Economic Policy and Exchange Rate Regimes: What Have We Learned in the Ten Years since Black Wednesday?
by Edwin M. Truman, Peterson Institute for International Economics
Speech at the European Monetary Symposium
London School of Economics
September 16, 2002
© Peterson Institute for International Economics
Was the September 1992 crisis in Exchange Rate Mechanism (ERM) of the European Monetary System (EMS) inevitable? Was the logic of the ERM fatally flawed, perhaps, because it was out of step with the evolution of the international financial system? Were the ERM's problems a decade ago a function of the regime itself or were they caused by national economic policy choices? What role did German unification, the weakness of the dollar and the global economy, and choice of sterling's ERM peg play in the drama? To what extent are answers to these historical questions relevant to debates about international financial architecture or the desirability and sustainability of the common currency regime under the European Monetary Union (EMU)?
My assigned perspective is from the other side of the pond. I was an observer of and marginal participant in the events of the ERM crisis as it unfolded in 1992 and its 1993 exclamation point (which should be treated as an integral part of the story) as well as too many other foreign exchange and financial crises in the decade before and in the decade after that turbulent period. From that viewpoint, I draw four principal, overall lessons about economic policy and exchange rate regimes.
First, regimes at the hard end of the bipolar array, involving pegged exchange rates, currency boards, adoption of another country's currency, or a common currency, are more demanding than more flexible rate regimes. However, no regime offers a complete solution to the economic and financial challenges that an economy may face; without adequate supporting macroeconomic and microeconomic (institutional or structural) policies, any regime periodically will be subjected to pressures that may lead to crises. The virulence of the crisis depends on the political economy of the country at the moment, its economic circumstances, and the scope its authorities have to respond to exchange market pressures.
Second, the ERM crisis, not the Mexican crisis that followed in 1994-95, was the first external financial crisis of the Twenty-First Century not only in terms of the scale of the associated private and official capital movements both before and during the crisis but also in terms of the roles played by market securities, derivative instruments, and financial market dynamics in the buildup to and the unfolding of the crisis. The ERM crisis of 1992-93 should be viewed in the context not of the events of the 1970s and 1980s but of the events over the following ten years. A failure to appreciate this fact has set back subsequent efforts at crisis prevention and crisis management. With respect to crisis prevention, in the wake of the ERM crisis, the European view of exchange rate regimes continued strongly, if defensively, to endorse the virtues of pegged exchange rates. With respect to crisis management, the ERM crisis contributed to a European double standard: in their crisis, countries and investors were bailed out with unprecedented amounts of official financing, but European authorities applied a different test and were much stingier about official financing when it came to similar crises in the rest of the world in subsequent years.
Third, markets are not always right, but they are also not always wrong. Policy makers tend to praise the behavior of markets when it appears to be providing a vote of confidence in their policies while disparaging them when the message is one of no confidence. This tension is reflected in the ongoing debate about the relative roles of economic fundamentals and the behavior of markets in such crises.
Fourth, an exchange rate crisis could happen again in Europe. I am not predicting another crisis like that of 1992-93, in part, because history rarely repeats itself with precision. However, I attach a small probability to a crisis in Euroland's experiment with a common currency and a larger probability to an echo of the ERM crisis of a decade ago during the prospective enlargement of the European Union (EU) and the European Monetary Union.
In section I of this paper I summarize my US perspective on the events of a decade ago and the immediate aftermath, including the 1993 Group of Ten (G-10) report on the subject International Capital Movements and Foreign Exchange Markets and the important aftershock in July-August 1993. In section II, I discuss what we have learned about the interaction of economic and financial policies and exchange rate regimes in the decade since September 16, 1992.
I. A US Perspective on the ERM Crisis of 1992-93
The ERM crisis that reached its first climax in September 1992 was not entirely unanticipated in official circles in the United States. Its potential virulence was underestimated, as is the case with most actual international financial crises; on the other hand, the virulence of potential crises that do not occur is overestimated. Thus, William McDonough, commenting on the aftermath of the Danish rejection of the Maastricht Treaty on June 2, told the Federal Open Market Committee (FOMC) on June 30, "European central bankers think that they can hold the present rate structure together, but there are clear signs of strain."(Board of Governors of the Federal Reserve System 1992a, William J. McDonough, "Notes for FOMC Meeting," 5) In McDonough's report on "Treasury and Federal Reserve Foreign Exchange Operations: May through July 1992," which was released on September 10, 1992, he commented on the shift in investor sentiment following June 2, "Market participants viewed the Danish rejection as a blow to the prospects for a single currency in the foreseeable future. In their view, abandonment of the agreed-upon timetable for monetary union would loosen the tight discipline that the Maastricht treaty had implied for European inflation rates and budget deficits and thus raise doubts about the likelihood of continued convergence of European fiscal and monetary policies." He described the flow of funds out of higher-yielding European currencies into investments in German marks, a flow contributing to "talk of an imminent EMS realignment." (McDonough 1992, 739)
Notwithstanding this evidence that the US official sector was not entirely asleep at the switch, US official thinking during this period was not focused primarily on the ERM and the deutsche mark (DM). It certainly was not focused on sterling. The principal focus was on the US economy, whose recovery from the 1990-91 recession had been weaker than expected, and the implications of economic and financial developments for the impending US presidential election.
The Bush administration had already established a track record favoring heavy foreign exchange market intervention, for example having operated on 97 separate days in 1989. It was concerned about weakness in the dollar, at least with respect to European currencies,2 and the associated implications about the strength and leadership of the United States. At the same time, the US administration was eager for the Federal Reserve to stimulate the economy. Secretary Brady had been publicly critical of the Federal Reserve for failing to lower the federal funds rate far enough or fast enough.
The Federal Reserve reduced the federal funds rate by a further 50 basis points to 3.25 percent on July 2 after the release of weak data on US employment. The communiqué from the Munich Summit disappointed market participants because of the absence of any references to exchange rates. Secretary Brady's subsequent statement that the United States "is not seeking to depreciate the dollar" did not dispel the impression of G-7 unconcern. On July 16, the Bundesbank surprised the market by the size of the increase in its discount rate, three quarters of a percentage point to a record high of 8 percent. Against this background and the broader political economy considerations outlined above, the US monetary authorities operated in concert with a number of foreign central banks to buy dollars and sell DM on Monday, July 20. US participation was principally motivated by a desire to reinforce Secretary Brady's point about US policy on the dollar, not one of benign neglect. The Bundesbank was reluctant to participate in this operation, but responded positively to a direct invitation from the Federal Reserve (Board of Governors of the Federal Reserve System 1992b, William J. McDonough, "Notes for FOMC Meeting," 3). In large part because it was unexpected and, therefore, caught market participants with exposed intra-day positions, the operation was successful in appreciating and stabilizing the dollar-DM rate for the balance of the month at about 1.50 DM/dollar.
The dollar-DM rate came under renewed downward pressure on Friday, August 7, following another release of weak data on US employment, and the US monetary authorities again sold DM. The following Tuesday, August 11, they operated again, this time in concert with other central banks. However, those efforts "did not interrupt the tendency of the dollar to decline." (McDonough 1993, 12) In fact, the Federal Reserve dissuaded the Treasury from a further follow-up operation on August 13 (Board of Governors of the Federal Reserve System 1992b, William J. McDonough, "Notes for FOMC Meeting," 3). The Bundesbank did engage in "oral intervention" on August 14, when Hans Tietmeyer commented "we certainly are not interested in a weak dollar." (Board of Governors of the Federal Reserve System 1992b, William J. McDonough, "Notes for FOMC Meeting," 7)
The US Treasury continued to want to demonstrate US concern about the dollar's weakness, and the Federal Reserve reluctantly joined the Treasury in coordinated sales of DM on August 21 and August 24. However, the operations had little impact; the dollar-DM rate weakened from 1.4413 on August 20 to 1.3995 on August 25. The Treasury was finally persuaded that further US operations in the face of obvious strains on exchange rates within Europe would be ineffective and possibly counterproductive in terms of adverse spillover effects in US bond and equity markets. US monetary authorities did not participate in the heavy intra-European intervention operations at the end of August when sterling came under heavy pressure and the lira dipped below its ERM floor. They did not operate again in the foreign exchange markets for the balance of 1992.
The dollar subsequently hit an historic low against the DM of 1.3862 on September 2 but recovered a bit on September 3. The dollar dipped on September 4 following a third successive release of disappointing data on US employment and the Federal Reserve's final reduction in the federal funds rate to 3 percent, but the DM rate did not breach the historic low. Through the balance of the September turbulence affecting European exchange rates, the dollar strengthened, reaching high for the month of DM1.5180 on Black Wednesday, September 16, perhaps reflecting in part safe-haven considerations.
In the immediate aftermath of the September developments affecting intra-European exchange rates, the view at the Federal Reserve, for example, during the October 6 meeting of the FOMC, was that the Italian lira's devaluation within the ERM that was agreed on September 13 and the Bundesbank's tepid interest rate response were inadequate to deal with mounting pressures; the remaining investors in currencies with high interest rates "went roaring toward the same exit and ordered their commercial or investment banks to get them out of their foreign exchange positions immediately at virtually any exchange rate." (Board of Governors of the Federal Reserve System 1992c, William J. McDonough, "Notes for FOMC Meeting," 5) The FOMC meeting was three weeks after Black Wednesday, which may be one reason why sterling's departure from the ERM received only passing notice in McDonough's report, "After very heavy losses of reserves, the British had to pull out of the exchange rate mechanism on September 16, followed by the Italians." (Board of Governors of the Federal Reserve System 1992c, William J. McDonough, "Notes for FOMC Meeting," 4)
As events unfolded in August and September, staff at the Federal Reserve watched the pressures build in the ERM with a combination of disbelief and awe. By September 16 and the suspension of sterling's participation in the ERM, we were already quite numb. When we got to the FOMC meeting on October 6, so much more had happened, including the UK announcement on September 19 that it would not return sterling to the ERM until the mechanism was reformed, that sterling's trials and tribulations were seen as relatively minor. I did tell the FOMC that the staff judgment was that the "slight easing of European monetary conditions," having the UK situation very much in mind, should boost European growth unless "continued turmoil within Europe" generated substantial uncertainty that, in turn, worked to reduce investment in Europe. (Board of Governors of the Federal Reserve System 1992c, E. M. Truman, "FOMC Presentation-International Developments," 1). It turned out that we were right in our basic forecast for the United Kingdom, where there was a recovery in 1993, and also right to be worried about the rest of the European Union, where on average economic activity contracted in 1993.
When the dust had settled as of early October, it was apparent "[t]he Bundesbank achieved its goal of restoring flexibility to the EMS and lessening the likelihood that they would lose control of monetary policy as they did in September." (Board of Governors of the Federal Reserve System 1992c, William J. McDonough, "Notes for FOMC Meeting," 7) Although this comment could be interpreted as referring to the easing action that the Bundesbank was persuaded to take on September 14 following the lira's depreciation, a more complex view is that the Bundesbank had been able to invoke the 1978 Emminger letter and limit the extent of its intervention which it saw as threatening monetary control. In fact, the Bundesbank fairly easily sterilized its unprecedented intervention and financing operations during this period; it was the other European central banks, not the Bundesbank, which adjusted, i.e., tightened, their monetary policies, in effect conducting intervention that was not sterilized.
The view at the US Treasury differed somewhat from that at the Federal Reserve. Secretary Brady stated in a letter to his G-10 colleagues on October 1, "Recent events demonstrate forcefully that the increased size and complexity of international financial markets over the past few years have resulted in a volume of transactions which dwarfs the resources governments can bring to bear in the markets." In his capacity as Chairman of the G-10 Ministers and Governors, he proposed that the G-10 Deputies cooperate with the International Monetary Fund (IMF) on a "study of international capital markets and the implications of recent trends and developments." (Letter reprinted in Group of Ten 1993) The G-10 acted on Secretary Brady's proposal and produced a solid and, what is more notable, publicly released study of the ERM crisis.3
The report on the G-10 study reached three principal conclusions that remain valid today:
First, "However welcome on broad economic grounds, the growth in the size, integration, and agility of international financial markets has greatly increased the scale of the pressure that can be exerted against an exchange rate when market sentiment shifts." (Group of Ten, paragraph 74)
Second, with respect to the effectiveness of traditional exchange rate policies and intervention strategies, "the effectiveness of such policies and strategies depends crucially on the circumstances in which they are employed." Factors that were cited included the nature of the exchange rate system, the role of interest rates, and coordination and cooperation on policies (Group of Ten, paragraph 75). Furthermore, the effectiveness of exchange market intervention "under any type of exchange rate arrangement" depends on "whether market participants perceive that existing exchange rates are reasonably consistent with macroeconomic fundamentals, and whether the commitment of policy authorities to limit exchange rate movements is credible." (Group of Ten, paragraph 76)
Third, for a pegged exchange rate system involving strong commitments to avoiding realignments, "Convergent macroeconomic policies and performances are necessary to avoid exchange rate adjustments on a more durable basis. . . . Timely adjustment [of parities] may involve lower economic costs and a smaller erosion of political credibility than attempting to resist adjustment for as long as possible. . . . The policy implication, accordingly, is that the degree of resistance to exchange rate adjustment should be kept consistent with the degree of convergence in macroeconomic policies and performances." (Group of Ten, paragraph 79)
From a US perspective, it also was important that the G-10 report observed, accurately in our view, "the weakness of the US recovery, combined with reductions in US interest rates and the associated depreciation of the dollar through early September, as well as the low growth of activity in most other OECD countries, generated a less favorable environment for the maintenance of exchange rate stability within the ERM. However, the major causes of the recent exchange rate tensions were to be found in developments within ERM countries themselves." (Group of Ten, paragraph 41)
The G-10 report did a nice job in describing the market phenomena that were revealed by the ERM crisis. In the end, it also reached reasonably balanced conclusions about the potential role for exchange market intervention in dealing with financial market pressures.4 Where it fell short, in my view at the time and certainly as I now look back on that period, was in its emphasis on the behavior of financial markets and the pressures that they can bring to bear on the authorities and in its lack of emphasis on economic fundamentals, including the political implications of economic policies and performances.
This failure was partly a consequence of the way that Secretary Brady posed the issues to be addressed by the Group of Ten. It was also a reflection of the view of the representatives of some G-10 countries that the crisis itself demonstrated a market failure and that it was important that the report on the crisis not convey the impression that governments are powerless in the face of market irrationality. Recall French Finance Minister Sapin's comment that during the French Revolution the authorities knew how to deal with speculators, they used the guillotine.
The Research Department of the IMF, in its contribution to the work of the G-10 Deputies, put forward a more balanced assessment, "The turbulent events in foreign exchange markets in the late summer and autumn of 1992 demonstrated the enormous pressures that can be brought to bear against official exchange rate parities when market participants perceive that significant exchange rate realignments may be imminent [following the June 2 defeat of the Maastricht referendum in Denmark] . . . In a fundamental sense, however, the causes of the recent exchange rate turbulence were not the market pressures that were unleashed late last summer , but rather the underlying macroeconomic divergences that developed during the preceding few years and ultimately provided the foundation for these market pressures." (International Monetary Fund 1993a, 141)
In the aftermath of Chapter Two of the ERM crisis in the summer of 1993, this emphasis on fundamentals was detailed further in the World Economic Outlook produced in October 1993 (International Monetary Fund 1993b). In that publication, the IMF staff pointed to external factors that had contributed to the ERM crisis in 1992, the deterioration in cost competitiveness and widening of current account deficits. They noted that the external dimension added to the conflict between the domestic requirements of monetary policy to address lackluster economic activity and rising unemployment in an environment of limited inflation pressures and, in some cases, strained banking systems and the external requirements of monetary policy that were largely dictated by the efforts by the Germany monetary authorities to deal with the Germany's sub-optimal economic policy choices in the context of German monetary union-the one-to-one conversion of Ostmark into deutsche mark, the ensuing economic boom driven in part by an expansionary fiscal policy, and no upward adjustment of the DM within the ERM.
Chapter two of the ERM crisis served to reinforce the view that fundamentals are more important than market phenomena. Following the extraordinary and in some cases, for Belgium, Denmark, France and the Netherlands, superficially successful efforts to preserve their ERM parities in late 1992 and through the turbulent first seven months of 1993, the exchange rate band in the arrangement was widened to plus or minus 15 percent. Ireland in January finally chose to devalue the pound within the ERM as unemployment rose. Rising unemployment in Spain and Portugal led to further devaluations of those currencies in May of 1993. Unemployment was also rising in Denmark and France contributing to "perceived tension between the domestic needs of monetary policy and the ability to endure high interest rates to defend official parities." (International Monetary Fund 1993b, 32) Moreover, "[I]n economic conditions that had worsened considerably over the preceding year, it would be far more difficult to raise interest rates substantially and durably to combat renewed exchange market pressures." (International Monetary Fund 1993b, 33) The German economy was finally cooling down, the Bundesbank had embarked on a course of reducing interest rates, but the reductions were considered inadequate to address in a timely manner deteriorating economic conditions in some of Germany's partners.
The EU finance ministers and central bank governors on August 1, 1993, in effect, threw in the towel on the ERM as a hard-peg regime. In fairness, in announcing the wider ERM bands, the ministers and governors did reaffirm existing central parities, which helped to provide a continuing policy and market focus. Less realistically they asserted their intention of returning to a system with narrower bands as soon as circumstances permitted. It is interesting to note that a decade later, circumstances have not yet permitted a generalized return to narrower bands as far as actual and potential participants in ERM-II are concerned. It is also interesting to recall that contrary to the behavior of the dollar in Chapter One of the ERM crisis of 1992-93, when the dollar weakened in advance of the climax in September and strengthened in its aftermath, in Chapter Two the dollar strengthened in advance of the July-August climax and subsequently weakened. It would seem that events in the ERM were not the only force affecting the evolution of the dollar's value on average in either period.
II. What Have We Learned About Economic Policy and Exchange Rate Regimes?
The first overall lesson about economic policy and exchange rate regimes that we have learned from the ERM-crisis and similar crises since 1992 has two parts: (1) pegged exchange rate regimes are very demanding and (2) no regime offers a complete solution to the economic and financial challenges that a country is likely to face. That pegged exchange rate regimes are very demanding is the lesson of the ERM crisis of 1992-93, as well as the Mexican crisis of 1994-95, the Asian financial crises of 1997-98, the Russian crisis of 1998, the Brazilian crisis of 1998-99, the Turkish crisis of 2000-01, and the Argentine crisis of 2001-02. It is the lesson of all the major international financial crises of the past ten years with the exception of the current Brazilian crisis; Brazil has an exchange rate regime that is flexible both de facto and de jure.
The Brazilian crisis of 2002 reminds us of the second part of this lesson: no exchange rate regime, by itself, can reasonably promise a complete solution to all of a country's potential economic and financial challenges. From the fact that no exchange rate regime is best for all countries, at all times, or in all circumstances, it follows that any exchange rate regime must be supported by macroeconomic and microeconomic (institutional or structural) policies that are consistent with and compatible with that regime as well as with the political economy and economic circumstances of the country at the time.
In the case of Brazil today, a flexible exchange rate may not be enough to avoid an external financial crisis. Brazil's chances of avoiding a full-fledged crisis would have been enhanced and will be enhanced if it were able to follow an even more disciplined fiscal policy, one that would put the ratio of its public sector debt to GDP on a downward path that would be more resilient to disturbances such as an Argentine devaluation or the uncertainties of a presidential election. Brazil's chances of avoiding a full-fledged crisis also would have been enhanced and will be enhanced if its underlying economy were more competitive. If the economy were more open, if there were less explicit and implicit reliance on governmental intervention in the economy, and if the rate of private investment were higher, which, in turn, would be aided by less government debt and a lower overall fiscal deficit. With a more flexible and competitive economy, the ratio of Brazil's external debt to its exports of goods and services or to GDP would be on a downward path that would be more resilient to disturbances.
A similar story can be told about Argentina. It is not that its choice of an exchange rate regime was in error. At the time the choice was made in 1991, it was brilliant; witness the initial success despite the skepticism of many, including myself, at the time.5 Although Argentina might have been better served if the authorities had chosen to exit from its exchange rate regime sooner, Argentina's problems over the past three years were not primarily caused by its choice of an exchange rate regime. Those problems were caused by insufficiently supportive macroeconomic and microeconomic policies. The economy lacked the flexibility and competitiveness to adjust to the real appreciation of the peso as the US dollar appreciated and when the Brazilian real depreciated. As has been well documented by Michael Mussa (2002), fiscal policy was too loose in the boom period of 1997-99, and the program of the de la Rua government that came to power in early 2000 was never fully implemented. The result was protracted adjustment through recession and deflation that undermined political support for sound policies and exacerbated the trend in the ratio of public sector debt to GDP. The numerator grew while the denominator shrank. In the end, the political economy could not sustain the effort.
The Mexican and East Asian crises were similar in that in the context of fixed exchange rate regimes macroeconomic policy was insufficiently restrained to prevent the emergence of large external deficits (Mexico and Thailand) or the build up of highly leveraged balance sheets in part through external borrowing (Korea and Indonesia). Moreover, by the time the problems were evident, banking systems were insufficiently robust to support the tight monetary policies that would have been essential to avoid an exchange rate crisis.
The 1992-93 ERM crisis was similar to the crises that followed in that macroeconomic and microeconomic policies proved to be insufficiently supportive of the exchange rate rigidity that built up within the ERM and associated European exchange rate arrangements since 1987. The ERM had not undergone a general realignment since 1987.6 As a consequence, the ERM had evolved into a very rigid regime, and that rigidity was reinforced by the Maastricht treaty's emphasis on exchange rate stability as a condition for participation in stage three of EMU. Neither the authorities nor their bodies politic had fully embraced the implications of that rigidity for economic policies and performance. A disconnect resulted between the choice of the exchange rate regime and the economic policies necessary to support that choice along with the potential adverse implications for economic performance.
The second overall lesson we have learned about economic policy and exchange rate regimes in the decade since the 1992-93 ERM crisis is that the ERM crisis was the first external financial crisis of the Twenty-First Century not only in terms of the scale of the associated private and official capital movements both before and during the crisis but also in terms of the roles played by market securities, derivative instruments, and financial market dynamics in the buildup to and the unfolding of the crisis. This lesson was largely underappreciated at the time, and, I would submit, it is under appreciated today to the detriment of the international financial system.
Consider the basic elements of the ERM crisis; not all apply to every country that was caught up in the crisis, but consider them as a whole. First is the fixed exchange rate regime. The regime constrained national monetary policies. It also contributed to excessive borrowing abroad at "cheaper" interest rates by domestic residents and to excessive cross-border investments seeking higher yields; both borrowers and lenders ignored potential exchange rate risk. Given that most of the European countries ultimately abandoned the rigidly fixed exchange rate regime, through the adoption of floating or the 15 percent band, the ex post critics argued that the regime should have been abandoned or modified earlier in order to avoid the damage to the real economy-the absence of an exit strategy.7
Second, consider macroeconomic policies. Fiscal policy, particularly in Germany, was faulted as one of the principal causes of the crisis. Fiscal policies were tightened both during and/or in the wake of the crisis, for example, in Italy, Sweden, Finland, Portugal and even Germany. Monetary policies were tightened by all of the central banks whose currencies came under downward pressure, and in several cases (Sweden, Spain, Portugal and Norway, for example) monetary policies were employed to limit subsequent currency depreciation.
Third, banking system problems were common and in some cases severe, especially in the Scandinavian countries. This situation limited the scope for the use of monetary policy to defend exchange rates.
Fourth, three countries experimented unsuccessfully with exchange controls. Spain imposed them, and Ireland and Portugal moved to enforce controls that were already on the books.8
Fifth, political uncertainty played a major role, for example, in Denmark, France, Italy, and Spain. As often in such crises, in the end politics-lack of political support for policies judged to be necessary to ride out the storm and dissatisfaction with economic performance-trumped economics.
Sixth, contagion was rampant in the ERM crisis. When the Finnish markka was floated, it put pressure on the Swedish krona. When the lira was devalued within the ERM, it put pressure on sterling. When sterling was floated, it put pressure on the Irish pound. The European contagion involved competitive pressures, proxy hedging, lack of differentiation among countries, and increased aversion to risk.
Seventh, the ERM crisis was characterized by exchange market intervention and official assistance on an unprecedented scale, net sales of deutsche mark by European central banks between June and December 1992 were estimated at DM284 billion, approximately $200 billion. A portion of those sales came from owned reserves, but an even larger portion reflected the extension of bilateral and multilateral credit. Mutual support eventually included an $11.2 billion EU loan to Italy and an array of new loan programs by the European Investment Bank. Although the official financing during and following the crisis did not involve the Bretton Woods institutions, it was an unprecedented volume of international financial assistance. That assistance should be recognized for what it was: a collaborative effort to deal with what is conventionally called a "liquidity crisis." In the absence of that support, ERM parities would have been abandoned under even more chaotic conditions. Moreover, many investors were bailed out by the liquidity that was provided by central banks and governments.
Eighth, without the unprecedented assistance the economic and financial consequences of the ERM crisis almost certainly would have been even more severe in many of the countries caught up in the crisis. Many regard the ERM crisis as a benign event in terms of its economic consequences, but they are mistaken. Their memories may be distorted by the UK experience where real GDP rose 2.2 percent in 1993 following a 0.5 percent contraction the previous year.9 For the European Union as a whole, real GDP declined 0.4 percent in 1993, following modest 1.0 percent growth in 1992, and real total domestic demand declined 1.6 percent following 1.1 percent growth the year before. Finland and Sweden, which were not yet members of the EU, both experienced in 1993 their third consecutive years of negative growth.
Ninth, the ERM crisis was not caused by one event, such as Denmark's initial rejection of the Maastricht treaty. The IMF's Research Department (International Monetary Fund 1993a, 144) suggested at the end of Chapter One of the ERM crisis that the best analogy is with an "an earthquake, where the underlying pressures are built up over a period of time and then suddenly unleashed by some triggering event in a huge shock, followed by a series of after shocks." They went on to comment that little is to be gained from trying to avoid particular triggering events; "Instead, attention might more profitably focus on understanding the economic pressures underlying the recent European exchange market crisis, and on designing mechanisms and strategies to avoid a build up of such pressures in the future." In other words, improved crisis prevention. The ERM crisis involved many elements of the subsequent debate over crisis management as well. However, little was done to generalize those lessons.
In retrospect, it has been unfortunate that the ERM crisis was not and has not been viewed as a precursor of the international financial crises of the following ten years, unfortunate that governments and institutions did not start earlier to learn lessons for crisis prevention and management.
Two unlearned lessons deserve attention. With respect to crisis prevention, most European officials in the wake of the ERM crisis continued to preach the virtues of pegged exchange rate regimes, perhaps, because they felt the need to defend their decisions. The European preoccupation with the hard-pole answer to the exchange-rate-regime question helped to paralyze the IMF from seriously examining the viability of pegged exchange rate regimes in Mexico and East Asia, thereby, contributing to their crises. The European fixation with pegged rates also prevented the establishment of a G-7 consensus as part of the new international financial architecture in the wake of the East Asian crisis.
With respect to crisis management, in the ERM crisis, countries and investors were bailed out with official financing. Apparently, the well-known European aversion to large-scale packages of external financial support did not extend to intra-European operations. The explanation, no doubt, is that efforts to save the ERM were considered to be an internal matter. However, a similar explanation was rejected in 1997 when Asian countries proposed the establishment of their own monetary fund. (See Henning 2002.) In addition, many Europeans who were heavily implicated in the ERM bailout of 1992-93 opposed the rescue operation for Mexico in 1995, turned a deaf ear to pleas from East Asian countries for larger packages of external financial support in 1997-98, and today continue to press for firm limits on IMF lending programs, arguing that investors must be taught to bear the consequences of their decisions.10 Thus, the European double standard emerged on international rescue packages.
The third overall lesson for economic policy and exchange rate regimes from the ERM crisis and subsequent international financial crises is that markets are not always right, but they are also not always wrong. Much of the debate about financial crises over the past decade has revolved around the question of whether their causes lie in the economic and financial fundamentals or the behavior of financial markets. The G-10 debate about the ERM crisis was no different. The 1993 G-10 study was motivated, in large part, by a concern that the markets had misbehaved. The G-10 report largely absolved the markets, but it failed adequately to pin the blame on the fundamentals.
Nevertheless, consider the proposition that the markets, meaning market participants, were at fault in the ERM crisis. When did markets get it wrong? Were they wrong to bet on a continuation of the economic convergence process or were they wrong to hedge their bets after the failure of the Danish referendum? Posing the question this way demonstrates the propensity of policy makers to praise the behavior of markets when it appears to be providing a vote of confidence for their policies while disparaging the behavior of markets when the message is one of diminished or no confidence.
Again, the Asian financial crisis provides an interesting parallel with the ERM crisis in this respect. The view in much of Asia continues to be that the fundamentals of economic policy were not at fault in 1997-98; the crisis was caused by the irrational behavior of financial markets. As noted in connection with the second lesson, the resulting Asian prescription called for massive external financial assistance to roll back the speculative tide, but it was denied to them with the Europeans taking the lead.
The fourth overall lesson from the ERM crisis about economic policy and exchange rate regimes relates not so much to the past but to the future, to the future of exchange rate arrangements in Europe. What is the likelihood of another European foreign exchange crisis over, say, the next ten years? What is the probability of a crisis that would affect the common currency at the core of European Monetary Union? What is the probability of a crisis that would affect the currencies and economies of accession countries and thereby negatively impact upon the economy of Euroland itself?
The euro, the common currency of Euroland, is a unique type of hard-pole exchange rate regime, harder than was the ERM itself, but nevertheless a regime that could face a crisis, and a regime that could be dissolved. As with the ERM of 1992-93, the most likely source of constitutional crisis for the euro would be dissatisfaction with economic performance in a significant part of Euroland and association of the euro and the common monetary policy as the source of the disappointing performance. The crisis might be brought on by structural rigidities that inhibit the internal adjustment process, it might be caused by a monetary policy that focuses too much on inflation and not enough on deflation until it is too late, or it might be caused by a further buildup in public sector debt in one or more of the national economies and a rising risk of default.
Public opinion in the United Kingdom apparently attaches a higher probability to such an adverse development than does informed opinion outside the United Kingdom, in part, no doubt, because of memories of the traumatic and humiliating experience leading up to Black Wednesday but also, in part, because of a failure to recognize the central role that German monetary union and the associated German policy choices played in the ERM crisis. Germany remains a point of vulnerability for the euro experiment because, again, the common monetary policy is less than fully appropriate for Germany's economic circumstances, including the rigidity of its labor and product markets. However, Germany no longer has as much control over the levers of Euroland policy. Moreover, it is difficult to imagine an economic disturbance, internal to the euro area, on the scale of German unification. Therefore, I attach only a small probability to a crisis in Euroland's experiment with a common currency. I suspect the British are excessively preoccupied and insufficiently appreciative of the true lessons of the ERM crisis: the need for appropriate supporting economic and financial policies.
I attach a larger probability to an echo of the ERM crisis of a decade ago during the EU enlargement process as it proceeds in the period ahead. Markets and policy makers both now assume it will go smoothly. On the other hand, I suspect that excessive weight has been given to exchange rate stability as one of the criteria for joining both the EU and the common currency, and I would not rule out the possibility of a revision in market expectations to the consternation of policy makers. Consider one example, Hungary. (I do not single out Hungary because it is uniquely vulnerable; similar scenarios could be developed for other countries that are candidates for EU accession, and the situation is ripe for contagion among them.) In Hungary, the public sector debt is more than 50 percent of GDP, the fiscal deficit is about 6 percent with an objective of bringing it below 3 percent by 2004 or 2005, growth is 4 percent, consumer price inflation is 5.5 percent and is also to be reduced to below 3 percent by 2004 or 2005, the exchange rate is pegged to the euro with a wide band of plus or minus 15 percent, but the forint is now strong within that band, in part, because Hungary's short-term interest rate is 9.50 percent compared with 3.25 percent for the European Central Bank and the long-term (ten year) interest rate is about 7.50 percent compared with about 4.60 percent in Italy. Hungary is the recipient of substantial capital inflows, and as a result, Hungary is ripe for a convergence drama like that we saw in Europe in 1991-93.
What will happen if the progress toward enlargement slows down because present EU members get cold feet? What will happen if Hungarian enthusiasm for joining the EU and the euro area is dampened because the economic and financial terms of the deal become less attractive or economic prospects sour? What will happen if the political economy of Hungary does not support its ambitious program of fiscal retrenchment and monetary restraint? The result would be a mini European exchange rate crisis. A crisis could well spread to several other candidates for EU accession and impact on both the economic and financial stability of Euroland and the viability of the euro itself.
Crisis II for Euroland may be the subject for a symposium ten years from now especially if we have not learned the lessons of the past decade.
Bini Smaghi, Lorenzo and Giovanni Ferri. 2001. Was the Provision of Liquidity Support Asymmetric in the ERM? New Light on an Old Issue. Working Paper of the Ministry of the Treasury. Rome, Italy: Ministry of the Treasury (March).
Board of Governors of the Federal Reserve System. 1992a. Transcript of Meeting of the Federal Open Market Committee: June 30-July 1, 1992. Washington, DC: Board of Governors of the Federal Reserve System.
Board of Governors of the Federal Reserve System. 1992b. Transcript of Meeting of the Federal Open Market Committee: August 18, 1992. Washington, DC: Board of Governors of the Federal Reserve System.
Board of Governors of the Federal Reserve System. 1992c. Transcript of Meeting of the Federal Open Market Committee: October 6, 1992. Washington, DC: Board of Governors of the Federal Reserve System.
Group of Ten. 1993. International Capital Movements and Foreign Exchange Markets: Report to the Ministers and Governors by the Group of Deputies. Basel, Switzerland: Bank for International Settlements.
International Monetary Fund. 1993a. "A Note on Macroeconomic Causes of Recent Exchange Market Turbulence." Annex V in Group of Ten, International Capital Movements and Foreign Exchange Markets: Report to the Ministers and Governors by the Group of Deputies. Basel, Switzerland: Bank for International Settlements.
Henning, C. Randall. 2002. East Asian Financial Cooperation. Washington, DC: Institute for International Economics.
International Monetary Fund. 1993b. World Economic Outlook. Washington, DC: International Monetary Fund (October).
McDonough, William J. 1992. "Treasury and Federal Reserve Foreign Exchange Operations: May through July 1992." Federal Reserve Bulletin (October): 738-742.
McDonough, William J. 1993. "Treasury and Federal Reserve Foreign Exchange Operations: August through October, 1992." Federal Reserve Bulletin (January): 11-14.
Mussa, Michael. 2002. Argentina and the Fund: From Triumph to Tragedy. Washington, DC: Institute for International Economics.
1. I am grateful for the comments from my colleague John Williamson on a previous draft of this note. He and the Institute for International Economics, of course, are not responsible for any of the views or remaining errors the paper contains.
2. The U.S. posture toward the yen during 1992 was more ambivalent than that with respect to European currencies. As the yen weakened during the first quarter of 1992, the US Treasury joined the Japanese authorities in selling dollars in Tokyo on Monday, February 17 when the US markets were closed for a holiday and the rate was about 128 yen/dollar. The Federal Reserve declined to participate in this operation, because of skepticism about its likely effectiveness and technical concerns about operating in the Tokyo market. The Federal Reserve did join a follow-up operation in Tokyo on February 20, having made its point to the Treasury through its decision not to participate earlier in the week. Nevertheless, the dollar hit a high for the year of ¥134.97 on April 2. Later in the year, the dollar hit an historic low of ¥118.60 on September 30, and there were no sales of yen by the US authorities.
3. Secretary Brady proposal, however, gave rise to Truman's rule on international working groups: potential outgoing Treasury Secretaries should not propose international studies. The result in this case was that the incoming Clinton administration had little time or interest in the study, and the burden of quality control fell upon the Federal Reserve.
4. The balanced results were largely the consequence of negotiations at the Bank for International Settlements in Basel that lasted until 3:00 a.m.
5. Nevertheless, the Turkish experience has demonstrated that an exchange-rate based stabilization strategy may have had a reasonable chance of succeeding in 1991, but eight years later it was much more difficult to pull off even with, though some might argue because of, a pre-announced exit strategy.
6. The lira was devalued within the ERM in 1990, but that was viewed as a minor correction.
7. Finland floated the markka on September 8. At the time, Finland's action was criticized by its European partners as an unfriendly act (recall Taiwan's pre-emptive exchange rate move in October 1997), but in retrospect it was a well-timed move.
8. The Group of Ten (1993, paragraph 55) concluded that the controls were ineffective, counterproductive, and "damaging in economies with large financial sectors that have been operating in an environment free of such controls."
9. The growth of real total domestic demand was positive at 0.3 percent in 1992 and it increased to 2.2 percent in 1993. Thus, the swing in the external sector accounted for 0.8 percentage points (about 30 percent) of the pick up in real GDP.
10. Lorenzo Bini Smaghi and Giovanni Ferri (2001) have demonstrated that the provision of liquidity support within the ERM was not as symmetric as defenders of that system pretended was the case. It was not a "rules based" system; considerable discretion was exercised by the center country (Germany) with a tendency to favor assistance for countries closer to Germany, for smaller countries, and in situations in which there was less inflation pressure in Germany.