Here I sketch out what I think is one plausible intermediate view and consider its implications for issues of international discipline and crises. I then consider how the combination of this view of capital flows and the view of political behavior discussed above helps explain some of the salient aspects of the recent crises in Asia, the EMS, and Mexico.

IV. The Too Much–Too Late Hypothesis ₂₈

In an earlier paper Clas Wihlborg and I (1991) sketched a view of international capital flows in which, due to uncertainty, financial investors discounted the future heavily and thus typically acted on relatively short time horizons. Many have suggested, although there is little firm evidence, that a short time horizon of financial markets puts pressure on business and portfolio managers in the United States to likewise give considerable weight to short-term earnings. Combined with this may be some degree of hubris on the part of individual investors and portfolio investors that they will be able to detect problems and safely pull their money out before others do. In this view, investors do not generate unwarranted speculative attacks. They are not the independent causes of speculative crises as depicted by many critics of flexible exchange rates and financial markets more generally. ₂₉ But neither do they behave according to farsighted rational expectations. They may not sufficiently anticipate problems, and then may overreact to them when they occur.

This view suggests that investors are not as strong a source of discipline as advocates (including myself in earlier writings, e.g., Willett 1980) have argued, rather many investors focus predominantly on the short-term outlook. As David Andrews and I (1997) have described, a short-term focus leads capital flows to have different short-term discipline effects over monetary and fiscal effects. While interest-sensitive capital flows will still tend to discipline monetary expansion in the short term, they make fiscal deficits unaccompanied by monetary expansion easier to finance—under both flexible and pegged exchange rates. ₃₀ The so-called twin deficits appreciation of the dollar during the first half of the 1980s and the funding of the huge Italian budget deficits within the EMS are major examples of where international capital flows helped finance “excessive” budget deficits for a period of years and thus were an escape from rather than a source of discipline. ₃₁

This time horizon problem can be exacerbated under pegged exchange rates. As long as there is little immediate short-term danger of devaluation, high interest rates that reflect inflation premia may not carry expected depreciation discounts in the minds of international investors, thus generating capital inflows based on questionable perceptions of the differences between nominal and real interest rates. This source of potentially disequilibrating capital flows has been labeled the Walters effect (Walters 1990), and unlike the Nurkse critique of disequilibrating capital flows based on the assumption that speculators had excessive expectations of the likelihood of parity changes, Walters effect flows are based on “excessive” short-run credibility of exchange rate parities.

There is considerable anecdotal evidence of this occurring with respect to U.S. capital flows to Mexico as well as within the EMS. In the U.S. it appears that many young portfolio managers gave excessive credibility to the promises of Mexican government officials that the peso would not be devalued. ₃₂ In one case an economist’s projection that the peso was substantially overvalued and ripe for devaluation was dismissed out of hand with the comment, “Do you realize how much money my fund made last year? Why should I listen to you?” The old adage that “when the finance minister feels the need to pledge that he won’t devalue, you better get the rest of your money out of the country” apparently hadn’t been taught at whatever business school that portfolio manager attended. More understandable was the tendency of frequently inexperienced bank and corporate financial managers in Asia to ignore or substantially underestimate the exchange risk involved in borrowing at low foreign currency denominated interest rates. For example, Park (1998: 16) notes that “Korea’s financial institutions had no expertise in credit analysis, risk management, and due diligence. They also had little experience in foreign exchange and securities trading and international banking in general.” The situation was similar in the other Asian crisis countries. With the substantially higher interest rates in Asia than in industrial country money markets (typically a gap of 300 to 500 basis points) there were strong incentives to borrow abroad if one ignored or heavily discounted exchange risk. Indeed, it is possible that the pegged rates of many of the Asian countries themselves created moral hazard problems. For example, Greenwood (1998:12) argues “the fixed exchange rates created a problem of moral hazard for corporations operating in the area.” The relative importance of beliefs that there was no exchange risk because of government commitments to pegged rates versus assumptions that they would be bailed out if problems arose is unclear, but there is no question that the combination led to substantial overborrowing.

In the Mexican case it seems quite likely that for many investors adaptive rather than rational expectations were at work. The success of the fixed pegs in the EMS during the latter part of the 1980s and early 1990s appears to have had a major impact on the lenses or mental models through which many government officials and market participants viewed the feasibility of maintaining pegged rates. Traditional economists’ views that in a world without capital controls pegged rates would be difficult to maintain without severe loss of domestic monetary sovereignty, the so-called iron triangle or unholy trinity, were forgotten or heavily discounted. (Indeed Richard Portes (1993) recounts statements by officials after the EMS crises chiding economists for not warning that this could happen and counters with examples of numerous warnings by economists that went unheeded.) Ffrench-Davis (1998:30) refers to “the shortsightedness of suppliers [of capital], who saw only the merits of the many achievements made by Mexico...but did not recognize the many problems that were still there...or were being created (external deficit and a growing stock of volatile liabilities).” During 1994 there were several brief attacks on the peso, but at the end of October international reserves still totaled $18 billion and Ffrench-Davis (1998:26) reports that “in November, once the presidential elections were over, international risk analysts gave Mexico good grades financially and recommended investing in its assets.” It would be particularly interesting to discover why so few U.S. firms investing heavily in Mexico failed to learn from the EMS example and why those investing in Asia failed to learn from either.

It may be a psychological corollary that if one has remained optimistic for a long time, and things then turn out badly, one will then overreact. The old saw that financial markets always overreact has been clearly falsified by a massive amount of empirical research. It does not normally hold, but it does seem to often hold in the extreme circumstances of financial crises. Any overshooting that accompanied the depreciations of the pound and the lira when they were forced out of the ERM in 1992 was relatively mild, as was the case for the depreciation of the French franc when the exchange rate bands were substantially widened in 1993. ₃₃ But, as will be discussed in Section V, the depreciation of the Mexican peso in 1995 and that of the Asian currencies forced off their pegs in 1997 went far beyond typical previous estimates of their amount of overvaluation. ₃₄

As a young economist working at the Treasury in the 1970s I remember feeling very dismissive of Undersecretary Ed Yeo’s remark that international financial markets would not be a good source of discipline because they would wait too long and then overreact. Here was one of those typical bankers who didn’t understand economics, I thought. Twenty years of following the behavior of international financial markets has taught me to give much more credence to Yeo’s comment and to similar views expressed by the Bank of Indonesia’s Governor Soedradjad Djiwandono that “the market is very extreme. When they trust us, they trust us like crazy. And when somebody’s confidence is shaken, everyone follows.” ₃₅ Similarly, while granting that the basic cause of the Asian crisis was inconsistent government policies, an editorial by the Financial Times, by no means an anti-market oriented publication, argues that the punishment was far greater than the crime and suggests that “a big part of the explanation has to lie with the fickleness of external investors, who first behaved as if east Asian economies could do nothing wrong and, shortly thereafter, as if they could do nothing right” (“East Asian Shipwreck,” Financial Times, February 16, 1998, p. 15.)

Particularly in the case of U.S. investors in Mexico there appears to have been a strong sense of disillusionment that they had been misled. Badly burnt, and faced with considerable uncertainty, any sign of mismanagement in the handling the crisis, of which there were many, sent the Mexican stock market and the peso plunging further. Such behavior was mirrored in the Asian cases.

Just what mental model was broken by devaluation? That of course varied across investors. All shared in common the assumption or belief that the exchange rate regime would not be changed. For some the “model” may have been as simple as that. Many investors may have had in mind models or reasoning which was little more sophisticated than the adage which contributed so much to the 1982 debt crises—that sovereign governments don’t go broke–a view that was easily falsified by a little knowledge of history, albeit history of more than the previous two decades.

A likely candidate for more “sophisticated” models or views is the Lawsen dogma, named after the former U.K. Chancellor of the Exchequer who advocated it forcefully. This held that as long as the host country wasn’t running a large fiscal deficit and the capital inflows were private rather than public, the resulting balance of payments and exchange rate positions were equilibrium phenomena and thus could not be the source of problems. The first part of this proposition is true in terms of short-run equilibrium, but the second part does not follow logically from the first, since one needs to worry about the sustainability of the capital inflows and the costs of adjustment if they suddenly dry up. It is quite true that a sizable current account deficit can be an equilibrium position for a developing country for some time, but that does not mean that any size current account deficit should not be a source of worry. For example, Dornbusch et al. (1995) estimate that sustainable current account deficits for Mexico in the 1990s would have been on the order of 2 to 4 percent of GDP, while actual deficits ran as high as 8 percent. ₃₆

The basic correctness of the Lawsen dogma was argued quite forcefully by Mexican officials and accepted by many investors. Indeed, one can see how many investment managers could feel that they were being quite sophisticated in understanding that the traditional view that a large current account deficit was always a problem didn’t hold in the “new era” of high capital mobility. As Ffrench-Davis (1998:33) notes, “...in 1996 many outstanding researchers and observers asserted that ...[Asian current account] deficits were not relevant if investment ratios and growth were high.” Unfortunately, as events proved, a little bit of sophistication can sometimes be quite dangerous. While this widely adopted model was insufficiently sophisticated by the standards of modern international monetary economics, its attraction to “practical” investors, who often display great skepticism toward academic economics, is understandable. To understand the growing riskiness of investing in Asia, one needed to look beyond aggregate investment ratios and detect that an increasing proportion of this investment was flowing into relatively unproductive uses (see Barth et al. 1998 and Makin 1998b). Such monitoring obviously requires a much greater investment in data collection and evaluation. Both governments and investors were also slow to pick up on shifts in the composition of capital flows away from direct investment and in the destination of direct investment away from export sectors toward the nontradable sectors. (See Parker 1998). For investors to miss these problems is not wildly irrational despite its clear inconsistency with the standard efficient markets assumption that all publicly available information is employed. ₃₇

Another “model” would focus on the composition of capital flows on the assumption that portfolio investment is likely to be quite volatile but direct investment is not. There is certainly some basis for this view. For example, Frankel and Rose (1996) find a high proportion of foreign direct investment inflows to be associated with a substantially lower probability of crisis. Indeed, the Thai government reportedly undertook a self-study in the wake of the Mexican crisis and concluded that despite the similarity of large current account deficits, their situation was much safer because of the high proportion of direct investment in their capital inflows. Unfortunately, they apparently did not reevaluate their position when later data showed the proportion of direct investment had dropped substantially.

It is not clear why so many banks and investors adopted mental models suggesting that so many exchange rate pegs were credible. To be sure, there were many promising signs in the EMS countries, Mexico, and Asia, but there were also numerous danger signals. Contrary to the economists’ typical assumption that financial markets view government policies with considerable skepticism, in these cases the markets appeared to highlight the positive and ignore the negative. I suspect there was an element of herd behavior in the acceptance of exchange rate credibility views combined with a serious lack of knowledge of history. ₃₈ Given the amount of effort devoted to financial research in aggregate, the adoption of such mental models is quite surprising and should be an important topic for case study research. Given the widespread acceptance of such mental models, however, their falsification would seem likely to induce international investors to run for safety much more than if they had been taking a deliberate, calculated approach to the risks of devaluation and just discovered that they had made the wrong bet. In effect, risk would turn into uncertainty and would generate a strong tendency to pull back substantially on ones’ investments until the new situation was better understood. In such rare cases, financial market overreaction should not be surprising.

The special properties of the foreign exchange market increase the likelihood of exchange rates overshooting in crises. A good deal of the depth of the depreciation of some of the Asian currencies was likely due to a lack of sufficient stabilizing speculation rather than active destabilizing speculation. Many of these countries were running substantial current account deficits, so that substantial capital inflows were necessary to keep the currency from depreciating. Furthermore, because trade elasticities tend to be low in the short run, depreciation is likely to initially worsen the trade balance. This is the well-known J-curve effect. In the absence of considerable stabilizing speculation, the low short-run trade elasticities could cause the initial depreciation to substantially overshoot the long-run equilibrium rate without there being any destabilizing speculation. For many investors with relatively limited information, a sharp decline in a market may act as a signal that they better get out. Where information is limited, the investor may perceive the costs of staying to be greater than those of pulling out. As discussed above, poor or limited information may make herd responses quite reasonable.

When investors have discovered that their view of the world was wrong, such conservative responses seem especially likely. In such conditions of high uncertainty and considerable risk aversion, it could take a substantial expected appreciation to induce investors to provide the capital necessary to balance a current account deficit. Thus in short-run equilibrium, the exchange rate fall might have to substantially overshoot its long-run equilibrium. This amount of overshooting would be greater the larger was the initial current account deficit, the lower are trade elasticities, the greater is the uncertainty perceived by investors, and the greater is their degree of risk aversion. A recent World Bank study (1997) concludes that “there is...some evidence to suggest that emerging markets are prone to foreign investor herding, but...the peak of such herding is relatively short lived” (p. 26). “The experience of countries following the Mexican crisis suggests that pure contagion is a relatively brief phenomenon, international markets were able to differentiate among emerging markets, so that those countries with the strongest economic fundamentals saw a resumption in flows quite quickly” (p. 27).

As Calvo (1996) argues, portfolio diversification can also help account for this phenomenon. Where investors are following a strategy of diversification across many countries and assets, the incentive to seek information is sharply diminished. The combination of the typically high information costs about developing countries and lower incentives to invest in information acquisition may make the markets less forward looking with respect to monitoring developing countries and more reactive to major developments. Calvo and Végh (1998) conclude that “Consequently, in a globalized capital market, investment in emerging markets’ assets is likely to be highly sensitive to rumors and relatively unresponsive to fundamentals” (p. 57).

At some point along the path of depreciation, expectations that the currency needs to be depreciated further will be quenched, but where there is a substantial risk premium due to great uncertainty and an excess supply of currency due to current account deficits and loans maturing the currency may keep falling a good deal beyond the market’s best estimate of its short-run equilibrium value based on the fundamentals. Especially in the case of developing countries, it may require a substantial amount of over depreciation and hence expected appreciation to generate a sufficient incentives for brave speculators to stabilize the currency in the short run. In cases of strong active speculation, official intervention in the foreign exchange market is usually of limited effectiveness. Where the problem is insufficient stabilizing speculation, however, official intervention can be effective. ₃₉

In summary, because of event-produced shifts in mental models, combined with low short-run trade elasticities in the foreign exchange markets and inelasticity of the supply of stabilizing private speculators, the market may overreact in the sense of driving the value of the currency down below its medium term equilibrium value, without there being panic or bandwagon market psychology.

V. A Selective Review of Recent Crises ₄₀

The “Too Much, Too Late” hypothesis doesn’t fit all of the complexities of the EMS, Mexican, and Asian currency crises, but it does seem to explain a good deal. In the vast majority of these cases it is broadly consistent with the gross facts. Markets typically did not give strong early warning signals, and when the crisis did hit, for the Mexican and Asian cases they hit much harder than was expected by many of those who had thought crises were likely. In substantial part the magnitude of these crises resulted from the interactions of the currency crises with domestic financial structures whose weaknesses had not been fully appreciated. But there also appears to have been some element of market overreaction in a number of the cases. The conclusion of Dornbusch, Goldfajn, and Valdés (1995:220) after reviewing earlier crises, that “an overly accommodating capital market aggravates the potential for maintaining the exchange rate and amplifies the ultimate costs of collapse,” appears to apply to the Asian case as well.

Did Markets Overreact?

As shown in Chart 1, for Indonesia, Korea, and Malaysia the behavior of the exchange rate fits well with the hypothesis. The Malaysian ringgit dropped from approximately 2.5 to the dollar in July 1997, to 4.6 in January 1998 before rebounding to 3.6 in March. The Indonesian rupiah fell to over 13,000 to the dollar in late January before stabilizing in the range of 7000 to 9000 to the dollar in February, and the Korean won fell to almost 2000 to the dollar in December before stabilizing in the 1400 to 1600 range several months later. Not all currencies fit the overshooting pattern, however. As previously discussed, any substantial overshooting of the lira was very brief and moderate. Even assuming that none of the exchange rate movements were due to news, the maximum possible overdepreciation was around eight percent for a few days, and three to four percent for about a month. For the pound there was not pattern that could plausibly be interpreted as overshooting. For France there was also at most overshooting by three to four percent for a few weeks. Thus for the ERM currencies we must conclude that the “too much” part of the hypothesis does not hold up strongly.

For Mexico and Asia the “too much” aspect is more difficult to evaluate. In terms of short-term overshooting, the pattern is quite mixed. As noted, the patterns of the ringgit, the rupiah, and the won are consistent with overshooting, but these are dominated by the behavior of exchange rates during January, in which a good bit of the exchange rate fluctuations can be plausibly interpreted in terms of news about the extent of financial problems and the policy intentions of governments. Nor except for Korea are the major overshooting episodes associated with the initial stages of the crisis. For both the Thai baht and the Malaysian ringgit, it is notable that the currencies declined rather steadily from June and July 1997 respectively before the large temporary depreciations in January 1998. The same is true for the Indonesian rupiah during November and December. One suspects that official intervention under flexible rates had a good bit to do with the persistence of these declines during the initial stages of the crisis, but this requires further investigation.

Interestingly, there is also little evidence of initial overshooting in the case of the Mexican peso. As shown in Chart 2, there was a brief spike at the end of January 1995, and a swing from roughly 6 to 7 pesos to the dollar in March that was not fully reversed until April, that are candidates or an overshooting explanation, but neither occurred at the very beginning of the crisis as the market overreaction hypothesis would suggest.

In terms of temporary overshooting, we are left with January 1998 in Asia as the major candidate. Stronger evidence for the “too much” part of the hypothesis comes from the more sustained amounts of depreciation suffered by Mexico and the four Asian countries covered in this study.

Typical estimates of the possible overvaluation of the Mexican peso in 1994 were 20 to 30 percent, ₄₁ but the peso more than halved in value against the dollar. Likewise the Asian currencies fell by far more than standard previous estimates of the amount of overvaluation. Part of large discrepancies between ex ante estimates and the ex post behavior of these exchanges may be due subsequent high inflation in some of these countries, Mexico and Indonesia in particular. Some part can also be explained by subsequent revelations that financial situations were far worse than had been generally recognized. A third element is that typical estimates of overvaluation were based on guestimates of sustainable levels of capital inflows which, while far lower than actual inflows prior to the crises, would still be positive. A complete drying up of net capital inflows or the emergence of net capital outflows would of course require a much larger depreciation to generate balance of payments equilibrium. Estimates of overvaluation are typically based on non-crisis corrections of exchange rates, i.e., they are based on estimates of medium term equilibrium relationships given no major changes in the fundamentals. The Mexican and Asian crises, however, didn’t just change exchange rates, but also perceptions of the fundamentals. Recognizing this, it should not really be surprising that depreciation were much greater than typical estimates of the amounts of overvaluation. From this perspective, excessive depreciation of the currencies is a counter part to an excessive fall of net capital inflows, and as the 1980s debt crisis suggests, this is a condition that could be maintained for a substantial period of time.

Such a view of appropriate normal levels of capital inflows may underlie the judgments of leading IMF officials that the “too much” part of the hypothesis does indeed hold. For example, in an interview in the Financial Times (February 9, 1998, p. 130) Michel Camdessus, Managing Director of the IMF, commented that “In our view currencies have depreciated far more than is warranted.” Or as First Deputy Managing Director Stanley Fisher (1998b:19) put it, “But even if individual market participants behaved rationally, the degree of currency depreciation that has taken place exceeds by a wide margin any reasonable estimate of what might have been required to correct the initial overvaluation of the Thai baht, the Indonesian rupiah, and the Korean won, among other currencies. To put it bluntly, markets overreacted.” Likewise the Deputy Prime Minister of Thailand, Dr. Supachai Panitchpakdi, in remarks to the Western Economics Association commented: “I think most of us agree that the exchange rate corrections have gone too far... The correction has overshot. Everyone in the world knows this, but because of the lack of confidence and because of the easily spread panic, we understand that the present overshoot will not go away easily” (in Saving et al. 1998:139).

While there was undoubtedly elements of panic among some investors, even during the height of the crises the market kept its ability to respond positively to good news as well as negatively to bad news. Thus while the market may well have overreacted on average once the crises began, it still usually responded sensibly to economic and political news. Thus once into the crisis, the foreign exchange and stock markets did appear to provide healthy incentives for the adoption of sound economic policies, typically strengthening on news that reforms were being implemented and falling when indications of backsliding were reported.

Were the Speculative Attacks Unwarranted?

Despite the criticisms of Dr. Mahathir and others, including some economists, that the Asian crisis was largely caused by unwarranted destabilizing speculation, this view does not appear to fit the facts. It is true that none of the target countries were following the highly inflationary policies that have typically been the predominant cause of past currency crises. But they all shared exchange rates which it could be plausibly argued had become overvalued, and all had substantial weaknesses in their domestic financial structures.

The speculative attacks may well have had elements of self-fulfilling speculation in the sense that if confidence had been maintained the crises might have been avoided. In other words, the underlying fundamentals were not necessarily so bad that speculative attacks were inevitable, but all of the countries were in a zone of vulnerability. Dornbusch et al. (1995:256) conclude that for the collapses they reviewed, “none...is associated in any way with an unwarranted attack.” ₄₂ While not all economists would agree, I believe that the same conclusion holds for the Asian case (with the possible exception of Korea, to be discussed below).

In all of the recent crises under review one can point to external developments that helped trigger the worsening in the countries’ balance of payments positions. In Europe it was the Maastricht ratification votes against the backdrop of German reunification which led to large German budget deficits. The Bundesbank’s unwillingness to fuel inflation by financing these deficits by money creation resulted in high interest rates which attracted capital from Germany’s EMS partners, forcing on them the dilemma between tightening monetary policy when their economies were in recession or allowing currency depreciation. Against this vulnerable background, political events triggered speculative attacks. For Mexico a rise in U.S. interest rates was also partially responsible for a slowdown in capital flows to Mexico, as was the assassination of the front-running presidential candidate Luis Donaldo Colosio. In Asia the major external shock was the strong appreciation of the dollar. By continuing to peg to the dollar without major parity adjustments, governments allowed their currencies to become substantially overvalued against other currencies.

Unlike typical international financial crisis, most of the countries involved in the recent crises had low and/or decelerating inflation rates. Thus there were no obvious reasons why these countries should not have been able to adjust their parities in response to external shocks without this leading to a widespread undermining of investor confidence. Indeed, standard rational expectations theory predicts that after a devaluation a country will be a more attractive place to invest. Thus, a devaluation should be followed by stabilizing capital inflows. Frequently this does occur, but not in the cases considered here.

We can fairly easily explain the initial Italian and Mexican devaluations. In both cases the devaluation was considerably less than many estimates of the extent of overvaluation. Thus the market expected that more devaluation was needed and had seen that the government was willing to devalue. Because the devaluations were too small, the consequent further destabilizing of speculative expectations was not surprising, and it would be interesting to know why this wasn’t foreseen by the governments in question. ₄₃ Similarly, the Koreans likely contributed to a worsening of expectations by widening the won’s daily limit of permissible depreciation and then seeing it depreciate quickly to the limit several days in a row. In all three cases the government soon let the currency float, but only after they had further undermined market confidence in their economic policies.

Were these examples of unstable markets subject to irrational swings of optimism and pessimism? We cannot rule out the possibility that nonrational swings in psychological moods may have played some role. But much of the shifts in investor sentiment we can explain in terms of fairly rational agents operating under conditions of poor information and incorrect assumptions or mental models. Part of the strength of investor reactions may have been due to belated recognition of the limited and/or false information on which they had based their investment decisions. Such recognition could easily explain a tendency to pull back on any investments that seemed in any way analogous and may help explain the contagion effects which followed both the Mexican and Asian crises. Of course, as will be discussed below, it is often difficult to determine to what extent contagion effects are justified economically and there is a strong incentive for national officials to blame difficulties on speculators. This is illustrated in the May 1998 Russian crisis where over the course of the month reserves fell by almost $10 billion, a roughly 40 percent decline, and short-term interest rates were raised to 150 percent. Williams (1998:A4) reports that “the country’s chief economic gurus had been blaming Asian financial crises for upending Russia’s fragile markets...But leading reformers said after the Kremlin meeting that it was largely Moscow’s own mismanagement of resources that had pushed the national economy to the edge.”

The Lack of Early Market Discipline

In all of the cases under review governments led the markets to believe that they were firmly committed to keeping their exchange rates fixed or, in the cases of Indonesia, Korea, and Mexico, limited to slow rates of crawl. Economists and policymakers have noted the difficulty of gaining credibility in the market place for government stabilization efforts. Here, however, we appear to have had a rare set of cases of the opposite problem: “excessive credibility” of government exchange rate policies. ₄₄ In most of the cases reviewed here forward markets and interest rate differentials did not give strong signals of speculative concerns about devaluation until shortly before the full crisis hit. For example, Rose and Svennson (1995) conclude that “The currency crisis (of 1992) was not preceded by a gradual deterioration in ERM credibility...despite the increasingly dire economic news from both Germany...and the other ERM participants (including the British recession, the cumulative deterioration in Italian public finances, and competitiveness problems in both Italy and the United Kingdom” (p. 107). ₄₅ Italy devalued on September 14, and both Italy and the U.K. were forced out of the ERM on September 16. Modest market signals of increased expectations of devaluation did not begin to show until late August. In the Mexican case in early 1994 the combination of rising US interest rates, fears about whether NAFTA would pass the US Congress and the assassination of the leading presidential candidate in Mexico generated a substantial reduction in capital inflows to Mexico and stimulated a speculative run on the peso. With the passage of NAFTA and US support of the peso through a swaps arrangement, the markets calmed and the interest differential against dollar assets fell substantially. It did not rise again until further problems in Chiapas generated massive speculative attacks in December. ₄₆ While the pre-December interest differentials were sufficient to indicate a lack of full credibility in the peso exchange rate band, the general impression on Wall Street prior to the December speculative run was that the Mexican situation had stabilized. ₄₇ Krugman (1996:374-5) concludes that “It is hard to avoid the suspicion that financial markets were simply myopic in the runup to both the ERM and the Mexican crisis...Unfortunately, this conclusion wrecks havoc with all of the currently popular models.” While this excessive credibility is somewhat difficult to understand in the European case, it is particularly hard to understand why the lessons of the EMS crisis were not remembered by investors in Mexico and even more so why neither crisis appears to have changed the mental models of many investors in Asia. There was, of course, a considerable amount of initial reevaluation after the Mexican crisis, but the general conclusion appears to have been that Asia was different.

It is certainly understandable that the high investment ratios in Asia and the widespread perceptions of the Asian miracle would lead to this conclusion. What is much more problematic is the apparent assumption that therefore Asia could not also get into trouble. Initial evaluations in 1995 that Asia was still a good investment were likely sound. From the standpoint of rational economic theory the problem is rather that once these reevaluations had been made, they do not appear to have been effectively reconsidered as further information became available. As is discussed below, this may have been due to the psychological tendency toward confirmation bias. This would be an interesting subject for an in-depth field study.

Of course, especially with respect to banks, moral hazard problems certainly played a role. Much of the initiative for capital flows came from borrowing by Asian banks and corporations and was made in foreign currency. ₄₈ Thus much of the industrial country lending was not directly subject to exchange risk. Its risk was from bankruptcy or general financial crisis and was likely subject to both the disaster myopia discussed by Guttentag and Herring (1987) and beliefs in the likelihood of an at least partial bailout in the face of a crisis. On the borrowers side, many were relatively inexperienced in financial management and likely put too much faith both in the durability of fixed exchange rates and in government support if they did get into serious trouble. As Radelet and Sachs (1997:47) comment: “The pegged rate regimes gave misplaced confidence to the [Asian] financiers to accept such [exchange] risk.”

While the learning process with respect to mental models appears to have been suspect, the processing of information through defective mental models may well have been rational. The efficient markets literature is only now coming to grips with the issues of information processing where one is not sure of the true model. The problem with many investors in these cases may be even more difficult to treat formally. They appear to have believed quite strongly in mental models that were false (which of course tends to create much bigger mistakes than when agents recognize there is uncertainty about the correct model). ₄₉

Many international investors appear to have felt personally betrayed when promises by senior officials to maintain exchange rate parities were not kept. Many Wall Street investors in Mexico appeared to have this type of reaction. In such cases one faces not only the shock of having one’s mental model falsified, but also a feeling of betrayal and breaking of trust. Especially in cases such as Mexico where the exchange rate policy had been widely billed as a crucial part of the government’s overall economic strategy, it is easy to see how devaluation would generate a widespread loss of confidence. Much of the depth of the resultant losses of confidence may have been a case of the governments reaping what they had sowed. In both the Mexican and Asian cases the depth of the crisis also reflected the interaction of exchange rate problems with severe domestic financial problems. With respect to both problems it became clear that much of the information available had substantially understated their true magnitudes.

One likely contributor to the slowness of investors to attack currencies that many economists believed were suspect, was the ambiguity of the traditional signals of impending crisis. This was compounded by the well-known difficulties of estimating equilibrium exchange rates. The 1990s crises under review did not involve the standard cases of countries with large budget deficits, rapid monetary expansion and high inflation. ₅₀ (These circumstances do fit the 1990s experiences of a number of the former communist countries.) And in many cases these countries international reserve positions were strong and growing, not declining rapidly until the crisis was virtually upon them.

The signals that were worrisome were substantial real appreciation and large current account deficits. But this is just what we would expect as a consequence of large capital inflows to developing countries with bright prospects for continued economic growth. ₅₁ Thus diagnosis of looming problems required emphasis on the composition of capital inflows, ₅₂ the productivity of investments, and questions of how large was “too large” for the sustainability of current account deficits. ₅₃ Thus these cases were more difficult to evaluate.

It is interesting to note that using a model of optimal consumption smoothing, Ostry (1997) in an IMF Working Paper estimated that for 1995 Malaysia’s and Indonesia’s current account deficits were excessive by only moderate amounts (1 and 1/2 percents of GDP respectively). The Philippines and Thailand did not have excessive deficits on this criterion, but the equations for Thailand did not work well. Overall, this study presents a relatively optimistic outlook for these countries. It notes problems of financial market fragility, but argues that “the ASEAN countries are acutely aware of the importance of appropriate supervision and prudential regulation” (p. 20) and favorably discusses reforms which are under weigh. After discussing a range of factors including high domestic investment ratios, debt service burdens and the composition of capital inflows which have “created concerns of increased foreign exchange risk” (p. 21), the study concludes that “while the evidence in favor of sustainability (both from the model and the other qualitative indicators) is therefore relatively strong, the deficits nonetheless pose risks in the event of unfavorable shocks” (p. 23). In my judgment just such a shock came with the increasing real appreciation over the post 1995 period. ₅₄

Still, the lack of uniformity among the traditional signals of impending crisis gave considerable room for reasonable disagreement between the views of optimists and pessimists. It is interesting to note, however, that among investors—as opposed to economists—the optimists substantially outnumbered the pessimists. With ambiguous signals from the fundamentals, it seems likely that many borrowers and investors fell prey “to the familiar psychological effect of confirmation bias, namely, the tendency to ignore, discount or creatively reinterpret evidence...that goes against some hypothesis or model that we already have in place” (Clark 1997:286). Likewise, group dynamics seem to generate “a powerful drive to rapidly discover a shared interpretation of the data” (Clark 1997:288). Herding behavior seems likely to appear not only in market reactions, but also in the selection of the “models” that underlie such actions. Krugman (1998a) suggests that the tendency for portfolio managers to be evaluated in relation to the performance of their peers likely stimulates herding behavior which contributes not only to the strength of negative contagion effects in response to crisis, but also to “positive contagion effects” which can lead to excessive confidence in currency pegs. Thus he argues that “the apparent myopia about Asian risks seems to have been fed by a general sense of optimism about Asian economies in general” (p. 10).

Moral hazard problems undoubtedly also played a role in the process of overlending and borrowing. It should be noted, however, that these were generated primarily by the policies of governments in the borrowing countries. It is popular to argue that the IMF bail out of Mexico substantially increased moral hazard problems with respect to Asia. There is of course some truth to the argument, but it is easily overstated. ₅₅ Despite the IMF bailout in Mexico, many foreign investors were badly burned. Much more serious were the incentive problems generated by the governments of the borrowing countries, both in directing credit on political and rent seeking grounds and in encouraging Asian borrowers to think that foreign currency loans were virtually risk free.

Despite the importance of the moral hazard problems, we are still left with the puzzle of why other segments of the financial markets, such as international portfolio managers, did not do a better job of evaluating risk and sending early warning signals. ₅₆ In a thoughtful analysis of the moral hazard issue, Fischer (1998b: 27) concludes that while it played a role, “it was not mainly moral hazard that led to the unwise lending that underlies the Asian crisis.” His candidate for the explanation is “irrational exuberance.” Clearly the relative roles of these factors, along with the other aspects of the micro incentive structures facing international lenders and borrowers, is an important area for further study.

In the Asian case, banks appear to have begun to be concerned about the situation somewhat in advance of the crisis. While “bank lending to southeast Asian countries soared between 1994 and 1996, [it] fell off sharply in 1997” (Los Angeles Times, January 23, 1998). However, William Cline of the Institute for International Economics concludes that the risk premia charged by banks declined excessively between 1995 and 1997 (Los Angeles Times, “Banks’ Responsibility in IMF Bail Out Debated,” January 23, 1998, D1). Charles Dallora, Managing Director of the Institute for International Finance, commented “I wouldn’t say the banks were reckless, but risk management wasn’t what it should have been and they clearly were part of the problem” (Los Angeles Times, January 23, 1998, D1.)

In Thailand there were somewhat earlier warning signals. Questions began to be raised by the market in late 1996 as export growth stalled, but initial speculative attacks in November and December 1996 proved premature. The domestic financial and real estate booms generated continuing capital inflows, much of it short term. Rumors of financial difficulties at Finance One, a leading Thai firm, began to circulate in February 1997. The managing director, Pin Chakkaphak, publicly denied the rumors, but privately began selling his own shares. This was symptomatic of serious problems in the Thai financial sector. Unknown to the general public, by June the central bank’s Financial Institution Development Fund had lent over $8 billion to major financial institutions. In late June it was rumored that government support for Finance One would be terminated. On July 2 the baht was floated. Consistent with our emphasis on the role of shifts in mental models, Financial Times journalist Ted Bardacke (1998) wrote that “Investors had been told both that the baht’s value would remain stable, and that leading financial institutions would not be allowed to fail. By abandoning Finance One, Thai authorities altered the way foreign investors assessed Thailand in such a fundamental way that when the baht was floated, it inevitably sank” (p. 7).

One of the key factors to be highlighted by both the Mexican and Asian cases was how much relevant financial information was not available to the markets. Many countries, such as Mexico, published data on levels of international reserves and rates of money growth only with a considerable lag. ₅₇ In the case of Thailand, and to a lesser extent Korea, the published figures on international reserves were highly misleading because of large forward sales that were not reported. Only a few days after taking office in June 1997, the new Thai finance minister, Thanong Bidaya, forced the central bank to admit that while published international reserve figures stood at over $30 billion, secret forward sales of dollars undertaken by the central bank to prop up the baht were almost equally large, leaving a net reserve position of only slightly over $1 billion, scarcely two days worth of imports. ₅₈ In the cases of both Korea and Indonesia, the amount of short-term international debt had been substantially underestimated and in many of the crisis countries the amount of domestic bad debt had also been substantially underestimated. In Indonesia, little attention was focused on its international debt position until after the crisis had begun. Journalist Sander Thoenes (1998) writes, “Poorly reported, the offshore borrowings surfaced only as a topic when the depreciation of the Indonesian currency boosted their value in rupiah terms.” Indonesian private foreign debt had been estimated at $55 billion, “but many brokers feared that this figure could be $20 bn to $50 bn higher, with much of the additional debt in short-term commercial paper” (p. 7). Current efforts by the IMF to improve the quality and timeliness of relevant data may pay substantial dividends in the future. ₅₉

As Ffrench-Davis (1998) correctly points out, however, too much can be made of the concealment or absence of information. As he comments with respect to the Mexican case: “The problem was that neither those on the supply side nor those on the demand side paid enough attention to the available information and that they did not take it seriously until after the crisis exploded” (p. 22). ₆₀ As a recent World Bank study (1997:3) concludes, “although international investors are becoming more discerning, market discipline tends to be much more stringent when investor confidence is lost...than during a buildup to a potential problem.”

Thus even with improved data quality, however, it seems unlikely that we will be able to rely on the international financial markets or credit rating agencies to provide effective early warning systems which are a strong timely source of international discipline. ₆₁ They are likely to remain a blunt instrument whose shadow provides not early signals but costly incentives to avoid the continuation of policies likely to lead to crises. As a warning system it appears that the International Monetary Fund has been much more effective, but its confidential warnings have frequently not been heeded.

Contagion Effects

The amount of information available likely also plays an important role in the amount of herding behavior and contagion effects. A recent World Bank study (1997) suggests “that the potential for foreign investment herding at the country level diminishes as the investors become more familiar with the markets” (p. 130). The likelihood of a U-shaped pattern is suggested in response to the process of financial liberalization in emerging markets. Initial entry into these markets is typically by specialized investment firms with good knowledge and a low propensity to herd. This is followed by regional funds with less specific country and company expertise and hence a greater tendency to herd. Finally in the third stage a class of better informed investors becomes less susceptible to herding. This is consistent with some findings of greater volatility following liberalization, but an overall result that “data on the short-term variability of returns do not suggest that there has been a generalized increase in the overall volatility of returns following liberalization of emerging markets” (p. 134). ₆₂

It should be noted that the contagion effects that were quite indiscriminate, i.e., hitting most developing and former communist countries, were quite brief. ₆₃ Those which were more sustained tended to focus on countries where there were reasonably strong reasons to believe that currencies were overvalued. ₆₄ It is certainly possible that contagion effects could be caused by irrational herd behavior by speculators, similar to the type of behavior that many have argued at times creates bandwagon price runs in individual markets. But it appears that much of the contagion fallout observed in the Mexican and Asian crises can be accounted for by rational behavior in the face of great uncertainty. As a recent IMF study comments

As Paul Krugman points out

In such an atmosphere of uncertainty, unconfirmed rumors are likely to have a much greater effect. A prime example occurred in early December 1997. According to Sander Thoener of The Financial Times, “Indonesia’s battered rupiah lost 10 percent yesterday on rumours that President Suharto had died,...Traders said they suspected colleagues in Singapore of spreading the rumour to push down the currency just as Indonesian companies were struggling to pay off hard currency debts” (Financial Times, December 10, 1997.) Ironically, the following month rumors that Mr. Suharto would seek reelection again sent the rupiah falling.

Furthermore, one of the problems with adjustably pegged exchange rate regimes is that a substantial devaluation by one country can generate overvaluation among some of its trading partners. Some economists have pointed to the Chinese devaluation of 1994 as an important factor in the loss in export competitiveness of Thailand and other Asian countries, ₆₅ and the depreciation of Taiwan’s currency despite its huge foreign exchange reserves has been suggested as a major factor stimulating the speculative attacks on the Korean won.

A great deal of the magnitude of the shifts in investor sentiments can be blamed on the actions of the governments themselves, interacting with what to an economist was a surprising amount of gullibility on the part of both domestic and international financial sectors. In the cases of Korea, Mexico, and Taiwan, election-induced policies also played a role.

There was considerable contagion within Asia, but much of this can be explained by the direct economic linkages among the countries involved and the initial overvaluation of a number of the Asian currencies. ₆₆ Furthermore, as Lipsky (1998) argues, the initial policy response in a number of the Asian countries tended to generate greater rather than less uncertainty, contributing to a climate conducive to pessimistic contagion effects. For countries who are within a zone of vulnerability, adverse shocks elsewhere can be the spark which sets off self-fulfilling speculative attacks. Dean (1998), however, distinguishes between the Indonesian, Malaysian, and Thai cases on one hand, and Korea on the other, arguing that “whereas currency crises and consequent illiquidity in Thailand, Indonesia, and Malaysia were arguably of their own making, South Korea’s currency crisis spread irrationally from the others....” (p. 47). ₆₇ “Unlike in Thailand, Malaysia and Indonesia, the real exchange rate in Korea was not overvalued. Before the won collapsed, Korea’s real effective exchange rate was some 5 percent lower than in the early 1990s, and the real exchange rate against the dollar was 25 percent lower” (p. 58).

Support for Dean’s view is given in a recent econometric study by Chinn (1998) who finds consistent evidence of overvaluation before the crisis for the Malaysian ringgit, Philippine peso, Taiwanese dollar, and Thai baht, and mixed evidence for the Indonesian rupiah. On the other hand, he found that the Korean won was already undervalued. ₆₈

Undoubtedly, some of the contagion to Korea was prompted by tendencies of semi-informed investors to view countries within a region as a group. In addition, Dean points to what he calls the ‘reverse free-rider’ phenomenon among foreign lenders (p. 47). “When portfolio capital is short-term, as much of it is, the incentives for ‘reverse free-rider’ behavior are high if an individual investor sees other investors pulling out, that alone is an incentive for him to pull out, irrespective of the merits of the investment itself” (p. 50). “Asian-Pacific countries in 1997 were vulnerable to banking and financial crises. But contagious exchange rate crises were the trigger” (p. 520. “It was quite rational for individual short-term lenders not to renew, even though collectively it was very much in their interests to do so since that would close the liquidity gap and end the immediate crisis....concerted liquidity relief...was crucial to preventing widespread insolvency” (p. 58).

The Korean case is clearly different from Thailand in the sense that defective exchange rate policy was not at the heart of the matter, but this does not mean that the contagion effect from Thailand to Korea did not have a rational basis. The jolt of the early stages of the Asian crises forced investors to take a hard look not only at exchange rate misalignments but also the health of corporations and financial institutions. Even at an appropriate initial exchange rate, concerns about the latter would generate capital outflows and hence pressures on the currency. It is not surprising that “officials at the Finance and Economy Ministry suggested that panic triggered by ‘sheer speculative foreign press articles’ was driving foreign investors out,” but more likely to be on target are the comments that “international investors are dismissing South Korea’s claims that its financial system is sound...the latest market action highlights a shift in focus by increasingly risk-averse global investors who are turning their attention to South Korea’s financial position, and becoming frightened by what they see” (Roubini, Asia Chronology, Nov. 8, 1997). In mid-November the failure of the Parliament to pass the pending financial reform bills worsened expectations as did initial government announcements that it saw no need to turn to the IMF for assistance. Finally, on December 4, “a record loan package [of $57 billion] led by the International Monetary fund to bail out South Korea helped calm jitters in most regional markets” (Roubini, Asia Chronology, 1997). This temporary calm was soon shattered by uncertainties about the upcoming presidential election and reports that short-term foreign indebtedness was much greater than had been generally thought. Overall, I am doubtful that one has to invoke irrational contagion to explain the Korean crisis. ₆₉

Krugman (1998b) notes the substantial differences in the macroeconomic impacts of the 1990s crises and conjectures that “the key difference was how well informed markets were about the policy environment in the respective sets of countries.” A British devaluation, while it may have shattered the credibility of the current Chancellor of the Exchequer, did not shake confidence in British institutions in general...In Mexico, by contrast, the devaluation made markets question the whole premise that the country was not run by reliable, reformist technocrats” (p. 16) and raised concerns about a political backlash against the reformers. A second important explanatory factor is that the degree of Mexican commitment to its pegged rate strategy was much stronger than the U.K.’s. The crawling peg regime for the peso was widely seen as the centerpiece of Mexico’s economic policy strategy, and the government explicitly fostered this belief. Thus it is not surprising that the breakdown of this policy shattered confidence in Mexican economic policy in general. The U.K.’s peg was widely perceived to be much less of a centerpiece for British policy (although some statements exaggerating its importance can be found). Thus these two cases fit well the Cukierman et al. (1995) hypothesis that the costs of abandoning a peg will be a direct function of the amount of commitment that is perceive to have been made to it.

As a number of economists have emphasized, another major difference between the European and the Asian and Latin American crises is that the latter have been financial as well as exchange rate crises. ₇₀ Dooley (1997) presents a model in which the inconsistency generating the crisis is not between macro policies and the exchange rate, but between national government-based guarantees of financial intermediaries and the backing for such insurance. He also shows that where IMF lending is perceived to be a part of the asset backing for these national insurance programs, then financial crises are likely to be contagious. Miskin (1997) also stresses the important role which weak financial systems can play in the contagion process.

While financial markets may have generally overreacted to the crises due to broken mental models (or panic and herd behavior), the market did not lose the ability to react in a sensible manner to additional news about the economic and financial situation or about government policy initiatives. Nor were these responses strongly asymmetrical, responding only to bad news but not to good news. In many cases the continued depreciation of currencies was due to a string of disclosures revealing that the underlying situation was worse than had been realized; but in general when good news occurred, such as agreements on IMF stabilization packages and national government policy reforms, markets strengthened. ₇₁ Consistent with this view the Wall Street Journal (January 9, 1998) reports “While some blame the irrationality of the markets for Indonesia’s suffering, it is clear that Asia’s currencies are reacting to the varying policy responses in Southeast Asian capitals. The recently plunging Malaysian ringgit, for instance, stabilized this week after Deputy Finance Minster Anwar Ibrahim gave a speech addressing the market’s concerns about Malaysia’s unwillingness to slow economic growth to a more sustainable pace...” (p. A12). Similarly on the same day that the Indonesian rupiah dropped on the news of Suharto’s 1998 budget (which will be discussed below), the Korean won rose on improving expectations about the rollover of bank debt and possible conversion to a new bond issue.

Perhaps a stronger candidate for non-rational contagion effects than the initial devaluation of the Thai baht, were the sharp temporary depreciations of all of the Asian currencies under investigation in January 1998. Policy developments in Korea and Indonesia can easily explain much, if not all, of the depreciations of the won and the rupiah over this period, but the causes of the large depreciations of the ringgit and the baht are not so clear cut. Ironically, Thailand might well have been as much an importer as an exporter of contagion effects. This is clearly an important area for further study.

In summary, while the average level of exchange rates during the crisis periods may have depreciated “too much,” financial markets typically still responded in a reasonable way to economic and political news. In this sense, despite their general failure to provide sufficient discipline to head off the crisis, once the crisis had broken out the financial markets do appear to have generally behaved in a way that produced substantial incentives for governments to adopt stability-producing policies. Thus, belatedly, markets did provide incentives for discipline. It is interesting to observe, however, the substantial differences in how governments responded to these incentives.

VI. Government Responses to Crises

One particularly vivid lesson of the crisis of the 1990s is that the nation state is not dead. The forces of globalization are important, but they are not always dominant. Financial markets can impose substantial economic and political costs on governments, but governments still have considerable freedom of action in how to respond. This is illustrated by the wide range of responses by developing country governments to the financial crisis of the 1990s. While none of the governments involved showed much foresight in anticipating the adverse consequences of their policies which contributed to the crises, under the force of its presence some governments, such as Mexico, did rather rapid about faces and gave high priority to pursuing the painful policies which would calm markets and lay the groundwork for restoring a prosperous economy. Thailand appears to be an intermediate case. The Thai government has not indulged in the anti-speculator rhetoric of Mr. Suharto and Dr. Mahathir, and U. S. Secretary of the Treasury Rubin publicly commented in January 1998 that the Thais “have expressed real commitment to working with the IMF and World Bank.” However, the Financial Times reported that privately some foreign economic officials were concerned that “There is still lack of appreciation in Bangkok about just how serious this is” (Redding and Kynge, January 13, 1998, p. 8).

Still, compared with the Indonesian and Malaysian cases to be discussed below, the Thai government showed considerable responsiveness to financial markets and the IMF. ₇₂ It may not be coincidental that in both the Mexican and Thai cases new governments were in power. (In the Mexican case the crisis started shortly after a new President had taken office. In the Thai case, the crisis contributed to the resignation of the government of Prime Minister Chavalit Yongchaiyudn.) This pattern was repeated to a considerable extent in Korea. There the first signs of impending problems were the bankruptcies of Hanbo Steel in January and Sammi Steel in March 1997. As in almost all of the Asian cases, investors failed to recognize the full extent of the problems, and in several cases this was due in part to deliberate government efforts to disguise the magnitude of the problem. The crisis hit Korea during the run-up to the presidential election, and it seems clear that being in this sensitive stage contributed to the slowness of the initial government responses and the difficulty in making acceptance of an IMF rescue program a bipartisan issue.

Electoral considerations were surely at least part of the motivation behind President Kim’s delays in asking the IMF for help and in freeing the won. Likewise they must have contributed to the leading opposition party’s presidential candidate Kim Dae-Jung’s hesitancy in promising to endorse the IMF rescue package if he were elected. These actions, along with the delayed announcement that Korea’s available international reserve position was much lower and its short-term international debt obligations much higher than had been reported, contributed to a worsening of the crisis. Nor did the close election of Mr. Kim, who was viewed as being on the left and as having little expertise on economic matters, help calm the markets. Some initial clumsy statements contributed to such concerns, and on several days both the stock market and the won recorded dramatic drops. Quite quickly, however President-elect Kim toughened his stand on the need for reform and by the second week of January foreign funds began to return to the country.

As Milner and Keohane (1996:20) stress, “Political Leaders have a degree of latitude in how they respond to internationalization. In large part, this range of choice is a function of the domestic institutional framework in which they must operate.” ₇₃ The long-time strongmen governing Indonesian and Malaysia were unwilling to emulate the seriousness of the policy response of their neighbors in Korea and Thailand. Dr. Mahathir Mohamad, Prime Minister of Malaysia, shocked international financial markets by blaming the crisis on Westerners and Jews who were jealous of Malaysia’s economic success and were trying to topple the economy. Investor George Soros, famous for contributing to the toppling of the pound sterling in 1992, was singled out as a “moron” and the “arch villain in conspiring to impoverish south east Asian countries by attacking their currencies.” Foreigners selling Malaysian shares were described as racist rumor mongers “who should be shot.” ₇₄ Mahathir even proposed a ban on currency trading which he described as being “unnecessary, unproductive, and immoral.” Financial Times journalists John Redding and James Kynge (1998) note that while “the rhetoric helped shore up Dr. Mahathir’s already formidable domestic power base...almost every time he attacked his perceived enemies, the ringgit and stock prices fell” (p. 38). ₇₅

While Indonesian President Suharto’s initial public statements were not as challenging to international financial markets as Dr. Mahathir’s, his actions, if anything, became even more so. The initial government response in Indonesia was positive. The rupiah was floated on August 14, and on September 3 an ambitious reform package was announced. This included a freeze on a number of major infrastructure projects and a promise of banking reform which included shutting down insolvent banks. The Governor of the Bank of Indonesia described the government response as a “self-imposed IMF programme,” and both the rupiah and the stock market responded with a substantial strengthening. However, while the government and many domestic investors thought that the situation had stabilized, Stephen Rodgers of UBS Securities in Jakarta commented that “The perception among foreign investors was that this package was a reasonable first step but they had a hell of a long way to go. Still a lot of people at that stage were giving the government the benefit of the doubt” (Thoenes 1998). As private estimates of the amount of foreign debt mounted, however, this benefit of the doubt began to rapidly erode. Fears that such debt might well be double the official figure severely undermined confidence. Paul Shang of Lehman Brothers commented that “Investors are looking for a clear picture in terms of how deep and serious the asset problems are and how they [the government] are going to deal with them. The government never gave this picture... [and] domestic and foreign investors alike rushed for the exits” (Thoenes 1998). On October 8 the government turned to the IMF for advice, but not money. As the rupiah continued to depreciate, however, the government was forced to apply for a formal IMF stabilization program which was announced October 31. The temporary stabilization of the rupiah was brought to a screeching halt, however, by the announcement of the government’s new budget. Markets were stunned when in early January Mr. Suharto’s new “austerity budget,” revealed only a few months after the negotiation of a $43 billion IMF rescue package, called for a 24 percent increase in government spending. That day (January 8) the rupiah fell by 26 percent against the dollar and the stock market tumbled.

Nor was the resulting further loss of confidence limited to the financial markets. The Wall Street Journal (January 9, 1998) reported widespread panic buying by the public: “...many workers in the capital left early in the afternoon—following panicky phone calls from family and friends—to try to stock up on essential goods. Supermarkets and shops were jammed by customers fighting to buy rice, noodles, cooking oil, sugar and powered milk.”

Apparently oblivious to how his own actions were undercutting the chances of the IMF program to bolster confidence, Mr. Suharto appeared to accept critics’ views that the IMF approach was not working. Prodded by members of his family, and against the advice of most of his own economic advisors, the international financial community, the IMF, the World Bank and the U.S. Treasury, Suharto decided to gamble on the adoption of a currency board that would fix the rupiah to the dollar at a rate rumored to be between Rp 5,000 to Rp 6,000. ₇₆ Sander Thoenes reports that “One Indonesian government official said most members of the cabinet and presidential advisors opposed a hasty introduction of a currency board” (Thoenes, February 12, 1998, p.1). Despite strong international opposition, including a public statement by IMF Managing Director Michel Camdessus, speaking with the unanimous backing of the IMF Executive Board, that “the failure of a currency board would completely undermine credibility and policy making and severely damage the country’s growth prospects” (Financial Times, February 14-15, 1998), Finance Minister Mar’ie Muhammad announced to Parliament that the government “was preparing steps toward setting up a currency board system, including legislation to support the board” (Wall Street Journal, February 12, 1998, p. A14).

Anticipation that the rate would be fixed at a high value led to a substantial appreciation of the rupiah, but because there was little evidence that the board would be backed by the strong political commitments which had helped currency boards in Argentina, Estonia, and Lithuania to gain considerable credibility, Mr. Suharto’s actions did not have confidence-building effects. The Wall Street Journal reported that “Even if Indonesia does adopt a currency board, speculators are eager to test the government’s resolve in holding the peg.” While Indonesia’s international reserves were quite adequate for initially backing rupiah currency in circulation, they fell far short of total bank deposits, which the government had recently guaranteed. Some economists such as Stephen Hanke, who Suharto’s family brought in to advise the government, argued that a currency board would quickly establish market confidence.

On this view one could interpret Mr. Suharto’s actions as bowing to the discipline of the market. More likely, however, Mr. Suharto was hoping that the currency board would provide a quick fix that would allow the economy to escape painful adjustment. Few international economists shared this optimistic view. Many pointed out that the conditions which had made currency boards successful in countries such as Hong Kong, Estonia, and Lithuania were not present in Indonesia. With Mr. Suharto’s intentions on reform still suspect, the likely result would be high interest rates and deep recession. Thoenes reports, “Top Indonesian officials, the IMF, World Bank and foreign lenders have all warned in recent weeks that a currency board system would force up interest rates and damage the economy” (Financial Times, February 12, 1998, p. 1). ₇₇

The currency board-induced strengthening of the rupiah lasted only a few days. By letter on Friday, February 13, and public statement on Monday, February 16, Mr. Camdessus made clear that the IMF would halt disbursement from the $43 billion stabilization package if Indonesia proceeded with plans to establish a currency board. Backed by the finance ministers of the major industrial countries, Mr. Camdessus commented, “When a person is very ill you shouldn’t kill him with your medicine, and this could kill” (Wall Street Journal, February 17, 1998, p. A14). President Clinton called Mr. Suharto to back the IMF’s position. The heads of state of Germany, Japan, and the United Kingdom also communicated support for the IMF. The market reacted to this news with a 22 percent drop in the value of the rupiah against the dollar, before recovering slightly to close on Monday the 16th down 18 percent from Friday’s close.

Why did Suharto pay so little attention to the pressures of the financial markets and the IMF? A Wall Street Journal editorial suggests two possibilities.

One suspects that both motives may have been at work. Five of Suharto’s children have business empires and several have continued to receive favorable treatment during the crisis. One son publicly criticized the closure of his bank and soon after was allowed to purchase the license of an inactive bank and reopen in his old bank building under a new name. Other examples include the continuation of an electric power plant on Java where there is a surplus of electricity and the award of large tax breaks and pressuring of banks to provide financing for an automobile project of his youngest son (Wall Street Journal, January 9, 1998, p. A12).

Two explanations have surfaced for Mr. Suharto’s family’s support for a currency board. The Wall Street Journal reported the views of U.S. economist David Cole that the primary purpose was retribution against the central bank which had “clashed frequently with banks controlled by Suharto family members. According to Cole “the whole purpose of it is not to stabilize the currency; its to try to destroy the central bank” (Wall Street Journal, February 10, 1998, p. A13). A complementary view, reported by Sander Thoenes of the Financial Times, is that “The currency board is supported by Mr. Suharto’s family and by his wealthy business associates...A high rupiah would help them pay off their offshore debt” (Financial Times, February 12, 1998, p.1). Either explanation suggests that there may be substantial economic costs as well as benefits to “strong” governments. Further evidence of Mr. Suharto’s lack of concern with market discipline was offered a few days later when he chose for his vice-presidential candidate, Mr. Habibie, the Minister for Science and Technology. According to the Financial Times, “Mr. Habibie is known for his high-cost prestige projects and hostility toward trade deregulation” and was “widely opposed by foreign and domestic investors” (Financial Times, February 13, 1998). Furthermore, General Wiranto, new chief of the armed forces, “is believed to have told Mr. Suharto that most military officers opposed Mr. Habibie’s nomination” (Financial Times, February 13, 1998). Not surprisingly, the Jakarta stock market fell by 9.8 percent on the day of the announcement.

Eventually, Mr. Suharto backed off of the currency board proposal and his government began to negotiate seriously with the IMF again, but only after considerable damage had been done. It may not be coincidental that the switch to a more conciliatory attitude by Mr. Suharto did not begin until after he had been safely “re-elected” to another term as president. In a clash between financial market pressures and domestic political pressures, it would not seem wise to always bet on the dominance of market pressure, especially around election time.

Reelection, however, was not sufficient to keep Mr. Suharto in power. Ultimately, continuing economic distress and widespread domestic opposition to cronyism and support for more democratic government led to sufficient pressures to induce Mr. Suharto to step down. Unfortunately, uncertainties about his successor, Mr. Habibie’s degree of independence and commitment to political and economic reforms, have kept the rupiah from making a quick recovery from its spring plunge in response to the demonstrations and rioting which led up to Mr. Suharto’s resignation.

VII. Concluding Remarks

The exploratory analysis presented in this paper suggests that the “Too Much, Too Late” hypothesis, while clearly needing refinements and certainly not providing a full explanation for the recent Mexican and Asian crises, appears to fit the facts much better than the major alternative hypothesis about the behavior of international financial markets. It also suggests that the nation-state, for better or worse, is still far from dead.

The general approach suggested here offers a rich menu for political economy and financial research. The best ways to model these phenomena are still far from clear, but it does seem virtually certain that such efforts should begin with the assumptions that we likely cannot rely on international capital flows to be either a reliable early warning system or a source of continuous discipline over governments’ macroeconomic policies. Like voters, they will ultimately sanction politicians for bad economic policies, but only after things have gotten out of hand.

Nor are pegged (as opposed to fixed) exchange rates likely to be a consistent source of discipline. ₇₈ There are cases where a commitment to a pegged rate has been an important source of domestic discipline for a sustained period of time. Austria, France, the Netherlands from the 1980s to the present, and Mexico from 1989 to 1993 are all examples. But Mexico, the Czech Republic, Thailand, the Philippines, Malaysia, Korea and Taiwan in the mid 1990s are all examples where this discipline did not work. In many of these cases at least some of the effects were perverse. ₇₉ In the absence of well-informed forward-looking behavior on the part of voters or the financial markets, pegged exchange rates can increase the short-run incentives to destabilize the economy. ₈₀ By delaying devaluation for too long, the eventual crises are made much worse.

The proper lesson is that necessary devaluations should not be put off. Unfortunately, many commentators and officials draw just the opposite conclusion. Because the effects of overly delayed devaluations are often catastrophic, it is argued that they must be fought against even harder. This is a typical theme of Wall Street Journal editorials. Discussing the May 1998 Russian crisis, former Prime Minister Yegor Gaidar argued that “even a slight downward adjustment of the ruble’s value would deal a death blow to stabilization...Only half-wits think that a managed devaluation of the ruble is possible here. In reality...that would guarantee an unmanageable plunge of the ruble, depleting Central Bank reserves, dealing a heavy blow to the banking system and causing total panic” (Williams 1998:A4). Thus does a misreading of the lessons of currency crises contribute to the conditions for their repetition.

Our analysis suggests that short of adopting “permanently” fixed exchange rates or greatly expanding the power and authority of the IMF, international sources of discipline will often be weak until crises develop. For countries where adopting a genuinely fixed exchange rate does not make sense on optimum currency areas grounds, this suggests that the best ways to avoid international financial crisis are to adopt considerable flexibility in their exchange rate policies and to focus on domestic reforms to counter the incentives to pursue economic policies which provide short-term political gains at the risk of producing longer term economic instability. ₈₁

This analysis also suggests the importance of carefully analyzing the role of the IMF as an agent of discipline and of crisis avoidance and containment. ₈₂ This topic goes far beyond the scope of this paper, but a few brief observations may be useful. These recent episodes suggest that on average the IMF has been a better forecaster or early warning signal of emerging policy inconsistencies than have either national governments or private financial markets. But the IMF has not been terribly successful in persuading national governments to act in a timely fashion. Perhaps the pain generated by the current crisis will serve to increase the attention that national governments will “voluntarily” pay to IMF advice. But sadly, the Asian crisis episode yields little evidence of learning by either national governments or private investors from the Mexican crisis. The crises of the 1990s have provided a healthy spur to the more timely and accurate provision of financial data. This will undoubtedly help, but is far from a complete solution.

Some have suggested that the IMF has at its disposal a powerful instrument to increase its clout. There is little doubt that making public its warnings would unleash the wrath of market forces against recalcitrant governments, but there would also be many costs to such a strategy. Spurred particularly by the governments of France and Germany, at its April 1998 meetings the Board of Directors of the IMF agreed that a system of more graduated warnings should be developed, and that the possibility of the use of public warnings in extreme cases should not be ruled out. The IMF itself has understandably displayed considerable reluctance to embrace the threat of formal public warnings. Knowing that the information they provided might be publicly disclosed national officials would likely be much less frank in their discussions with the IMF and a more adversarial stance would almost inevitably develop. Furthermore, this approach would leave the IMF open to charges of having generated rather than calmed international financial crises. Whether these costs would be worth the benefits of increased IMF clout is not immediately obvious. Both the possible desirability and political feasibility of such a change in the role of the IMF deserve careful attention. ₈₃

In summary, as Goldstein and Calvo (1996: 258) argue, “both private market discipline and official surveillance are prone to weaknesses that can reduce their effectiveness; therefore relying exclusively on one or the other would be ill advised.” To promote discipline most countries will need to look primarily to domestic institutional reforms. As for capital flows, while they cannot be consistently relied upon to provide sufficient discipline, neither are they major culprits, destabilizing otherwise stable situations. They do need to be monitored carefully, however, and while a return to widespread capital controls would be neither feasible or desirable, we have many examples of how complete freedom of capital flows can interact quite dangerously with distorted incentive structures in weak domestic financial systems. The management of domestic and international financial liberalization in the emerging market economies is one of the greatest challenges facing the global economy today. ₈₄

References

Aitken, Brian. 1998. “Have Institutional Investors Destabilized Emerging Markets?” Contemporary Economic Policy, April: 173-184.

Alesina, Alberto, and Nouriel Roubini with Daniel Cohen 1997. Political Cycles and the Macroeconomy. Cambridge, Mass.: MIT Press.

Andrews, David. 1994. “Capital Mobility and State Autonomy: Toward a Structural Theory of International Monetary Relations.” International Studies Quarterly 38: 193-218.

Andrews, David, and Thomas D. Willett. 1997. “Financial Interdependence and the State.” International Organization 52(3): 470-511.

Artus, P., and H. Bourguinat. 1994. “The Stability of the EMS,” in A Steinherr (ed.), Thirty Years of European Monetary Integration. London: Longman.

Auerbach, Nancy Nieman. 1997. “The Mexican Peso Crisis: Constituent Pressure and Exchange Rate Policy.” Claremont Policy Briefs No. 97-02, December.

Baillie, Richard T, and Patrick McMahon. 1989. The Foreign Exchange Market: Theory and Evidence. Cambridge: Cambridge University Press.

Baliño, Tomás, et al. 1997. “Currency Board Arrangements.” IMF Occasional Paper No. 151. Washington, D.C.: International Monetary Fund.

Bardacke, Ted. 1998. The Financial Times, various issues.

Barro, Robert, and David B. Gordon. 1983. “Rules, Discretion and Reputation in a Model of Monetary Policy.” Journal of Monetary Economics 12: 101-21.

Barth, James, R. Dan Drumbaugh, Lalita Ramesh, and Glenn Yago. 1998. “The Role of Governments and Markets in International Banking Crises: The Case of East Asia.” Presented at the Western Economics Association Meetings, Lake Tahoe, June 28-July 2.

Baughton, James. 1997. “From Suez to Tequilla: The IMF as Crisis Manager.” IMF Working Paper WP/97/90. Washington, D.C.: International Monetary Fund.

Bekaert, Geert, and Campbell R. Harvey. 1997. “Emerging Equity Market Volatility.” Journal of Financial Economics, January: 29-77.

Bekaert, Geert, and Campbell R. Harvey. 1995. “Emerging Equity Market Volatility.” NBER Working Paper No. 5307. Cambridge, Mass.: National Bureau for Economic Research.

Bhagwati, Jagdish. 1998. “The Capital Myth.” Foreign Affairs, May/June: 7-12.

Blaine, Michael. 1998. “Déjà Vu All Over Again: Explaining Mexico’s 1994 Financial Crisis.” The World Economy, January: 31-55.

Blanchard, Olivier J., and Mark W. Watson. 1982. “Bubbles, Rational Expectations, and Financial Markets,” in Paul Wachtel, ed., Crises in the Economic and Financial Structure. Lexington, Mass.: Lexington Books.

Bordo, Michael, and Anna Schwartz. 1996. “Why Clashes Between Internal and External Stability Goals End in Currency Crises.” Open Economies Review 7: 437-68.

Bowe, Michael, and James W. Dean. 1997. Has the Market Solved the Sovereign-Debt Crisis? Princeton Studies in International Finance No. 83, August.

Brau, Eduard. 1995. “External Financial Assistance,” in Daniel Citrin and Ashok Lahiri, eds., Policy Experiences and Issues in the Baltics, Russia, and Other Countries of the Former Soviet Union. IMF Occasional Paper No. 133. Pp. 103-16. Washington, D.C.: International Monetary Fund.

Braun, R. Anton, Arijit Mukherji, and David Runkle. 1996. “Delayed Financial Disclosure: Mexico’s Recent Experience.” Federal Reserve Bank of Minneapolis Quarterly Review, Fall: 13-21.

Brown, Stephen, William Goetzmann and James Park. 1998. “Hedge Funds and the Asian Currency Crisis of 1997.” NBER Working Paper No. 6427. Cambridge, Mass.: National Bureau for Economic Research, April.

Bruno, Michael. 1995. “Comment.” Brookings Papers on Economic Activity 2: 278-85.

———. 1991. “High Inflation and the Nominal Anchors of an Open Economy.” Essays in International Finance No. 183. Princeton: Princeton University Press.

Buiter, Willem, Giancarlo Corselti, and Paolo A. Pesenti. 1998. Financial Markets and European Monetary Cooperation. Cambridge: Cambridge University Press.

Burdekin, Richard D.K., Heidi Nelson, and Thomas D. Willett. 1998. “Central European Exchange Rate Policy and Inflation,” in Sweeney et al.

Calvo, Guillermo A. 1996. “Capital Flows and Macroeconomic Management: Tequila Lessons.” International Journal of Finance and Economics 1: 207-23.

Calvo, Guillermo A., and Enrique G. Mendoza. 1996. “Mexico’s Balance-of-Payments Crisis: A Chronicle of a Death Foretold.” Journal of International Economics 41(3-4) Supplement: 235–64.

Campbell, John Y., Sanford J. Grossman, Jiang Wang. 1993. “Trading Volume and Serial Correlation in Stock Returns.” Quarterly Journal of Economics, November: 905-39.

Campbell, John Y., and Albert S. Kyle. 1993. “Smart Money, Noise Trading, and Stock Price Behavior.” Review of Economic Studies 60: 1-34.

Caplin, Andrew, and John Leaky. 1994. “Business as Usual, Market Crashes, and Wisdom after the Fall.” American Economic Review, June: 548-65.

Caves, Richard, Jeffrey Frankel, and Ron Jones. 1996. World Trade and Payments. New York: Harper Collins.

Chann, Huh, and Kenneth Kasa. 1998. “Economic Competition and Contagious Currency Crisis.” Federal Reserve Bank of San Francisco Economic Letters No. 98-1, January 16.

Chinn, Mengie. 1998. “Before the Fall: Were East Asian Currencies Overvalued?” NBER Working Paper No. 6491. Cambridge, Mass.: National Bureau of Economic Research.

Chuhan, Punam, Gabriel Perez-Quiras and Helen Popper. 1997. “The Capital Flow Mix.” Santa Clara University Working Paper.

Clark, William, and Usha Reichert. 1998. “International and Domestic Constraints on Political Business Cycles in OECD Economies.” International Organization, Winter: 87-120.

Cline, William. 1995. International Debt Reexamined. Washington: Institute for International Economics.

Cohen, Benjamin J. 1996. “Phoenix Risen: The Resurrection of Global Finance.” World Politics 48: 268-96.

Cukierman, Alex, and M. Kiguel and Nissan Liviatan. 1995. “How Much to Commit to an Exchange Rate Rule: Balancing Credibility and Flexibility,” in Pierre Siklos, ed., Varieties of Monetary Reform. Boston: Kluwer.

Daniel, Betty C. “Fiscal Policy and the Predictability of Exchange Rate Collapse.” IMF Working Paper WP/97/133. Washington, D.C.: International Monetary Fund, October.

Dean, James. 1998. “Why Left-Wing Moralists and Right-Wing Academics Are Wrong about Asia.” Challenge 41(2): 44-60.

DeGrauwe, Paul. 1996. “Exchange Rates in Search of Fundamental Variables,” in Marc Uzon, ed., The Financial System Under Stress. Pp. 196-213. London: Routledge.

———.1992. The Economics of Monetary Integration. Oxford: Oxford University Press.

DeLong, J.B., Andrei Shleifer, Lawrence Summers and Robert Waldman 1990a. “The Economic Consequences of Noise Traders.” Journal of Political Economy 98: 703-38.

———. 1990b. “Positive Feedback Investment Strategies and Destabilizing Rational Speculation.” Journal of Finance 45: 379-95.

Dixit, Avinash. 1996. The Making of Economic Policy. Cambridge, Mass.: MIT Press.

Dominguez, Kathryn, and Jeffrey Frankel. 1993. Does Exchange Market Intervention Work? Washington, D.C.: Institute for International Economics.

Dooley, Michael. 1997. “A Model of Crises in Emerging Markets.” NBER Working Paper No. 6300. Cambridge, Mass.: National Bureau for Economic Research.

———. 1996. “Comment on Private Capital Flows to Emerging Markets after the Mexican Crisis,” in Guillermo Calvo, Morris Goldstein, and Eduard Hochreiter (eds.), Private Capital Flows to Emerging Markets After the Mexican Crisis. Pp. 147-150. Washington, DC: Institute for International Economics.

———. 1995. “A Retrospective on the Debt Crisis,” in Understanding Interdependence, edited by Peter Kenen. Pp. 262-88. Princeton, N.J.: Princeton University Press.

Dornbusch, Rudiger, and Alejandro Werner. 1994. “Mexico: Stabilization, Reform, and No Growth.” Brookings Papers on Economic Activity, 1:253-97.

Dornbusch, Rudiger, and Ilan Goldfajn, and Rodrigo Valdés. 1995. “Currency Crises and Collapses.” Brookings Papers on Economic Activity 2: 219-70.

Dow, James, and Gary Gorton. 1997. “Noise Trading, Delegated Portfolio Management, and Economic Welfare.” Journal of Political Economy 105(5): 1024-51.

Drobak, John, and John Nye, eds. 1997. The Frontier of the New Institutional Economics. New York: Academic Press.

Drucker, Peter. 1997. “The Global Economy and the Nation State.” Foreign Affairs, September/October: 159-171.

Edison, Hali. 1993. “The Effectiveness of Central Bank Intervention.” Special Papers in International Economics. Princeton, N.J.: Princeton University Press.

Edison, Hali, Pongsok Luangavon and Marcus Miller. 1998. “Asset Bubbles, Domino Effects and ‘Lifeboats’: Elements of the East Asian Crisis.” International Finance Discussion Papers No. 606. Washington, D.C.: Board of Governors of the Federal Reserve System.

Edwards, Sebastian. 1998. “The Mexican Peso Crisis: How Much Did We Know? When Did We Know It? The World Economy, January: 1-30.

———. 1995a. Capital Controls, Exchange Rates, and Monetary Policy in the World Economy. Cambridge: Cambridge University Press.

———. 1995b. Crisis and Reform in Latin America: From Despair to Hope. New York: Oxford University Press for the World Bank.

Edwards, Sebastian, and Miguel Savastano. 1998. “The Morning After: The Mexican Peso in the Aftermath of the 1994 Currency Crisis.” NBER Working Paper No. 6516. Cambridge, Mass.: National Bureau for Economic Research, April.

Eichengreen, Barry. 1994. International Monetary Arrangements for the 21st Century. Washington, D.C.: Brookings Institute.

Eichengreen, Barry, and Peter Lindert, eds. 1989. The International Debt Crisis in Historical Perspective. Cambridge, Mass.: MIT Press.

Eichengreen, Barry, and Richard Portes. 1997. “Managing Financial Crisis in Emerging Markets,” in Federal Reserve Bank of Kansas City. Pp. 193-226.

Eichengreen, Barry, Andrew Rose, and Charles Wyplosz. 1996. “Contagious Currency Crises.” NBER Working Paper No. 5681. Cambridge, Mass.: National Bureau for Economic Research.

———. 1996b. “Speculative Attacks on Pegged Exchange Rates: An Empirical Exploration with Special Reference to the European Monetary System,” in Matthew Canzoneri, Wilfred Eithier and Viltorio Grilli, eds., The New Transatlantic Economy. Pp. 191-228. Cambridge: Cambridge University Press.

———. 1995. “Exchange Market Mayhem: The Antecedents and Aftermath of Speculative Attacks.” Economic Policy 21: 249-312.

Eichengreen, Barry, and Charles Wyplosz. 1993. “The Unstable EMS.” Brookings Papers on Economic Activity: 51–143.

Engle, Robert F., Takatoshi Ito, and Wen-Ling Lin. 1990. “Meteor Showers or Heat Waves? Heteroscedastic Intro-Daily Volatility in the Foreign Exchange Market.” Econometrica 59: 525-41.

Federal Reserve Bank of Kansas City. 1997. Maintaining Financial Stability in a Global Economy. Kansas City.

Feldstein, Martin. 1988a. “Thinking About International Economic Cooperation.” Journal of Economic Perspectives 2 (Spring): 3-13.

———. 1988b. “Reforming the IMF.” Foreign Affairs, March/April: 20-33.

Fernald, John, Hale Edison, and Prakash Lugani. 1998. “Was China the First Domino?’ Board of Governors of the Federal Reserve System, International Finance Discussion Papers No. 604, March.

Ffrench-Davis, Ricardo. 1998. “Policy Implications of the Tequilla Effect.” Challenge 41(2): 15-43.

Fischer, Stanley. 1998a. “The Asian Crisis: A View from the IMF.” Reported in IMF Survey 27(2), January.

———. 1988b. “The IMF and the Asia Crisis.” Second Annual Forum Funds Lecture on Public Policy, UCLA, March 20.

———. 1988c. “The Asian Crisis and the changing Role of the IMF.” Finance and Development, June: 2-5.

Fischer, Stanley. 1997. “Commentary” in Maintaining Financial Stability in a Global Economy. Kansas City: Federal Reserve Bank of Kansas City.

Fischer, Stanley. 1995. “The IMF and the World Bank at Fifty,” in Hans Genberg, ed., The International Monetary System. Pp. 171-200. Berlin: Springer.

Flood, Robert and Nancy Marian. 1997. “Speculative Attacks: Fundamentals and Self-Fulfilling Prophecies.” NBER Working Paper No. 5789. Cambridge, Mass.: National Bureau of Economic Research.

Folkerts-Landau, David, Donald Mathieson, and Garry Schinasi. 1997. International Capital Markets: Developments, Prospects and Key Policy Issues. Washington, D.C.: International Monetary Fund.

Folkerts-Landau, David, et al. 1995. “Effects of Capital Flows on the Domestic Financial Sectors in APEC Developing Countries,” in Mohsin Kahn and Carmen Reinhart, eds., Capital Flows in the APEC Region. IMF Occasional Paper No. 122. Washington, D.C.: International Monetary Fund.

Frankel, Jeffrey, ed. 1994. The Internationalization of Equity Markets. Chicago: University of Chicago Press.

———. 1992. “Measuring International Capital Mobility: A Review.” American Economic Review, May: 197-202.

Frankel, Jeffrey, and Kenneth Froot. 1989. “Chartists, Fundamentalists, and the Demand for Dollars,” in Anthony Courakis and Mark Taylor, eds., Policy Issues for Interdependent Economies. London: MacMillan.

———. 1987. “Using Survey Data to Test Standard Propositions Regarding Exchange Rate Expectations.” American Economic Review 77: 133-53.

Frankel, Jeffrey, and Andrew Rose. 1996. “Currency Crashes in Emerging Markets: An Empirical Treatment.” Journal of International Economics 41: 351-66.

Froot, Kenneth A., David S. Scharfstein, and Jeremy C. Stein. 1990. “Herd on the Street: Informational Inefficiencies in a Market with Short-Term Speculation.” NBER Paper No. 3250. Washington, D.C.: National Bureau for Economic Research, February.

Garber, Peter. 1996. “Discussion,” in Ben Bernanke, and Julio Rotenberg, eds., NBER Macroeconomics Annual 1996. Cambridge: MIT Press, pp. 403-6.

Garber, Peter. 1990. “Famous First Bubbles.” Journal of Economic Perspectives 4(2): 35-54.

Garrett, Geoffrey. 1995. “Capital Mobility, Trade, and the Domestic Politics of Economic Policy.” International Organization 49: 657-88.

Garrett, Geoffrey, and Peter Lange. 1996. “Internationalization, Institutions, and Political Change,” in Internationalization and Domestic Politics, edited by Robert O. Keohane and Helen V. Milner. Pp. 48–78. Cambridge: Cambridge University Press.

Giavazzi, Francesco, and Marco Pagano. 1988. “The Advantage of Tying One’s Hands: EMS Discipline and Central Bank Credibility.” European Economic Review 32 (June): 1055–82.

Goldfajn, Ilan, and Rodrigo O. Valdés. 1997. “Are Currency Crises Predictable?” IMF Working Paper WP/97/159. Washington, D.C.: International Monetary Fund.

Goldstein, Morris. 1998.

———. 1997a. “Commentary” in Maintaining Financial Stability in a Global Economy. Pp. 97–116. Kansas City: Federal Reserve Bank of Kansas City.

———. 1997b. “Presumptive Indicators of Vulnerability to Financial Crises in Emerging Markets,” in Priya Basu, ed., Creating Resilient Financial Regimes in Asia. Pp. 79-132. Oxford: Oxford University Press.

Goldstein, Morris, and Guillermo Calvo. 1996. “What Role for the Official Sector,” in Guillermo Calvo, Morris Goldstein, and Educard Hochreiter (eds.), Private Capital Flows to Emerging Markets After the Mexican Crisis. Pp. 233-282. Washington, DC: Institute for International Economics.

Goodman, John, and Louis Pauly. 1993. “The Obsolescence of Capital Controls?” World Politics 46: 50-82. The Asian Financial Crisis. Washington, D.C.: Institute for International Economics.

Greenwood, John C. 1998. “Asia’s Financial Crisis: Causes, Cures and Outlook.” Fraser Forum, May: 11-19.

Guttentag, Jack M., and Richard J. Herring. 1987. “Emerging Liquidity Assistance for International Banks,” in Richard Portes and Alexander Swadoba, eds., Threats to International Financial Stability. Pp. 150-85. Cambridge: Cambridge University Press.

———. 1986. “Disaster Myopia in International Banking.” Princeton Essays in International Finance No. 164.

———.1985. “Commercial Bank Lending to Developing Countries,” in G.W. Smith and J.T. Cuddington, eds., International Debt and the Developing Countries. Washington, D.C.: International Bank for Reconstruction and Development.

Haggard, Stephen, and Sylvia Maxfield. 1996. “The Political Economy of Financial Internationalization in the Developing World.” International Organization, Winter: 35-68.

Hale, David. 1996. “Comment on Private Capital Flows to Emerging Markets after the Mexican Crisis,” in Guillermo Calvo, Morris Goldstein, and Educard Hochreiter (eds.), Private Capital Flows to Emerging Markets After the Mexican Crisis. Pp. 132-146. Washington, DC: Institute for International Economics.

Helleiner, Eric. 1994. States and the Reemergence of Global Finance: From Bretton Woods to the 1990s. Ithaca, N.Y.: Cornell University Press.

Henning, C. Randall. 1996. “Political Economy of the Bretton Woods Institutions: Adopting to Financial Change.” The World Economy: 173-93.

Hogan, Kedreth C., Jr. and Michael Melvin. 1994. “Sources of Meteor Showers and Heat Waves in the Foreign Exchange Market.” Journal of International Economics 37: 239-47.

Hopper, Gregory. 1997. “What Determines the Exchange Rate: Economic Factors or Market Sentiment?” Federal Reserve Bank of Philadelphia Business Review, September/October: 17-29.

Isard, Peter. 1995. Exchange Rate Economics. Cambridge: Cambridge University Press.

Ito, Takatoshi, Marcel Cassard, and David Folkerts-Landau. 1996. International Capital Markets. Washington, D.C: International Monetary Fund.

Johnson, Christopher, and Stefan Collignon, eds., 1994. The Monetary Economics of Europe: Causes of the EMS Crisis. Rutherford: Fairleigh Dickinson University Press.

Kahler, Miles. 1992. “External Influence, Conditionality, and the Politics of Adjustment,” in Stephan Haggard and Robert Kaufman, eds., The Politics of Economic Adjustment. Pp. 90–138. Princeton: Princeton University Press.

Kalpena, Kochhav, et al. 1996. “Thailand: The Road to Sustained Growth.” IMF Occasional Paper No. 146. Washington, D.C.: International Monetary Fund, December.

Kaminsky, Graciela, Saul Lizondo, and Carmen M. Reinhart. 1998. “Leading Indicators of Currency Crises.” IMF Staff Papers 45(1): 1-48.

Kaminsky, Graciela, and Carmen M. Reinhart. 1996. “The Twin Crises: The Causes of Banking and Balance of Payment Problems.” International Finance Discussion Paper No. 544. Washington, D.C.: Board of Governors of the Federal Reserve System, March.

Kamphuis, Robert, Roger Kormendi, and J.W. Watson, eds. 1989. Black Monday and the Future of Financial Markets. Homewood, Ill.: Dow-Jones-Irwin.

Kapor, Devish. 1998. “The IMF: A Cure or a Curse?” Foreign Policy, Summer: 114-131.

Kapstein, Ethan B. 1994. Governing the Global Economy: International Finance and the State. Cambridge, Mass.: Harvard University Press.

Katada, Saori. 1998. “Collective Management of International Financial Crises.” Center for International Studies, University of Southern California.

Kaufman, Henry. 1995. “Opening Remarks,” in Mexico: Why Didn’t Wall Street Sound the Alarm? Washington, DC.: Group of Thirty.

Kawai, Masahiro. 1998. “The East Asian Crisis: Causes and Lessons.” Contemporary Economic Policy, April: 157-72.

Kehoe, Timothy J. 1996. “Discussion,” in Ben Bernanke, and Julio Rotenberg, eds., NBER Macroeconomics Annual 1996. Pp. 378-92. Cambridge: MIT Press.

Keohane, Robert O., and Helen V. Milner, eds. 1996. Internationalization and Domestic Politics. Cambridge: Cambridge University Press.

Kindleberger, Charles. 1978. Manias, Panics and Crashes: A History of Financial Crises. New York: Basic Books.

Krugman, Paul. 1998a. “What Happened to Asia.” http://web.mit.edu/krugman/www/ DISINTER.html.

———. 1998b. “Will Asia Bounce Back?” http://web.mit.edu/krugman/www/suisse.html. March.

———. 1996. “Are Currency Crises Self-Fulfilling?” in Ben Bernanke and Julio Rotenberg, eds., NBER Macroeconomics Annual 1996. Pp. 345-77. Cambridge: MIT Press.

Krugman, Paul, and Maurice Obstfeld. 1994. International Economics. New York: Harper Collins.

Kuran, Timur. 1995. Private Truths, Public Lies. Cambridge, Mass.: Harvard University Press.

Kurzer, Paulett. 1993. Business and Banking: Political Change and Economic Integration in Western Europe. Ithaca, N.Y.: Cornell University Press.

Lane, Timothy. 1993. “Market Discipline.” IMF Staff Papers, March: 53-88.

Larrain, G., Helmut Reisen, and J. von Maltzon. 1997. “Emerging Market Rish and Sovereign Credit Ratings.” OECD Development Centre Technical Paper No. 124.

Levich, Richard. 1994. “International Equity Transactions,” in Jeffrey Frankel, ed., The Internationalization of Equity Markets. Pp. 221-27. Chicago: University of Chicago Press.

Lewis, Karen. 1989. “Can Learning Affect Exchange Rate Behavior?” Journal of Monetary Economics, January: 79-100.

Lewis, Michael. 1998. “The World’s Biggest Going-Out-of-Business Sale.” New York Times Magazine, May 31, p. 2.

Lewis-Beck, Michael. 1988. Economics and Elections. Ann Arbor: University of Michigan Press.

Lindert, Peter and Thomas Pugel. 1996. International Economics. Chicago: Irwin.

Lipsky, John. 1998. “Asia’s Crisis: A Market Perspective.” Finance and Development, June: 10-13.

Makin, John. 1998a. “Is the Asian Crisis Over?” American Enterprise Institute Economic Outlook, February.

———. 1998b. “Asia’s Crisis Is Not a Currency Crisis.” American Enterprise Institute Economic Outlook, May.

Malkiel, Burton and J. P. Mei. 1998. Global Bargain Hunting. New York: Simon and Schuster.

Masson, Paul, and Mark Taylor, eds. 1993. Policy Issues in the Operation of Currency Unions. Cambridge: Cambridge University Press.

McKinnon, Ronald, and Huw Pill. 1996. “Credible Liberalization and International Capital Flows: The ‘Overborrowing Syndrome’” in Takatoshi Ito and Anne O. Kruger, eds., Financial Deregulation and Regulation in East Asia. Pp. 7-42. Chicago: University of Chicago Press.

Meese, Richard, and Andrew Rose. 1998. “Exchange Rate Instability: Determinants and Predictability,” in Reuven Glick (ed.), Managing Capital Flows and Exchange Rates. Cambridge: Cambridge University Press.

Meigs, A. James. 1997. “Mexican Monetary Lessons.” Cato Journal 17(1): 35-72.

Milesi-Ferretti, Gian Maria, and Assaf Razin. 1996a. “Current-Account Sustainability: Selected East Asian and Latin American Experiences.” IMF Working Paper WP/96/110. Washington, D.C.: International Monetary Fund, October.

Milesi-Ferretti, Gian Maria, and Assaf Razin. 1996b. Current-Account Sustainability. Princeton Studies in International Finance No. 81. Princeton, N.J.: Princeton University Press.

Miller, Norman. 1998. “Explaining Exchange Rate Puzzles.” Presented at the Western Economics Association Meetings, Lake Tahoe, June 28-July 2.

Milner, Helen. 1997. Interests, Institutions, and Information. Princeton, N.J.: Princeton University Press.

Milner, Helen V., and Robert O. Keohane, eds. 1996. “Internationalization and Domestic Politics: An Introduction” in Internationalization and Domestic Politics, edited by Robert O. Keohane and Helen V. Milner. Pp. 3–24. Cambridge: Cambridge University Press.

Mishkin, Frederic S. 1997. “The Causes and Propagation of Financial Instability,” in Federal Reserve Bank of Kansas City. Pp. 55-96.

Neeley, Christopher J. 1997. “Technical Analysis in the Foreign Exchange Market.” Federal Reserve Bank of St. Louis Review, September/October: 23-38.

Nordhaus, William D. 1989. “Alternative Approaches to the Political Business Cycle.” Brookings Papers on Economic Activity 2: 1-49.

Nurske, Ragnar. 1947. International Currency Experience. New York: United Nations.

Obstfeld, Maurice. 1996. “Discussion,” in Ben Bernanke, and Julio Rotenberg, eds., NBER Macroeconomics Annual 1996. Pp. 393-403. Cambridge: MIT Press.

———. 1995a. “The Logic of Currency Crises,” in Barry Eichengreen, Jeffry Frieden, and Jürgen von Hagen, eds., Monetary and Fiscal Policy in an Integrated Europe. Pp. 62-90. Berlin: Springer.

———. 1995b. “International Capital Markets in the 1990s” in Peter Kenen, ed., Understanding Interdependence. Pp. 201–61. Princeton, N.J.: Princeton University Press.

Obstfeld, Maurice, and Kenneth Rogoff. 1995. “The Mirage of Fixed Exchange Rates.” Journal of Economic Policy 9(4): 73-96.

Osler, Carol L. 1995. “Exchange Rate Dynamics and Speculator Horizons.” Journal of International Money and Finance 14: 695-720.

Ostry, Jonathan. 1997. “Current Account Imbalances in ASEAN Countries: Are They a Problem?” IMF Working Paper No. 51, April.

Ötker, Inci, and Ceyla Pazarbasioglu. 1995. “Speculative Attacks and Currency Crises: The Mexican Crises, the Mexican Experience.” IMF Working Paper WP/95/112. Washington, D.C.: International Monetary Fund.

Park, Yung Chul. 1998. “Investment Booms, Financial Bust.” Brookings Review, Summer: 14.

Parker, Stephen. 1998. “Southeast Asia’s Struggle through Crisis.” Brookings Review, Summer: 10-13.

Pauly, Lewis. 1997. Who Elected the Bankers? Ithaca, N.Y.: Cornell University Press.

Porter, Tony. 1996. “Capital Mobility and Currency Markets: Can They be Tamed?” International Journal, Autumn: 669-89.

Portes, Richard. 1993. “EMS and EMU After the Fall.” World Economy 16: 1-16.

Price, Simon. 1997. “Political Business Cycles and Macroeconomic Credibility: A Survey.” Public Choice 92: 407-27.

Radelet, Steven, and Jeffrey Sachs. 1997. “Asia’s Reemergence.” Foreign Affairs, November/December: 44-59.

Redding, John, and James Kynge. 1998. “Complacency Gives Way to Contagion.” Financial Times, January 13, p. 8.

Ries, Christine, and Richard J. Sweeney, eds. 1997. Capital Controls in Emerging Market Economies. Boulder, Colo.: Westview Press.

Reisen, Helmut. 1998. “Green Light for Danger.” Financial Times, February 3, p. 12.

Richards, Anthony. 1996. “Volatility and Predictability in National Stock Markets: How Do Emerging and Mature Markets Differ?” IMF Staff Papers, September: 456-501.

Rodrick, Dani. 1998. “Who Needs Capital-Account Convertibility?” in Peter B. Kenen, ed., Princeton Essays in International Finance.

———. 1996. “Understanding Economic Policy Reforms.” Journal of Economic Literature, March: 9-41.

Roll, Richard. 1988. “The International Crash of October 1987.” Financial Analysts Journal, September/October: 19-35.

Rose, Andrew, and Lars E. O. Svennson. 1995. “Macroeconomic and Political Determinants of Realignment Expectations: Some European Evidence,” in Barry Eichengreen, Jeffry Frieden, and Jürgen von Hagen, eds., Monetary and Fiscal Policy in an Integrated Europe. Pp. 91-117. Berlin: Springer.

Roubini, Nouriel. 1997-1998. “Asia Chronology.” http://www.stern.nyu.edu/~nroubini/asia/asia chronology.

Sachs, Jeffrey. 1997. “Commentary” in Maintaining Financial Stability in a Global Economy. Pp. 233–42. Kansas City: Federal Reserve Bank of Kansas City.

———. 1997. “IMF Is a Power Unto Itself.” Financial Times December 11.

Sachs, Jeffrey, Aaron Tornell and Andrés Velasco. 1996a. “The Collapse of the Mexican Peso: What Have We Learned?” Economic Policy, April: 15–56.

———. 1996b. “Financial Crises in Emerging Markets: The Lessons from 1995.” Brookings Papers on Economic Activity 1: 147-215.

Salinas-Léon, Roberto. 1997 “The Mexican Peso Crisis: Comment on Meigs.” Cato Journal, Spring/Summer: 73-80.

Sargent, Thomas. 1993. Bounded Rationality in Macroeconomics. Oxford: Clarendon Press.

Saving, Thomas, et al. 1998. “Thailand’s Exchange-Rate Crisis.” Contemporary Economic Policy, April: 136-56.

Schieler, Max. 1995. “Discussion,” in Hans Genberg, ed., The International Monetary System. Pp. 201-208. Berlin: Springer.

Schiller, Robert J. 1997. “Why Do People Dislike Inflation?” in Christina Romer and David Romer, eds., Reducing Inflation. Chicago: University of Chicago Press.

———.1995. “Conversation, Information, and Herd Behavior.” American Economic Review, May: 181-85.

———.1990. “Speculative Prices and Popular Models.” Journal of Economic Perspectives 4(3): 55-65.

———. 1984. “Stock Prices and Social Dynamics.” Brookings Papers on Economic Activity: 457-98.

Schuknecht, L. (1996). “Fiscal Cycles and the Exchange Rate Regime in Developing Countries.” Mimeo. Washington, D.C.: International Monetary Fund.

Schwartz, Anna. 1998. “Banking Crises in the Nineties.” Presented at the Western Economics Association Meetings, Lake Tahoe, June 28-July 2.

Shleifer, Andrei, and Lawrence Summers. 1990. “The Noise Trader Approach to Finance.” Journal of Economic Perspectives 4(2): 19-33.

Slater, Robert. 1996. Soros. New York: McGraw Hill.

Soros, George. 1987. Alchemy of Finance. New York: Simon and Schuster.

Stulz, René. 1994. “International Experiences with Securities Transactions Taxes: Comment,” in Jeffrey Frankel, ed., The Internationalization of Equity Markets. Pp. 303-308. Chicago: University of Chicago.

Sweeney, Richard J., Clas Wihlborg, and Thomas D. Willett, eds. 1998. Exchange Rate Policies for Emerging Market Economies. Boulder, Colo.: Westview Press.

Tavlas, George. 1993. “The ‘New’ Theory of Optimum Currency Areas.” The World Economy 33: 21-40.

Thoenes, Sander. 1998. “Indonesia: The Offshore Borrowing Trap.” Financial Times, January 12, p. 7.

Tornell, Aaron, and Andris Velasco. 1995. “Fixed versus Flexible Exchange Rates: Which Provides More Fiscal Discipline? NBER Working Paper No. 5108. Cambridge, Mass.: National Bureau of Economic Research.

Truman, Edwin. 1996. “Comment on What Role for the Official Sector,” in Guillermo Calvo, Morris Goldstein, and Educard Hochreiter (eds.), Private Capital Flows to Emerging Markets After the Mexican Crisis. Pp. 287-292. Washington, DC: Institute for International Economics.

Ul Haq, Mahbub, Ingre Kaul, and Isabelle Grunberg, eds. 1996. The Tobin Tax: Coping with Financial Volatility. Oxford: Oxford University Press.

Vaubel, Roland. 1991. “The Political Economy of the International Monetary Fund,” in Roland Vaubel and Thomas D. Willett, eds., The Political Economy of International Organizations. Boulder, Colo.: Westview Press.

Vertzberger, Yaacov. 1990. The World in Their Minds. Palo Alto: Stanford University Press.

Von Furstenberg, George. 1998. “From Worldwide Capital Mobility to International Financial Integration: A Review Essay.” Open Economies Review, January.

Von Hagen, Jürgen. 1995. “Discussion,” in Hans Genberg, ed., The International Monetary System. Pp. 209-15. Berlin: Springer.

Wadhwani, Sushil. 1996. “Explaining the Crisis and Preventing Future Crisis,” in Guillermo Calvo, Morris Goldstein, and Educard Hochreiter (eds.), Private Capital Flows to Emerging Markets After the Mexican Crisis. Pp. 302-306. Washington, DC: Institute for International Economics.

Walter, Alan. 1990. Sterling in Danger. London: Fontanu Paperbacks for the Institute of Economic Affairs.

Westbrook, Jilleen, and Thomas D. Willett. 1998. “Exchange Rates as Nominal Anchors,” in Richard J. Sweeney, Clas Wihlborg, and Thomas D. Willett, eds. 1998. Exchange Rate Policies for Emerging Market Economies. Boulder, Colo.: Westview Press, forthcoming.

Wigmore, Barrie. 1998. “Revisiting the October 1987 Crash.” Financial Analysts Journal, January/February: 36-48.

Wihlborg, Clas, and Thomas D. Willett. 1991. “Optimum Currency Areas Revisited on the Transition Path to a Monetary Union,” in Clas Wihlborg, Michele Fratianni, and Thomas D. Willett, eds., Financial Regulation and Monetary Arrangements After 1992. Pp. 279–97. Amsterdam: North Holland.

Willett, Thomas D. 1998a. “Credibility and Discipline Effects of Exchange Rates as Nominal Anchors.” Claremont Working Paper, Claremont Graduate University.

———. 1998b. “A Political Economy Analysis of the Maastricht and Stability Pact Fiscal Criteria,” in Andrew Hughes Hallett, Michael Hutchison and Svend Jensen, eds., Fiscal Aspects of Monetary Integration. Cambridge: Cambridge University Press.

———.1988. Political Business Cycles: The Political Economy of Money, Inflation and Unemployment. Durham, N.C.: Duke University Press for the Pacific Research Institute for Public Policy.

———.1980. International Liquidity Issues. Washington, D.C.: American Enterprise Institute.

———.1977. Floating Exchange Rates and International Monetary Reforms. Washington, D.C.: American Enterprise Institute.

Willett, Thomas D., and Young Seok Ahn. 1998. “Upward Biases in Estimates for Capital Mobility for Developing Countries.” Claremont Working Paper. Claremont Graduate University.

Willett, Thomas D., and King Banaian. 1988. “Models of the Political Process and their Implications for Stagflation,” in Political Business Cycles: The Political Economy of Money, Inflation and Unemployment. Durham, N.C.: Duke University Press for the Pacific Research Institute for Public Policy.

Willett, Thomas D., and John Mullen. 1982. “The Effects of Alternative International Monetary Systems on Macroeconomic Discipline and Inflationary Biases,” in Raymond E. Lombra and Willard E. Witte, eds., Political Economy of International and Domestic Monetary Relations. Pp. 143–55. Ames, Iowa: Iowa State University Press.

Willett, Thomas D., and Clas Wihlborg. 1991. “International Capital Flows, the Dollar, and U.S. Financial Policies,” in William Haraf and Thomas D. Willett, eds., Monetary Policy for a Volatile World Economy. Pp. 51–88. Washington, D.C.: American Enterprise Institute.

Williams, Carol. 1998. “Russian’s Top Bankers Vow to Help.” Los Angeles Times, June 3, p. A4.

Williamson, John. 1995. What Role for Currency Boards? Washington, D.C.: Institute for International Economics.

World Bank. 1997. Private Capital Flows to Developing Countries. Oxford: Oxford University Press.

Wyplosz, Charles. 1997. “EMU.” Journal of Economic Perspectives, Fall: 3-22.

Endnotes

*: I am indebted to participants in workshops at Claremont McKenna College and the University of Southern California and to David Andrews, Marc Bremmer, Aida Budiman, James Dean, Jo Gab-Je, Timur Kuran, Norman Miller, Ekniti Nitithenprapas, John Thomas and Jilleen Westbrook for helpful comments.

This paper was presented at the Western Economics Association Meetings, Lake Tahoe, June 29, 1998 Back.

Note 1: Or with Peter Drucker (1997:162) that “while the hope that governments will practice self-discipline is fantasy, the global economy imposes new and more severe restraints on government. It is forcing government back into fiscal responsibility.” Note however that Drucker does not see the standard economists’ world of capital flows based on rational expectations. Focusing on world money generated by currency trading, which he labels “virtual rather than real money,” he argues that “its power is real... This virtual money has total mobility because it serves no economic function...And because it serves no economic function and finances nothing, this money does not follow economic logic or rationality. It is volatile and easily panicked by a rumor or unexpected event” (p. 162). While concluding that “the global economy is the ultimate arbiter of monetary and fiscal policies,’ he argues that “currency runs...are not the appropriate cure for fiscal irresponsibility. In the case of Mexico, they were worse than the disease...But so far there is no other control on fiscal irresponsibility” (p. 163). Back.

Note 2: The farsighted rational expectations assumption is particularly widely used in game theory models of political economy. For a recent survey see Rodrick (1996). Charges that speculative capital flows generate major irrational bandwagon effects and are an independent source of financial instability come primarily from non-economists today, but, stimulated in part by Nurske’s (1947) influential study of instability during the interwar period, such views were common along economists at the time the Bretton Woods international monetary system was negotiated (see Helleiner 1994, and Willett 1977). This explains why capital account liberalization was not an objective of the negotiators. Since then, mainstream economic theory has swung in favor of freedom of capital flows. See the discussions and references in Ries and Sweeney (1997) and Edwards (1995). This has been reflected in a widespread move toward financial liberalization. See Cohen (1996), Goodman and Pauly (1993), Haggard and Maxfield (1996), and Helleiner (1994). With the recent crisis, however, the number of critics of free capital mobility is on the rise again. See, for example, Bhagwati (1998) and Rodrick (1998). Back.

Note 3: On the evolution of speculative crisis models and the difficulties of distinguishing between irrational, fundamentals-based or self-fulfilling speculative attacks, see the valuable exchanges among Krugman (1996), Kehoe (1996), Obstfeld (1996) and Garber (1996), and also Eichengreen, Rose and Wyploz (1995) and Obstfeld (1995). Back.

Note 4: The view that exchange rates can be a potent nominal anchor against inflation has been quite influential at the International Monetary Fund and has formed the basis of many countries’ exchange rate strategies. For an example of advocacy for using exchange rates as a nominal anchor see Bruno (1991), and for a survey and critical assessment of the literature see Westbrook and Willett (1998) and Willett (1998). Back.

Note 5: Of course, in many cases domestic and international approaches should be seen as complements rather than substitutes. See, for example, Willett (1998a, 1998b). Back.

Note 6: Recently in the United States both the far right and the far left have made common cause (for quite different reasons) against the IMF and have rallied sufficient numbers in Congress to make passage of the 1998 quota increase for the IMF an iffy proposition. For a range of recent academic views on the appropriate scope of activities for the IMF, see Feldstein (1998b), Fischer (1995), Henning (1996), Kapor (1998), Sachs (1997), Schieler (1995), Vaubel (1991), and Von Hagen (1995). For perhaps the most cogent defense of IMF policies, see Fischer (1998b, 1998c). Back.

Note 7: In addition to the questions about the effectiveness of international capital flows as a source of discipline explored here, there are also important issues concerning the potential threat of global financial markets to the political legitimacy of states. See Pauley (1997). Back.

Note 8: See, for example, Andrews (1994), Garrett (1995) and Westbrook and Willett (1998). For discussions of the conditions necessary for financial market discipline to be effective, see Goldstein and Calvo (1996) and Lane (1993). Back.

Note 9: For discussion of other possible explanations for the convergence of inflation among the industrial countries, see Andrews and Willett (1977). Back.

Note 10: On the history of the discipline argument see Willett and Mullen (1982). On the resurgence of interest in the argument put in modern theoretical terminology as a commitment device, see Willett (1998). A particularly influential contribution is Giavazzi and Pagino (1988). Back.

Note 11: As will be discussed below, as long as insolvency was not at issue, the effects of high capital mobility could be to actively loosen discipline over fiscal policy. In models with perfect capital mobility, recognition of a long-run inconsistency among policies would provoke an immediate crisis. See, for example, Daniels (1997). Back.

Note 12: The time inconsistency problem focuses on the prospective benefits to government of generating expectational surprises, thus leading to an inflationary bias. Barro and Gordon (1983) present an influential version of the argument. Although the technical formulation of these time inconsistency models is quite different, the basic idea is similar to the traditional political business cycle argument that governments have incentives to generate expansions before elections so that the benefits of increased employment and output come primarily before the election while the costs of higher inflation come primarily after the election is safely over. See, for example, Alesina, Roubini and Cohen (1997) and Willett (1988). Back.

Note 13: So far there has been relatively little formal testing of these propositions. Schuknecht (1996) found that political business cycles in fiscal policy were more prevalent under pegged than flexible exchange rates. Clark and Reichert (1998), however, find that high capital mobility makes political business cycles in output less likely for OECD countries under pegged rates. Back.

Note 14: These are some of the major criteria emphasized in the literature on the theory of optimum currency areas. See, for example, DeGrauwe (1992), Masson and Taylor (1993), Tavlas (1993) and Wihlborg and Willett (1991). Back.

Note 15: See, for example, Buiter, Corseti and Pesenti (1998), Obstfeld (1995, 1996) and Eichengreen, Rose and Wyplosz (1996). There is likely to be a continuing debate on this issue since, as Meese and Rose (1997) point out, the econometrician cannot distinguish between changed expectations about the future course of the fundamentals and self-fulfilling expectations. It is interesting to note that an early version of the self-fulfilling speculation hypothesis was developed by the noted speculator George Soros. See Soros (1987) and Slater (1996). Note that the self-fulfilling expectations models do not assume that speculation is divorced from concern with fundamentals, but rather that there may be a zone in which the fundamentals do not mandate a rational speculative attack but do not prevent it either. In the French case the concern was whether the French government would be willing to tolerate continued high interest rates and unemployment in order to defend the franc. See also Buiter et al. (1998) and Johnson and Collignon (1994). Back.

Note 16: On the moral hazard issues involved in lender of last resort activities, see the analysis and references in Guttentag and Herring (1987), and on their relationship to the 1982 debt crisis see Cline (1995), Dooley (1995) and Kapstein (1994). Back.

Note 17: For recent surveys and discussions of the empirical literature on the decree of international capital mobility and efficiency of the foreign exchange market, see Edwards (1995a), Frankel (1992), Hopper (1997), Isard (1995), Neeley (1997), Obstfeld (1995), Von Furstenberg (1998), Willett and Ahn (1998), and Willett and Wihlborg (1991). Back.

Note 18: It is interesting to note that the international credit rating agencies explicitly take into account estimates of the probability of government support in their ratings. Back.

Note 19: See Willett and Banaian (1988). Back.

Note 20: For empirical evidence on these issues, see the contributions and references in Alesina, Roubini and Cohen (1997), Lewis-Beck (1988), Price (1997) and Willett (1988). Back.

Note 21: For a valuable review of the implications of recent research from the new institutional economics for public policy analysis, see Dixit (1996). On the contributions of behavioral and the new institutional economics, see also Drobak and Nye (1997) and Kuran (1995). Back.

Note 22: See, for example, the surveys by Baillie and McMahon (1989) and Isard (1995). Back.

Note 23: Elements of this view are also held by the most famous speculator of our day, George Soros. See Soros (1987) and Slater (1996). For recent attacks on the case for free capital mobility, see Bhagwati (1998) and some of the contributions in Ul Hags, Kaul and Grunberg (1996). Back.

Note 24: Noise trading can also give rise to short-term volatility spillovers from one market to another, but it is difficult to distinguish this from shocks that have cross-border implications For efforts to empirically disentangle these effects in foreign exchange and equity markets, see Engle et al. (1990) and Hogan and Melvin (1994). On the contagion effects of currency crises, see Eichengreen et al. (1996a) and the discussion below. Back.

Note 25: Also see the confirming comments by Wadhwani (1996). Back.

Note 26: Some, however, hold that this overlending was caused primarily by rational responses to the moral hazard problem based on expectations, which proved to be partially true, of home government bailouts in the case that things went bad. See, for example, Dooley (1996). Back.

Note 27: Most of the literature takes the time horizon of speculators as exogenous. However, Osler (1995) addresses the issue of the endogenous determination of speculative time horizons, showing how the time horizons of rational speculators will depend on factors such as the number of speculators, their degree of risk aversion, and the patterns of disturbances to the fundamentals. Osler’s model is able to explain the tendency of the foreign exchange market to approximate a random walk during normal periods, but display mean reversion, i.e., overshooting, in the face of large shocks. Back.

Note 28: After this paper was substantially completed, I came across a concise statement of this view by Charles Wyplosz (1997:14). “History suggests skepticism about the ability of markets to impose discipline ... markets tend to throw good money after bad for a time...When markets do react, it is often too late and too violently.” Back.

Note 29: This is not to argue that there are never almost purely speculative swings of considerable magnitudes—especially in stock markets. For example, while we can point to some partial explanations of the Black Monday stock market crash in 1987, it is hard to square that episode with even the intermediate version of speculative behavior presented here, much less with full-fledged rational expectations views. (For discussions of possible explanations of the Black Monday crash, including conflicts over international policy coordination, see Feldstein 1988, Kamphuis et al. 1989, Roll 1988, and Schiller 1990,1995). Back.

Note 30: It should also be noted that contrary to the popular view in the literature on internationalization, see e.g., Keohane and Milner (1996) and Milner (1997), it is not true that high international capital mobility always reduces the effectiveness of domestic macroeconomic policies. Indeed, the standard results of the widely used Mundell-Fleming model are that high capital mobility will increase the strength of monetary policy under flexible exchange rates and the strength of fiscal policy under fixed exchange rates. See, for example, any standard international economics text such as Caves, Frankel, and Jones (1996), Krugman and Obstfeld (1994), or Lindert and Pugel (1996). Back.

Note 31: For discussions of these cases see the analysis and references in Andrews and Willett (1997) and Feldstein (1988). Back.

Note 32: My discussion here is based on newspaper reports and conversations with several Wall Street economists. This would be an interesting subject for an in-depth case study. Back.

Note 33: ndeed, in the cases of the ERM currencies it is not clear that there was more than momentary overshooting. An overshooting pattern is strongest for the French franc, but there may have been plausible ex ante, albeit incorrect ex post, expectations that the French would ease monetary policy in the face of their domestic recession. Gradual learning that the government was going to stick with a tight money policy would then account for the gradual strengthening of the franc back to approximately its initial level. Back.

Note 34: In part the severity of the market reactions can be explained by the revelations that levels of bad debt and international obligations were far greater than had been realized. Financial Times journalist Sander Thoenes (1988) concludes “The lesson of Indonesia’s experience is that the underlying problems of the region—particularly corporate foreign currency debt—were greater than had been realized back in the autumn” (p. 2). But it would be difficult to argue that there was not also an element of panic and herd behavior at play. Back.

Note 35: Quoted in Thoenes (1998:7). Similar statements have recently been made by current Secretary of the Treasury Rubin. See, for example, Art Pine, “Rubin Warns Against Forcing Banks to Take Bigger Asia Hit,” Los Angeles Times, January 22, 1988. Back.

Note 36: Edwards (1998) offers a similar estimate. On the political and economic factos generating the Mexican crisis, see Auerbach (1997), Blaine (1998), Calvo (1996), Calvo and Mendoza (1996), Dornbusch and Werner (1994), Edwards (1998), Meigs (1997), tker and Pazarbasioglu (19 ), Sachs, Tornell, and Velasco (1996) and Salinas-Lon (1997). Back.

Note 37: For further discussion of the Lawsen dogma and other “popular” exchange rate models, see Dornbusch et al. (1995). On the use of popular models in financial markets more generally, see Schiller (1990). Back.

Note 38: On the at least quasi-rationality of herd behavior in the acceptance of mental models and in financial market behavior, see the analysis and references in Kuran (1995) and Schiller (1990, 1995). Back.

Note 39: For surveys of the debate over the degree of effectiveness of sterilized intervention in the foreign exchange market, see Edison (1993), and Dominquez and Frankel (1993). Back.

Note 40: This review of the Asian episodes is based mainly on reports from The Economist, The Financial Times, The Los Angeles Times, The New York Times, and the Wall Street Journal. A useful chronology has been compiled by Nouriel Roubini and is available from his web site http://www.stern.nyo.edu/nroubini. Back.

Note 41: See, for example, Edwards and Savastano (1998). Back.

Note 42: See also Krugman (1998a, 1998b). Back.

Note 43: In the Mexican case, the lack of an initial clearly announced macroeconomic policy strategy was also important. Back.

Note 44: For discussions of excessive credibility in the context of the EMS, see Artus and Bourguinat (1994). Back.

Note 45: See also Buiter et al. (1998), Dornbusch et al. (1995), and Eichengreen et al. (1994, 1995, 1996). Back.

Note 46: See Goldstein and Calvo (1996), Edwards (1998) and Obstfelt and Rogoff (1995). Back.

Note 47: Interestingly, in regard to the view that international investment is a cause of instability, it was Mexican nationals and not international investors that led the exodus. This appears to have been due in part to consultations by the government with business leaders. See Dornbusch et al. (1995). Back.

Note 48: For a recent discussion of moral hazard problems in this context, see Dooley (1997), Krugman (1998a) and McKinnon and Pill (1996). Back.

Note 49: For examples of modeling the behavior of foreign exchange markets where there is uncertainty about whether there has been a shift in policy regimes, see Lewis (1989). Back.

Note 50: The crises of the 1990s have stimulated a great deal of empirical research on the determinants of exchange rate crises. See, for example, Frankel and Rose (1996) Goldfajn and Valdés (1997), Goldstein (1997a, 1997b) and Kaminsky et al. (1997). Back.

Note 51: In Korea and Thailand, however, slumping stock and real estate markets also gave clues that the risk of crisis was increasing. Back.

Note 52: Frankel and Rose (1996) find a high proportion of direct investment to total capital flows to be associated with a lower probability of crisis. It is widely assumed that direct investment will be less volatile than portfolio flows. However, as a recent World Bank study (1997:150) points out, “the distinction between FDI and portfolio flows should not however be exaggerated. It has been argued, for example, that an FDI investor who wants to get out of a country can borrow on the domestic market against his investment and take his money out” (p. 150). On this composition issue, see also the analysis and references in Chuhan, Perez-Quiras and Popper (1997). Back.

Note 53: On current account sustainability, see Milesi-Ferretti and Razin (1996). Back.

Note 54: It is surprising that Ostry, however, refers to “the absence of significant exchange rate misalignment” (p. 23). As is discussed below, this judgment does not coincide with most assessments. Back.

Note 55: See, for example, Lipsky (1998). Back.

Note 56: Brown et al. (1998) find that hedge funds did not cause the Asian crisis. This of course suggests that they were not ex post efficient speculators. Back.

Note 57: See Braun, Mukherji and Runkle (1996) and Meigs (1997). Back.

Note 58: Bardacke (1998). Back.

Note 59: See, for example, Fischer (1997, 1998). Back.

Note 60: Of at least equal importance were the ways in which the available data was analyzed. See, for example, Goldstein and Calvo (1996) and Truman (1996). Back.

Note 61: The credit-rating agencies primarily lag rather than anticipate developments. See Reisen (1998). One possible contributor to market overreactions is the requirement by many regulating agencies that financial firms unload their holdings of assets whose ratings drop below investment grade. Another source comes from the weighting schemes of index funds, which lead to increased holding of appreciating markets and decreased holdings of declining markets. On the former problems see Reisen (1998) and on the latter see Malkiel and Mei (1998). Back.

Note 62: See also Richards (1996) and Bekaert and Harvey (1995). Back.

Note 63: On the Mexican contagion effects see also Calvo et al. (1996), Dornbusch et al. (1995), Folkerts-Landau et al. (1995) and World Bank (1997). Back.

Note 64: In this regard the speculative attacks on the Hong Kong dollar were an intermediate case that may be an example of markets acting on limited information. There was good reason to believe that the depreciation of other Asian currencies had left the Hong Kong dollar overvalued, but in my judgment the combination of Hong Kong’s huge international reserve position and, more importantly, its many years operating its monetary policy under a type of currency board arrangement meant that devaluation was very unlikely. I suspect that many investors did not fully understand the differences between the adjustably pegged currencies of Thailand and the Philippines and the currency board based fixed exchange rate of Hong Kong. A Wall Street Journal editorial “Indonesian by the Board” (February 11, 1998, A22) makes a similar suggestion. There were also substantial attacks on Argentina’s currency board backed fixed exchange rate during the Mexican peso crisis, but, while proved wrong, these seem to have had a sounder basis since Argentina’s commitment to a currency board was much more recent and had not yet earned full credibility. For discussions of currency board backing for fixed exchange rates, see Baliño (1997), Sweeney, Wihlborg, and Willett (1998) and Williamson (1995). Back.

Note 65: See, for example, Chann and Kasa (1998), Makin (1998), and Radelet and Sachs (1997). For an argument that the effects of the Chinese devaluation have typically been overstated, see Fernald, Edison, and Longani (1998). Back.

Note 66: See, for example, Chinn (1998) and Chann and Kasa (1998). Back.

Note 67: Some would judge unwarranted contagion effects to have been broader. Kawai (1998:169), for example, expresses the view that “the severity of the currency attacks [on the Philippine peso the Indonesian rupiah, and the Malaysian ringgit] ...was puzzling.” Back.

Note 68: Of course, as Chinn emphasized, we do not have a single good operation concept of overvaluation. Thus there is considerable scope for studies to reach different conclusions. Back.

Note 69: Based on careful analysis on the problems in the Asian banking systems, Barth et al. (1998) and Schwartz (1998) also reach this conclusion. Back.

Note 70: See, for example Calvo (1996), Goldstein and Calvo (1996), Dooley (1997), Edison et al. (1998), and Kaminsky and Reinhart (1996). On the banking aspects of the Asian crisis, see also Barth et al. (1998) and Schwartz (1998). Makin (1998) puts the chief cause of the crisis as overinvestment. “Easy availability of funding in strong currencies has led Asians to increase the capacity to produce goods well above levels that would be economically viable” (p. 3). Lewis (1998:2) advances a similar argument. “The reason for the East Asian economic collapse was surprisingly simple; large Asian companies were failing to make money...They were selling hundreds of billions of dollar of consumer goods abroad without making a dime in profits. It wasn’t a business, it was a habit, and they gambled [on foreign currency borrowings] to support it” (p. 2). To the extent that domestic imbalances were generated more by excessive investment rather than excessive demand, it seems likely that it will take longer for economies to recover. On the problem of Asian overproduction, see also Lewis (1998). Back.

Note 71: Official announcements of IMF agreements and policy reforms did not always lead to currency appreciation, but this was typically because news of the likelihood of these events had already been widely circulated and hence already been discounted by the markets. It should also be noted that while the adoption of tough policy measure should strengthen the currency, they shouldn’t always be expected to strengthen the stock market. Thus, for example, the title of an article from the Financial Times (November 25, 1997) “Seoul Stock Market Plunges 7.22: Shares Hit 10-Year Low on Fears of Tough IMF Package.” Back.

Note 72: “Since a new government took over in November, Thailand has closed 56 finance companies, replaced the directors of two banks, and ordered banks to recapitalize—while shooting down any moratorium talk that could dismay investors and lenders...” (Wall Street Journal, January 9, 1998, p. A17). Back.

Note 73: See also Garrett and Lange (1996). Back.

Note 74: Krugman (1998b) notes that this attack on Soros was ironic, as it was widely believed in the market that he had not attacked the ringgit. Back.

Note 75: Blaming the speculators is of course hardly a practice limited only to developing countries. During the franc crisis a senior French official is reported to have lamented that speculators could no longer be sent to the guillotine. Back.

Note 76: This would have been twice or more as strong as the market rate at the time. Back.

Note 77: (See also the Financial Times, February 12, 13 and 14-15, 1998 and the Wall Street Journal, February 10 and 12, 1998). Back.

Note 78: See the analysis and references in Sweeney, Wihlborg and Willett (1998) and Willett (1998). Back.

Note 79: On the Czech case see the analysis and references in Burdekin, Nelson, and Willett (1998). Back.

Note 80: See Wihlborg and Willett (1991) and Willett and Mullen (1982). Back.

Note 81: Optimum currency area theory focuses on the factors that influence the relative costs and benefits of adopting fixed versus flexible exchange rates. Small size, a high degree of openness, factor mobility, and wage and price flexibility are major considerations favoring fixed exchange rates. For recent surveys of the literature see DeGrauwe (1992) and Tavlas (1993). For discussion of the relationships between OCA theory and the case for using fixed exchange rates to promote financial discipline see Westbook and Willett (1998). Back.

Note 82: For discussions of the revisions in IMF practices in the wake of the Mexican crisis, see Eichengreen and Portes (1997) and Fischer (1997). Back.

Note 83: Also important is the need to devise ways to reduce the extent to which perceptions of national government and IMF bailouts lead private sector actors to inappropriately evaluate the full risks of investing and borrowing. Such moral hazard problems are by no means the only reason for the myopia displayed by the international financial markets, but they are an important one.

The appropriateness of the IMF policy advice also has been a topic of considerable controversy. A typical critical comment is given by Makin (1998:2): “A hastily arranged and misconceived IMF program created a panic over South Korea and Asia by mid-December and forced U.S. Treasury Secretary Robert Rubin to take over the management of the Asian crisis outside Japan.” Writers of the left have frequently charged the IMF with promoting an excessively monetarist austerity policy which gives little attention to the plight of the poor, and The Wall Street Journal attacks the IMF for believing that devaluations work. Indeed, in an editorial on April 15, 1998, on “The IMF Crisis,” they charge that the IMF was the cause of the Asian crisis because of their urging of Thailand to devalue. While the composition of most IMF packages generally conforms to mainstream economic thinking, Keynesian and supply-side economists have also frequently argued that the IMF tries to apply too much austerity and discipline. Prominent economists such as Jeffrey Sachs and Joseph Stiglitz, Chief Economist at the World Bank, have been particularly critical of IMF policy conditions in the Asian crisis, because they argue the IMF has failed to understand that these are primarily due to financial market bad debt problems rather than overly expansionary macroeconomic policies. Not surprisingly the IMF disagrees with these charges, but it has displayed a commendable willingness to adjust its policies as the Asian crisis has developed. For recent discussion, see the references cited in footnote 3. For discussion of ways to reduce the moral hazard problems generated by IMF programs generated by IMF programs, see Goldstein (1998). Back.

Note 84: For recent discussion of these issues, see the contributions and references in Calvo, Goldstein and Hochreiter (1996), Glick (1998), Federal Reserve Bank of Kansas City (1997), Edwards (1995a) and Ries and Sweeney (1997). Back.