|
|
|
|
Selling Globalization. The Myth of the Global EconomyChapter 4. Currency Crises
Michael Veseth
We have been totally defeated by reality.
—Gabriel García Márquez1You don't often find Nobel Prize-winning novelists commenting on international economic news. Gabriel García Márquez, whose words are quoted in the above epigraph, received the 1982 Nobel Prize in Literature for his novels and short stories, which explore the dream qualities of reality and the reality of dreams. He's no economist, and yet he might understand the nature of global finance better than many of us who are trained in this discipline. We create dreams (theories) to describe a reality that he understands to be just a dream itself. Like a character in One Hundred Years of Solitude, we constantly mistake the imaginary for the real.
Was the peso crisis of December 1994, the object of Gabriel García Márquez's comments, reality or dream (or nightmare)? Is globalization—the process that tries to build solid economic, political, and social structures on the unsettled, crisis-plagued foundation of global finance—reality or dream?
Currency crises are no longer exceptional; they are now an accepted element of global finance. The Asian currency crisis of 1997, which infected Thailand, Indonesia, Malaysia, the Philippines, South Korea, and Japan with a contagious financial instability virus, is only the most recent in a long line of international financial plagues. Currency crises make true globalization impossible because the global integration of economic activity cannot take place in a world where currency values are subject to sudden jolts and long periods of instability, and where patterns of global capital movements are equally uncertain. You cannot build solid global businesses on so unsettled a foundation as this.
This chapter is built around short histories of two recent currency crises: Mexico in 1994 and Europe in 1992-1993. These real world case studies are wrapped around a theory core, where I survey the recent research on currency crises and speculative attacks. Both theory and experience persuade me that currency crises are a permanent part of global finance as we know it today and therefore a permanent constraint to the process of globalization.
The Peso Crisis of 1994-1995
The peso crisis of 1994-1995 illustrates the unstable nature of international financial flows today. Although the collapse itself was sudden, the crisis that led to it developed slowly over several months. News articles from this period indicate that the possibility of crisis was always very clear to market participants, who knowingly entered into arrangements that ultimately collapsed.
Perhaps the peso crisis is so discouraging because there is so little mystery to it. Mexico was the center of the media world for much of 1994. From the advent of the North American Free Trade Agreement (NAFTA) on January 1 through the controversial presidential election in November, all the important developments were covered in great depth by the press. The facts of the case are as follows.
For Mexico, 1994 was a roller-coaster year. It began with a combination of triumph and turmoil. The triumph came on the first of January as Mexico officially entered NAFTA. Mexico looked forward to the economic benefits of free trade with the United States and Canada, which, it was thought, would also help consolidate a liberal political regime in Mexico. Capitalism and democracy were finally on the horizon. But the turmoil began on the very same day in Chiapas, a poor area in Mexico's south, where revolutionaries had armed and organized themselves, seizing control of a region in protest of NAFTA and the free-market policies of the president, Carlos Salinas de Gortari. The year ended with the currency crises that are the focus of this account, which erupted shortly after Ernesto Zedillo Ponce de León took office as president of the Republic of Mexico on December 1, 1994.
Mexico had a persistent current account deficit of about $20 billion per year that was seen by many as its most pressing short-term economic problem. NAFTA-driven optimism spurred imports of both consumer goods and capital equipment. These imports were financed by hot money—short-term portfolio capital that was typically invested in peso-denominated government obligations called Cetes.
In March 1994, Donaldo Colosio was assassinated. Colosio was the presidential nominee of the dominant Institutional Revolutionary Party (PRI) and therefore the effective designated successor to Salinas. The assassination shook Mexico by its political roots, with significant economic fallout. The Colosio assassination raised the political risk premium required to attract and hold hot money in the Cetes. The inflow of foreign capital dried up, putting pressure on the central bank to raise interest rates.
The problem of attracting short-term foreign capital was made even worse by rumors that came and went that Mexico might soon be forced to devalue the peso. Peso devaluation would obviously reduce the dollar value of Mexican assets for short-term investors and would unleash inflationary forces and higher interest rates that would also affect long-term investments. The Mexican government pledged repeatedly that they would not resort to devaluation. Their policy was to keep the peso-dollar exchange rate within a slowly expanding band around a central rate of about 3.2 pesos per dollar.2
The question was how to attract this foreign capital without raising interest rates in the face of rising rates in the United States, devaluation fears, and political uncertainty. The solution appeared in the form of dollar-indexed short-term government bonds called Tesobonos. The Tesobonos were insurance against devaluation, both because of their guaranteed dollar value and by simple logic: No rational government would agree to make payment in dollars and then devalue relative to the dollar. Such an action would massively increase the debt. The Tesobonos were therefore a symbol of the Mexican government's commitment to exchange rate stability. After March, the Cetes were converted into the popular Tesobonos at a high rate. By December, the Tesobonos debt totaled about $28 billion. Mexico's government was solvent—it had sufficient assets to pay its Tesobonos debt—but it was illiquid. It had only about $12 billion in ready reserves to cover the hard currency base of the Tesobonos. This was a sustainable situation, however, as long as the hot money flows were not diverted and the short-term Tesobonos were rolled over into new securities by their foreign holders. In other words, the reserves were adequate providing there was no run on Tesobonos that might test the reserve limits.3
In fact, devaluation to reduce the current account deficit seemed decidedly unlikely in March, when the Tesobonos policy was initiated. Devaluation would have imposed significant political costs on the governing PRI, which could not help but view all actions in terms of their likely impact on the presidential election. Besides, devaluation would probably have resulted in rapid inflation, which would quickly negate any competitive benefits. Higher price levels would also have pushed up interests rates. The real and symbolic impacts of a devalued peso would have divided the party that Zedillio, Colosio's replacement as presidential candidate, was trying to lead into the election.
The political and economic environment worsened in November 1994, establishing the dynamics of the panic that followed. Mexico's Attorney General Ruiz Massieu revealed that the government had blocked the investigation of his brother's assassination in September. Massieu resigned from office amid rumors about plots within the PRI that might affect the political stability of the nation. Capital flight from Mexico began in serious volume as hot money investors, fearing a serious breakdown in order and governance, sought sanctuary in dollar assets.
Just when Mexico's domestic political situation seemed to be bleakest, Mexico's economic problem also worsened owing to an external event with severe internal implications. The U.S. Federal Reserve Board raised interest rates in an effort to keep inflation from gaining momentum in the United States. The higher U.S. interest rates made it even harder for Mexico to attract the funds it needed to finance its international payments deficit. Mexico came under even greater pressure to increase domestic interest rates to stem the accelerating capital flight. Mexico's authorities were extremely reluctant to raise their interest rates because of the domestic economic and political risks that this action entailed. The political risks were obvious: Raising interest rates so close to the election risked further alienating voters and weakening support for the PRI and its candidates. No incumbent party wants to go into the final weeks of an election with higher interest rates.
However, the economic risks, if anything, were greater. Mexico's newly privatized banking system was fragile and in danger of serious trouble if higher rates caused loan repayment problems. All they needed was a banking crisis! However, Mexico's banking system was also in great jeopardy for another reason. Banamex and other banks were borrowing dollars at the lower dollar interest rate and using funds to make peso loans at higher interest rates—a practice that was sustainable only as long as the peso was kept stable, which the PRI promised to do. This excursion into currency speculation made Mexico's big banks especially vulnerable to a peso devaluation. It was clear, therefore, that the financial system was balanced on a knife edge and could collapse either if interest rates were raised to defend the peso or if the peso was devalued to prevent higher interest rates. No one who followed the situation closely could help but be worried about Mexico's banks.
At this point, anyone who could add and subtract whole numbers could see the potential for a crisis. The surge into Tesobonos beginning in March was matched by a flight from them in November, which drained Mexico's supply of hard currency reserves. With $28 billion of Tesobonos outstanding and only $12 billion of reserves, it was clear that Mexico would run out of reserves before it ran out of obligations if capital flight continued. Investors sold Tesobonos as well as peso-denominated short-term investments, which started a run on the peso itself. This is how The Economist summed up the climax of the peso crisis:
At the Summit of the Americas in early December 1994, the Mexican miracle was lauded as a paradigm of successful economic reform and an example to the rest of the region. Ten days later, on December 20th, plummeting foreign-exchange reserves forced the newly elected president of Mexico, Ernesto Zedillo, to devalue the peso. The badly handled devaluation created panic among investors. Money poured out of the country; within a month the economic miracle was teetering, for the second time in 12 years, on the edge of default.4
With its reserves depleted, there was no way for the Banco de Mexico to keep the peso within its dollar band and the peso was devalued on December 20, 1994. The peso fell in an instant from an exchange rate of about 3.2 pesos per dollar to more than 5 pesos per dollar, eventually stabilizing at about 7.5 pesos per dollar—a fall of more than 50 percent. The dream of economic prosperity and political stability for Mexico was defeated by reality—the reality of the hot money investors who led the flight from Mexico and the cold, hard balance of payments arithmetic that made devaluation inevitable. The best that can be said is that the crisis was delayed until after Zedillo's election, which probably reduced the risk of political insurrection.
The peso panic provides a clear example of the sort of crisis to which international financial markets seem prone. Investment flows, trade flows, prices, interest rates, and exchange rates—all distorted in ways that heighten social tensions—distort political choices and discourage business investment.
Devaluing the peso did nothing to resolve either Mexico's short-term liquidity problem or its longer-term current account deficit dilemma. Inflation increased to over 50 percent per year, wiping out any competitive gains from a cheaper currency, and interest rates rose to startling levels. In the end, Mexico suffered both devaluation and higher interest rates, the two things it wanted to avoid. About 1.5 million workers lost their jobs as recession and austerity struck.
Banamex, the largest bank, survived the crash, but only thanks to a $2 billion government aid package. Other major banks also sought government aid. Mexico was able to stabilize the peso only with the help of a controversial assistance package from the United States. A Tequila effect hung over Latin America and other "emerging market" economies for more than a year, as investors remained cautious.
What is most interesting to me about the peso crisis is the fact that the problem was so clear and that at each step the cause and effect relationships were so obvious. The current account deficit problem, which was the ultimate source of the peso crisis, was general knowledge, especially among financial investors. When The Economist magazine surveyed the situation in Mexico in February 1993, almost every aspect of the story just told, aside from assassinations, was readily apparent.5 A survey by the Financial Times in March 1993 questioned whether Mexico was on its way to another debt crisis. The question, it said, "is not whether the deficit can be financed—because after the fact a country's current account deficit will always equal inflows of capital adjusted for changes in reserves. The question is at what level of interest rates and therefore economic growth will the equilibrium between capital inflows and current account deficit be reached."6 The same story cautioned that "For comfort's sake, too high a proportion of the inflows into Latin America over the past few years is 'hot money'—money attempting to capture profit from interest differentials or foreign exchange market inefficiencies, and which is likely to be withdrawn as soon as the perceived risk associated with the investment increases."7
The threat of a crisis was even clearer in April 1994 after the Colosio assassination. A Financial Times survey of Mexico was headlined "The Picture Darkens":
While financial markets proved unexpectedly resilient in the aftermath of the assassination of Mr. Colosio, the killing led to future increases in interest rates. If more violence lies ahead, the political uncertainty may force the government to choose between still higher interest rates that would choke the recovery or a faster rate of devaluation of the peso that could cause an upsurge in inflation.8
This is a sound analysis of the choice that had to be made in November 1994. Readers of the Financial Times were warned in April of the events of November and December. Yet even with these clear warning signs, the march to crisis continued.
On December 20, 1994, no one had much choice. Zedillo was forced to devalue—nothing else made sense. The investors were also forced to sell Tesobonos. The peso crisis seemed at that point inevitable. Yet this crisis resulted from actions that were rational, logical, and reasonable at the time and were taken with fairly complete knowledge of the risks and basic arithmetic of the situation.
This chapter's epigraph frames the peso crisis as a "defeat by reality." Gabriel García Márquez might better have said that they were betrayed by reason. How does reason create crisis and chaos? And is this condition an anomaly or is it a fundamental property of international capital markets and exchange rate systems? For the answers to these questions we must consult economic theory.
The Theory of Financial Crises
As stated in The Economist, "To many, Mexico was one more sign of how dangerously volatile the brave new world of free global capital flow is.... Yet for anyone with a sense of history, there was little new in the Mexican débâcle."9 Indeed, crises of this sort are a common feature in the history of international financial markets.
Currency crises are so much a natural part of international markets that it is perhaps true that the only way to truly eliminate them is to eliminate currencies themselves by regressing to a barter system or to eliminate exchange rates by adopting a single world currency. Either of these options trades one sort of crisis for another, however.10 Money and exchange rates will always be a feature of the international system, and so, therefore, will currency crises.
The best place to begin to learn about currency crises in particular and financial crises in general is in a little book by Charles P. Kindleberger, Manias, Panics, and Crashes: A History of Financial Crises.11 Kindleberger synthesizes the theoretical model of Hyman P. Minsky with his own deep understanding of financial history and experience of contemporary international economics. His insights help us put the peso crisis of 1993-1994 into context.
Like Tolstoy's happy families, financial crises all resemble one another—in certain basic characteristics, at least—and like the unhappy families they are each unique in particular circumstance. Don't be concerned that this makes currency crises both happy and unhappy—this is a dream world, remember? The common features or stages in the development of a financial crisis are these:
- Displacement
- Expansion
- Euphoria
- Distress
- Revulsion
- Crisis
- Contagion
Imagine for a moment a market for investments that has reached some sort of equilibrium, where investment flows are consistent with the information known to and expectations held by the market's participants. According to Minsky and Kindleberger, financial crises appear and then develop through these stages.
Displacement refers to an external shock or some "news" that fundamentally alters the economic outlook in a market, shifting expectations concerning future profits in some significant way. Displacement—in the sense of a change that affects expected profits—happens all the time, of course, and seldom leads to panic, crisis, or instability of any sort. The sorts of displacement we are concerned with create an object of speculation, some asset or financial instrument that becomes the focus of investors based on the news, creating a "boom."
Speculative objects appear and disappear with great frequency in financial markets, seldom creating panics or crises. So speculation and crisis may be related, but they are not the same phenomenon. Expansion is a necessary prerequisite for a financial crisis to rise out of a speculative episode. Expansion is the stage where the boom is fed by an increase in liquidity, which provides the means for the boom to grow, perhaps becoming a bubble. Although Kindleberger focused on increases in bank credit as a common source of expanding liquidity, there are many potential sources. Financial innovations, increased leverage, margin buying, and other techniques can stretch more buying power from a given monetary base.
Perhaps the most obvious form of expansion is the widening of the pool of potential investors or speculators, from a set of "insiders" to a larger group of "outsiders." Walter Bagehot, the great nineteenth-century political economist, suggested that panics form when an object of speculation attracts the greed of authors, rectors, and grandmothers. "At intervals, from causes which are not to the present purpose, the money from these people—the blind capital, as we call it, of the country—is particularly craving; it seeks for some one to devour it, and there is a 'plethora'; it finds some one, and there is 'speculation'; it is devoured, and there is 'panic.'"12
Expansion becomes euphoria when trading on the basis of price alone takes the place of investment based on fundamentals. The purpose of buying is to sell and take a capital gain as the price continually rises. The new buyer's motives are the same, and this euphoria continues as long as expectations do not change and liquidity holds out. This is the period of what Adam Smith called overtrading and Kindleberger termed pure speculation—that is, speculation on the basis of rising prices alone. A bubble (that will burst) or a mania (driven by wild-eyed investor maniacs) may here be created.
Distress is the next stage of a classic crisis. Distress is the stage between euphoria and revulsion when there is concern that the strength of the market may be fragile or that the limits of liquidity may be near. Distress is an unsettled time, and the reactions to this unsettled environment often deflate the bubble and defuse the mania. Distress can persist for lengths of time until the crisis is averted, or it can turn sharply into revulsion.
Revulsion is a sharp shift in actions and expectations caused by new information or a significant event. Insiders realize the importance of the news and sell first, perhaps at the top of the market, while outside authors and rectors are still buying. Liquidity dries up, especially bank lending, causing discredit. Walter Bagehot saw the rush to liquidity as the result of a loss of confidence in all but the most liquid assets. He wrote in The Economist that
It has not been sufficiently observed how very peculiar and technical is the sense in which we now talk of "panic." It would naturally signify a general destruction of all confidence, a universal distrust, a cessation of credit in general. But a panic is now come to mean a state in which there is confidence in the Bank of England, and in nothing but the Bank of England. There is an increased demand during a panic for Bank of England notes; at such times an enlarged trust is reposed in the Bank, but there is a much diminished confidence in everyone else. Distrust is diffused, but the Bank of England does not feel it; the use of its credit is augmented.
The reason is obvious. There is, in the ordinary working of banking in England, a refined mechanism of diffused credit which economises the use of bank notes, of visible instruments of exchange, of money in the ordinary sense of the word.... In a panic, this auxiliary and supplementary currency is at once in part annihilated. Its very foundation is taken away. That foundation is credit, and instead of credit there is discredit. . . .
In a panic, this currency of checks—this currency of refined credit—is much disturbed, and is in part destroyed; and, therefore, we fall back on credit of the first order—on credit of the coarser sort—upon bank notes. We require more bank notes, just because the feeling, the confidence which make few bank notes effectual has disappeared.13
In Minsky's model, revulsion and discredit lead to crisis, as outsiders join insiders in selling off. Kindleberger proposes the image created by the German term Torschlusspanik, gate-shut panic, to describe the rush to liquidity. The falling prices feed on themselves creating self-fulfilling prophecies. The result is a crisis, which may also be crash (collapse in price) or panic (sudden needless flight).
The crisis may be confined to a single market or it may spread, which is termed contagion. We are especially concerned with crises that spread from nation to nation through international linkages such as capital, currency, money, and commodity markets, trade interdependence effects, and shifting market psychology. Paul Krugman reserved the term contagion crisis for a financial crisis that spreads internationally to the extent that it causes a worldwide depression.14
What brings the crisis to an end? There are three possibilities, according to Kindleberger. The crisis may turn into a fire sale, with prices falling until buyers are eventually brought back into the market. Or trading may be halted by some authority, thereby limiting losses. Or, finally, a lender of last resort, of which we will hear more soon, may step in to provide the liquidity necessary to bring the crisis to a "soft landing."
The Peso Crisis in Perspective
Does Minsky's model of financial crisis describe the peso crisis? It is not difficult to frame the events of 1994 in terms of the seven stages of a financial crises just discussed.
Since we must begin somewhere, I choose Mexico's decision to join NAFTA as the critical displacement that created new objects of speculation in Mexico (how many of us bought Telefonos de Mexico stock on its way up?). It is more complicated than that, of course. Mexico's policies of liberalization, privatization, and openness created buoyant economic optimism (in foreign capital markets, anyway), and the NAFTA sealed the deal. Former "banana republic" Mexico seemed headed for a period of stable democracy, private enterprise, and growth. Everyone wanted to get in on the deal.
Expansion followed, driven by many factors. By the 1990s, financial markets were well-organized to mobilize the funds of authors, rectors, and grandmothers to invest in foreign countries that many would be hard-pressed to find on a map. The era of "global" and "emerging markets" mutual funds was here, creating the conditions for a classic speculative bubble. A modest recovery in economic prospects from the dismal 1980s led to large capital gains for those few investors who had been willing to put money into Third World markets. Their success led other investors to jump in, driving prices yet higher. And by 1993 or so "emerging markets funds" were being advertised on the television and the pages of some popular magazines.15 As speculation on price alone took off, even fund managers became uncritical of their investment decisions, driven as they were to invest the huge sums coming in from authors and rectors every day. "We went into Latin America not knowing anything about the place," one of them noted after the Mexican crisis. "Now we are leaving without knowing anything about it."16 You can see how disconnected investment became from any analysis of the realities involved.
"During the first half of the 1990s," according to Krugman, "a set of mutually reinforcing beliefs and expectations created a mood of euphoria about the prospects for the developing world. Markets poured money into developing countries, encouraged both by the capital gains they had already seen and by the belief that a wave of reform was unstoppable."17 We have reached euphoria, in terms of Minsky's model, the condition in which rising prices draw new investors who expect capital gains when prices rise again. This was true for Mexico and for the "emerging" markets of several other countries. One reason the tequila hangover of the peso crisis was so severe is that it revealed the possibility of many Mexicos and many crises.
Distress can be located in March 1994, with the assassination of Donaldo Colosio, which raised significant doubts among foreign investors about the political stability of Mexico. The era of political stability and economic expansion that Carlos Salinas had engineered was suddenly threatened. Short-term foreign investment shifted away from private securities and the peso-denominated Cetes, focusing instead on the dollar-indexed Tesobonos. Tesobonoswere investments that paid higher interest rates than other dollar investments, but without the exchange rate risk of peso securities. Tesobonos joined certain other Mexican assets as objects of speculation as investors shifted away from many other Mexican portfolio investments.
The critical importance of this distress is that it changed significantly the nature of Mexico's current account deficit problem. With the switch to dollar-backed Tesobonos as a way to finance the payments imbalance in the short term, Mexico's hard currency reserves were suddenly at risk. Mexico's $12 billion of reserves now guaranteed what became $28 billion of dollar-linked securities, with no international Federal Deposit Insurance Corporation (FDIC) insurance to prevent a run on the bank. It was at this point that the potential for crisis became clear. After March, in retrospect, the Mexican situation seems very delicate; it wouldn't take much to cause a sell-off.
The events of November 1994 caused revulsion. As an economist, I naturally look to the U.S. Federal Reserve's interest rate increase as the key factor. Insiders would have seen immediately the pressure that this put on Mexico's reserves, given their disinclination to raise interest rates to maintain foreign capital flows. It would be an error, however, to underestimate the importance of Attorney General Ruiz Massieu's resignation. A true PRI insider himself, Massieu's action was a signal of corruption and distrust inside the party and therefore the federal government. Some financial insiders might have interpreted this political crisis as an indication of the government's domestic weakness and, therefore, its eventual inability to raise interest rates once the presidential election was past.
Revulsion came, therefore, as Mexican authorities found themselves in a position of having to choose between their international financial responsibilities and their domestic political survival at a time when pressure was rising both inside and outside the country. The iron arithmetic of reserves versus dollar-indexed liabilities was suddenly a critical factor. Insiders caught the scent of a crisis and ran for the shutting gate doors.
The peso crisis was sharp and unexpected—or so the newspaper said. President Ernesto Zedillo certainly seemed unprepared for it. Fear of a devaluation caused the peso to drop, which put an even greater strain on Mexico's scarce reserve stock. When Zedillo abruptly canceled a television speech on the crisis, tension mounted and panic set in. When Zedillo finally did appear on December 20, he was forced to announce a very large devaluation of the peso.
Contagion occurred both within Mexico and between Mexico and other countries. The effects of peso depreciation, domestic inflation, and higher interest rates caused Mexico to experience a severe recession. Unemployment rose sharply from just 3.2 percent in December 1994 to 7.6 percent in August 1995 before falling to about 6.0 percent in 1996. Inflation, as measured by monthly changes in the National Consumer Price Index, rose from 3.8 percent per month in January 1995 to 8 percent in April, then fell back to the 2 to 3 percent per month range in 1996. Interbank interest rates soared, reaching 86.03 percent in March before falling back, although they remained above 40 percent until April 1996. It is hard to know what the final impact of capital flight from Mexico will be. Foreign capital deserted both weak and some strong investments.
Although the peso crisis recession in Mexico may turn out to be relatively short, it is also relatively deep, and its effects may be long lasting. Mexico's gross domestic product fell dramatically in 1995, effectively wiping out the short-run gains from the NAFTA boom and leaving Mexico's citizens not much better off—if at all—than in the old days before market reforms. Recovery was well under way by early 1996, but growth was highly concentrated in the export sector, which benefited from the peso's lower value. Mexico's internal economy—that part not directly affected by exports—remained deeply depressed by a combination of high interest rates, credit shortages, and general poverty.18
International contagion also occurred, notably to other "emerging market" nations that suffered from the tequila hangover effect. Krugman took a pessimistic view that because
the 1990-95 euphoria about developing countries was so overdrawn, the Mexican crisis is likely to be the trigger that sets the process in reverse. That is, the rest of the decade will probably be a downward cycle of deflating expectations.... This new reluctance will surely be directly self-reinforcing, in that it means that the huge capital gains in emerging market equities will not continue. It will also lead to a further slowing of growth in those economies comprising much of Latin America and several outside nations, whose hesitant recovery in the 1980s was driven largely by infusions of foreign capital.19
International contagion was not limited to emerging markets, however. Investors who worried that the falling peso might drag the dollar down with it joined in a run on U.S. currency. The dollar fell to historic lows against the yen (trading briefly below eighty yen per dollar) and the deutschmark in early 1995. The peso crisis created a dollar crisis, which put pressures on trade and financial structures around the world.
Lessons of the Peso Crisis
What did we learn from the peso crisis? Well, I don't think we learned anything really new. The peso crisis and its fallout were a classic case of the sort of crisis that financial markets, and especially international financial markets, are prone to experience.
One lesson that apparently had to be relearned was the difference between direct foreign investments, which are long-term real asset purchases, and hot money investment, which is short-term highly sensitive portfolio investment. All capital flows are not the same: "Another problem with the capital flowing to the region was its composition. While direct foreign investment in factories, utilities, and mines was boosted by privatization and deregulation, a record-breaking portion of capital flowed in as foreign portfolio investment in stock and bonds. Portfolio investment is always volatile, and in these days of electronically linked capital markets, it can leave a country literally at the speed of light."20
A second lesson was that our ability to deal more effectively with domestic financial crises does not translate into a similar ability on the international level. International crises are related to domestic crises but also differ in important respects. Deposit insurance, regulatory regimes, and the existence of a domestic lender of last resort have all but eliminated the seven-year cycles of financial boom and panic observed in the nineteenth century. International crises, however, have increased in frequency. Perhaps the sense that we are better able to control domestic financial problems has prevented us from appreciating the international problems that are harder to control.
Kindleberger saw this pattern driven in part by the very communications advances that have accelerated the trend toward "global" investments. Communications lags, he argued, slow down the process of bubble building and allow more time for pressures to diffuse. Faster communications also allow bubbles to build more quickly and burst faster. As Kindleberger said, "Today's troubles travel instantaneously."21
I think that Kindleberger was almost right. Technological changes probably have not altered the speed of these transactions very much, or as much as we like to think, but communications improvements probably have broadened the domain of the crisis, which makes it harder to control. As I noted in Chapter 2, I think the communications and technological revolutions have done little to increase information flows between nations from Keynes's day, but they have vastly improved the spread of information within nations. There is more at stake in a currency crisis now because of this widened domain of affected parties. Expansion is not quicker now; it is just wider. Rectors and authors aren't drawn in sooner, they are drawn in more often because of the wider distribution of financial information. This tends to make the environment necessary for the development of international crises more common today than in earlier years.
A third very important lesson was that the liberal market-oriented policies that Mexico and many other "emerging market" nations adopted in the 1990s were not enough in themselves to produce stability and growth, as some apparently believed. These policies, which Paul Krugman dubbed the Washington consensus of desirable market reforms, exist within a larger and unaccommodating financial environment.22 For emerging market economies, as for all economies, domestic stability is limited by the stability of the international financial markets. Mexico discovered that all its costly domestic reforms were vulnerable to a run by foreign investors. "As surprising as the dramatic swing in the perceptions of international investors about Latin America is," wrote Moisés Naím in 1995, "the region's vulnerability to such swings is even more surprising."23
Perhaps the last and most controversial lesson to be learned is that there is no international lender of last resort. Although the United States did eventually organize an international "bail out" of Mexico in 1995, which allowed the Tesobonos debt to be retired and Mexico's international reserve position to be restored, this action did nothing to calm the peso crisis at its peak and may have helped fuel the dollar crisis through errors of omission. For the most part, the peso crisis burned itself out independent of U.S. action.
The idea of a lender of last resort is usually attributed to Walter Bagehot. Bagehot argued that financial crises could be defused successfully by a lender of last resort that, in the face of a panic, would provide credit in exchange for good collateral at a high rate of discount. If a panic is a flight to liquidity in a dangerously uncertain situation, then a guarantee of liquidity would deflate the issue safely, Bagehot argued. Today, it is widely understood, although hard to prove, that a domestic lender of last resort can cut short financial crises and moderate their effects. What does remain controversial, however, is the stronger notion that the very existence of a lender of last resort causes crises to form because of the moral hazard this creates. Although the existence of a lender of last resort should prevent revulsion and crisis, it might also encourage riskier investment behavior and expand the speculative domain.
These same problems exist when we consider the notion of an international lender of last resort. Such an institution would have the resources necessary to lend reserves to nations during currency crises. If an international institution had stood ready to lend Mexico additional reserves (on good collateral, at a substantial discount, following Bagehot's dictum), would the peso crisis have happened at all? Remember that Mexico was illiquid, not insolvent; it had adequate assets, and only its ready hard currency reserve was insufficient.24
It can be argued, and some have argued it, that an international lender of last resort is unnecessary because financial crises tend to be domestic crises. The degree of international contagion is limited, it is argued, by the relatively low degree of true integration of the global economy.25 Even a collapse of Mexico is unlikely to bring down the United States, Japan, and the European Union. (Although a crisis in the United States would have significant external effects.) A currency crisis is therefore unlikely to turn into what Krugman called a contagion crisis—a global depression. A crisis can do considerable damage to trade and investment flows and living standards without causing a worldwide economic catastrophe.
The idea of a lender of last resort raises all sorts of practical policy questions. What country or international institution would play this role? (This job certainly does not fall under the job descriptions of either the International Monetary Fund or the World Bank, as presently organized.) How much money would be required? Who would put up the reserves? On what terms? What sort of collateral would be accepted? At what discount? To what limit? Under all circumstances? And what of conditionality—shouldn't there be "strings" attached to assistance to ensure an end to the behavior that created the crisis in the first place?
Each of these questions raises important issues and so it is not surprising that although the idea of an international lender of last resort has been widely discussed ever since Bretton Woods, no consensus has been reached and no action taken. This does not mean that the question is closed. Improvements in communications technology and financial innovation make questions relating to the stability of the system of global markets more, not less, important.
"I do not forecast world economic collapse, because I think that our profession of economics does not know the dynamics of the system well enough to do so," Kindleberger noted. "At the same time, the international lender of last resort seems worth thinking about, if only for contingency planning."26 The case of the peso crisis of 1994-1995 doesn't cause anyone to forecast global collapse, but it surely makes the case for "thinking about" an international lender of last resort stronger.
The European Currency Crisis of 1992
September 16, 1992, is called Black Wednesday by people who follow European currency movements. That was the date of a currency crisis distinctly different from Mexico's. Black Wednesday was no bubble or rational panic; it was a new kind of currency crisis distinctly different from anything that Walter Bagehot might have imagined. This was a speculative attack on the European Community's exchange rate mechanism (ERM).27
In 1992 the ERM was an agreement designed to limit exchange rate variation within the then twelve-member European Community (EC). The rationale for the ERM was that more stable exchange rates were needed to achieve greater regional economic integration with the EC. ERM states agreed to intervene in currency markets and to coordinate macroeconomic policies with an aim to keeping their exchange rates within (plus or minus) 2.25 percent of a central rate (a total variation of 4.5 percent was thus possible). The only exceptions to this rule were the Spanish peseta and the Portuguese escudo, which were subject to a broader band, reflecting their status as softer currencies. The ERM rules allowed for periodic revaluations of exchange rates—shifting the bands—when changing macroeconomic conditions required. In fact, however, the ERM had been stable since 1987.
The Maastricht Treaty intensified interest in the workings of the ERM. The Treaty on European Unity, which was negotiated in the Dutch town of Maastricht in October 1991, proposed tighter economic and political integration of EC members. When the treaty finally went into effect on November 1, 1993, the EC symbolically changed its name to European Union to reflect the intended closer relationship. A key element of the Maastricht Treaty was the proposal for monetary integration, with the eventual goal of a single European currency (eventually named the euro) and tight coordination of member state monetary policies.
The Maastricht Treaty proposed that the ERM was the first step in a process that would eventually lead to a single currency. As national economic policies and conditions converged, narrower exchange bands would further reduce currency variation until—voilà—a unified currency was achieved. The goal of a single currency was and, at this writing, still is very controversial because it asks nations to sacrifice one aspect of their sovereignty—self-determination of economic and especially exchange rate policies—to achieve the goal of partial exchange rate stability.28 Because this goal was so controversial, each EC member put the Maastricht decision to its citizens in the form of referenda. The Maastricht Treaty was not ratified until October 1993, when all member states' citizens had voted approval.
The importance of economic policy autonomy was especially apparent in 1992 because a period of relative policy convergence of the main European economies was coming to an end. The unification of eastern and western Germany was very expensive and potentially highly inflationary.29 Knowing that a tax increase to pay for unification would intensify domestic political conflicts, the German government chose to finance its actions through debt. The most important consequence for the current story was an increase in German interest rates, needed to both attract foreign capital and to contain inflationary pressures.
Germany's economic policies put them in conflict, in terms of the ERM, with those of France, Italy, and Great Britain. These countries, suffering from continued high unemployment, saw the need for lower interest rates to encourage economic expansion and job creation in their countries. This conflict of interests was a matter of concern within the ERM because higher German interest rates, if not matched by higher rates in France, Italy, and the United Kingdom (UK), would draw capital out of these countries and the resulting sell-off of their currencies would push them below the ERM floor. This is a classic case of domestic economic interests that conflict with international obligations as represented by the ERM commitment. Election cycles in these countries made their domestic interests potent, but the increased emphasis on exchange stability inherent in the Maastricht Treaty also created additional pressure to preserve the ERM relationships. Thus was created an environment conducive to a political-economic crisis.
The most important events leading up to Black Wednesday are these. On June 2, 1992, Denmark's voters unexpectedly rejected the Maastricht referendum, a vote that stunned Europe (Danish approval was gained later, in a second referendum). The Danish vote increased the doubts about the commitment of several nations to monetary unification, which also raised doubts about their commitment to defending the ERM.
These doubts grew deeper on July 16, when Germany's Bundesbank raised key interests rates in its continuing anti-inflation policy. This action was seen by many as an almost aggressive act, a statement that Germany would not sacrifice its domestic priorities and that other ERM states, especially France and the UK, would have to follow suit and raise their own interest rates, following the German example instead. It was not at all clear that they would do this. This set up a showdown on economic policy, with the credibility of the ERM in the balance.
"This policy conflict did not remain unnoticed by the speculators," noted Paul De Grauwe, economist and member of the Belgian parliament. "They realized that the UK and French authorities were tempted to cut their links with the deutschmark so as to be able to follow more expansionary monetary policies. Influential economists in fact openly urged the authorities to do just that. Thus speculators had good reasons to start speculating against the pound sterling and the French franc."30
In fact, several countries encountered great difficulty defending their ERM band commitments. Higher German interest rates pushed up the deutschmark and pushed down the other currencies. Unwilling to raise their interest rates, French, German, and Italian central bankers were forced to use their scarce reserves to defend their currencies from market forces. This was a battle they could win for only a limited time, however, because of the finite nature of their resources compared with the $1 trillion per day foreign exchange market.
On September 13, 1992, the commitment to ERM broke down. Italy devalued the lira by 7 percent but stayed in the ERM in exchange for a reduction in German interest rates. When the rate cut came the next day, it was a tiny one-quarter percent drop in the Lombard rate, clearly too small to turn around the market forces that were starting to tear the ERM apart.
France's referendum on the Maastricht Treaty was set for September 20, and this added to the pressure on the ERM. French voters, although traditionally pro-Europe, were not necessarily pro-unemployment, which is how many of them interpreted the prospect of a permanent connection to (or domination by?) the austere Germany Bundesbank. It looked like Maastricht might lose the vote in France, taking pressure off France to maintain its ERM stand. (Maastricht won, in fact, but by a slight 51 percent Yes to 49 percent No margin.)
Speculation intensified against the currencies held to have the weakest commitment to the ERM or the least reserves to defend their position. Hedge funds and others leveraged their assets and used them to sell short the target currencies. It was a one-way bet with enormous potential gain for someone with the resources and courage to bet really big. If the ERM held together, then the key currencies would not fall further than their floors, but they were unlikely to rise much above them, either. At worst, selling short could create a small loss. If the ERM collapsed under the speculative attack, however, short-sellers could fulfill their contracts at a much lower price and profit handsomely. Insignificant loss versus high gain. No wonder they bet against the pound and the lira.
Enormous speculative pressure was brought to bear against the pound, lira, and franc. Great Britain, facing severe reserve constraints, raised the base interest rate from 10 percent to 15 percent in two stages, but it wasn't enough.31 Britain's actions were more than adequate to bring their "fundamentals" into line with German financial returns, but the market's driving force had shifted from investment to speculation. Britain's desperate move made further defensive action seem unlikely, strengthening the speculative attack.
The crisis came to a head finally on September 16, when speculative pressures overwhelmed Britain and Italy. Unable to muster more reserves and unwilling to raise interest rates further, Britain and Italy dropped out of the ERM and allowed their currencies to float down to market levels. The speculators covered their short contracts at the lower rate and took their profits to the bank.
The devaluations, when they came, were not nearly so large as in the Mexico crisis. The pound's value fell from its ERM floor of 2.77 deutschmarks per pound on September 16 to 2.53 deutschmarks on September 21. The devaluation of the lira was smaller, but it had already been devalued earlier in the week. France, also subject to speculative attack, successfully defended the franc in the short run, but pressure continued until it cracked, too, in 1993.
Black Wednesday and its aftermath showed clearly that currency crises were not limited to the cases of less-developed countries with current account problems. Britain and France, the largest victims in the ERM's crises, in many ways had stronger economic fundamentals than Germany, with its unification problems and costs. This did not protect them, however, from crises driven by speculative attack.
The Theory of Speculative Attacks
The theory of speculative attacks, as conceived by Paul Krugman and developed by him and others, is deceptively simple.32 Assume an exchange rate system like that of the 1980s and 1990s. Capital is highly mobile across national boundaries, so that exchange markets are deep—the daily flow through exchange markets, a trillion dollars a day in the mid-1990s, vastly exceeds the reserves of the central banks. Exchange rates are flexible enough so that currency speculation is an active industry. Many governments intervene in the currency markets in an attempt to keep their currencies within target zones, which may be formally set, as with the ERM, or informal but real, as has often been true of the yen-dollar exchange rate. A system like this, designed to provide relative stability within a regime of flexible exchange rates, contains a systematic flaw: it is prone to speculative attacks that create exchange rate crises.
This conclusion is important because it means that the exchange rate system just described contains inherent instabilities. It is not just that rational bubbles sometimes form and burst or that the poor policies of deficit nations can cause soft currency crises. The tendency toward instability and crisis is built in. The theory of speculative attacks suggests that events like the ERM crisis of 1992 are inherent risks.
Imagine a situation in which a nation, say France, has set a target zone for its currency that it will defend using only its central bank reserves or some fraction of them. When the franc is at or near the border of its target zone, the dynamics of the speculative attack come into play. Speculators attempt to force a devaluation of the franc—a realignment of the target zone—by selling short massive amounts of francs, leveraging their own assets in an attempt to gain the central bank assets and cause a currency crisis. As noted earlier, this situation can be viewed as a "one-way bet." If the franc is devalued, then speculators win, repaying the francs they have sold with cheaper francs purchased on the spot market. If the central bank succeeds in fending off the attack, the speculators lose little, since they buy back francs for roughly what they paid for them, the target zone price. If the franc appreciates, which may be viewed as unlikely, speculators lose on their short contracts—perhaps significantly—buying back francs for more than they paid for them.
One way to think of the speculative attack scenario is that when the possibility of franc devaluation appears, non-franc reserves are a scarce and valuable asset, since they gain in value when depreciation occurs. The central bank's non-franc reserves are therefore the object of the attack. Speculators want these reserves so they can sell them back after devaluation at a profit. By engaging in a speculative attack, they force the central bank to yield to them their scarce and valuable reserves.
It is the perverse logic of the target zone system that no currency is safe once it is near enough its target zone border to make a speculative attack credible. If the central bank has a relatively small amount of reserves, a single attack of short-selling will work (as with Italy in 1992). If the central bank has greater reserves, as France did in 1992-1993, for example, it is still not safe. Since speculators do not know what fraction of its reserves the central bank is willing to use in defending its currency, a series of smaller attacks occur that eventually drain reserves to the critical level where a single large attack suffices to cause devaluation. This scenario fits well the story of the French franc, which survived the 1992 ERM crisis but not the attacks of July 1993.
Speculative attacks are self-fulfilling prophecies. Because speculators believe that a currency is vulnerable to devaluation, it is. And because they think that commitment to the target zone is not credible, it isn't. The rational, profit-maximizing actions of speculators create the results they expect, which are currency crises and sharp, discontinuous changes in exchange rates. According to De Grauwe, "Some continental European observers and politicians have claimed that the speculation against the pound sterling and especially against the French franc was irrational and driven by an 'Anglo-Saxon plot' against the process of monetary unification in Europe. Such explanations based on irrational motives cannot easily be disproved. One should be suspicious about these explanations, however. In general, speculators want to make money, and do not care about the colour of the money they expect to earn."33
Central bank reserves are not the only tool available to squelch speculative attacks. Central bank authorities can, in theory, mobilize private capital if they are willing to alter macroeconomic policy variables as part of a defensive strategy. Great Britain, for example, raised interest rates in 1992 in just such a move. If sterling's defenders had been willing to raise interest rates high enough, eventually they would have pulled sterling back from the brink of devaluation, and speculative attacks would have ended or moved on to other targets. The problem with this strategy is that higher interest rates have domestic as well as international effects. The growth-sapping impact of higher interest rates made them an unacceptable choice in 1992.
International cooperation is another strategy that can work to defend against speculative attacks. A small increase in French interest rates combined with lower German interest rates would have the same effect on the franc-deutschmark exchange rate with fewer unpleasant side effects within France. The difficulty, apparent in the ERM crises, is that there are many circumstances in which the domestic priorities of central banks will not be so aligned as to make effective cooperation possible. Recall Germany's quarter-point cut in the Lombard rate in 1992—cooperation in technical terms but insufficient to avert a crisis. If economic policymakers give domestic concerns high priority, which they do, and the nature of macroeconomic problems and priorities differ among nations, which they do, then the strategy of cooperation is an unreliable defense from speculative attacks.
Speculative attacks are a serious matter. They can have serious effects (e.g., lost reserves, distorted macroeconomic policies, inflation, unemployment, rising interest rates) on the particular country whose currency is under attack, but the negative effects are not limited to the target country. The evidence suggests that some impacts are spread to other economies through contagion. A recent study by Barry Eichengreen, Andrew K. Rose, and Charles Wyplosz, for example, examined the factors that make currency crises contagious and therefore an even more serious systemic problem.34 Using thirty years of panel data from twenty industrialized countries, they found evidence of contagion, which varied directly with the degree of interdependence among affected nations. "The evidence is striking," they wrote, "a variety of tests and a battery of sensitivity analyses uniformly suggest that a crisis elsewhere in the world increases the probability of a speculative attack by an economically and statistically significant amount... even after controlling for economic and political fundamentals in the country concerned."35 Crisis in any single currency therefore increases the risk of crisis generally.
The problems of currency crisis seem to pass from the "affected" to the "infected" nation especially through the trade and macroeconomic policy channels. The peso crisis, for example, affected Mexico directly, but it also caused problems for nations that are closely linked to Mexico by trade ties. The greater the trade linkage, the more powerful the contagion effect. The other way that contagion occurs is through macroeconomic policy effects. When a country such as Mexico suffers a crisis, other countries in roughly similar circumstances also come into play. They find themselves under pressure to adopt preventive policies so that, for example, higher interest rates in Mexico seem to be contagious. This is at least one aspect of the Tequila effect that followed Mexico's peso crisis.
Because, according to this study, contagion increases with the degree of international economic integration, the problem of currency crisis and speculative attack may be directly related to economic globalization. Globalization links nations more closely together, but the closer the links the greater the risk of instability due to contagion. This situation naturally limits the extent of globalization, if an equilibrium or balance between globalization and stability occurs. But it could also lead to cycles of globalization, instability, crisis, and chaos, followed by another round of globalization. In either case, currency crises and speculative attacks seem to make the notion of true globalization more myth than reality.
What we have seen thus far is that currency crises due to speculative attacks are based on the profitability of speculating against the central bank, the finite nature of central bank reserves, conflict between domestic priorities and the need to defend the currency, and the unreliable nature of international cooperation. The effects of crisis are contagious and increase with the degree of international economic integration. The conditions for successful speculative attacks are fairly common, and the resulting currency crises have become a regular feature of foreign exchange markets. Can the exchange rate system be modified to eliminate this threat or make it a less frequent occurrence? Four types of systemic changes come to mind.
An obvious first proposal, given the earlier discussions, is an international lender of last resort, willing and able to provide liquidity to central banks facing speculative attack. Such an institution would need to have virtually infinite reserves, however, and be willing to provide them freely in emergencies. An international lender of last resort of this magnitude is not practical in either political or economic terms, and it is unclear in any case whether the establishment of such an organization would result in more stable exchange rates or just more prosperous speculators.
A worldwide tax on foreign exchange transactions has been proposed as a solution to the speculative attack problem.36 Such a tax could conceivably alter the dynamics of the one-way bet and would tend to discourage speculation, especially highly leveraged speculation, by making it costlier than other sources of profit. Such a tax would also make all other foreign exchange transactions more costly, however, putting "sand in the wheels" of international trade and finance generally. The tax would need to be universal to be effective, and such worldwide regulatory agreement is considered unlikely.
A third solution would be monetary union, which would eliminate exchange rates through adoption of the single worldwide currency. This, too, is considered impractical at present, but it raises an important point. Once the European Union achieves monetary union, with a common monetary policy and a single currency, speculative attacks like those in 1992 and 1993 will not be possible. The goals of exchange rate stability will be achieved, presumably, by monetary union. But if the transition to monetary union must take place though a system of target zones, then the goal may never be achieved. It is the nature of speculative attacks on target zones that they produce currency crises that tear target zone structures apart. In Europe, the goal of a monetary union and the process of target zone convergence are thus inconsistent.37
Finally, the sharp point of speculative attacks can be blunted by target zone structures that feature broad bands, not narrow ones. Central bank intervention would tend to be less frequent under these circumstances, and so attacks of bank reserves would be less common as well. The logic of speculative attacks would remain, however, as would the problem of currency crises.
Exchange Market Mayhem
The exchange rate is the chink in the armour of modern-day macroeconomic policy-makers. Be it Italy and the United Kingdom in 1992, France in 1993, Mexico in 1994 or Spain in 1995, speculative pressures and the dire consequences of policy responses required to defend the exchange rate can bring a government's entire macroeconomic strategy tumbling down.... Even the United States, a relatively large closed economy committed to a policy of benign neglect, was forced in 1994-5 to consider sacrificing other policy goals on the altar of the exchange rate when the dollar declined precipitously against the yen. Without realizing it, many observers have derived an impossibility theorem: neither pegging like Sweden, nor occasionally realigning like Mexico and the EMS countries, nor floating like the United States is a tolerable option. Policy-makers seem to retain no acceptable international monetary alternative.38
How serious is the problem of currency crisis? Again, Eichengreen, Rose, and Wyplosz shed light on this question.39 Twenty OECD (Organization for Economic Cooperation and Development) nations were studied for the period 1959-1993 using quarterly data.40 The data included 2,516 observations of "tranquillity" (no significant change in market conditions), 61 failed speculative attacks, 81 instances of devaluation, 20 instances of revaluation (opposite of devaluation), 33 regime changes from fixed to floating exchange rates, 33 regime changes from floating to fixed, and 56 assorted nontranquil "other events."
The data on exchange crises and events are especially interesting because the theory can easily become so confusing. This chapter has focused on basic theories and case studies of currency crisis. Basically, theory suggests that currency crises should be happening all the time. The surprising event these days, from the standpoint of currency theory, is not the appearance but rather the lack of a currency crisis. The challenge for economists, therefore, is to try to figure out why some situations produce crises and speculative attacks whereas others do not. The search for answers to these questions has produced an amazing array of theoretical angles.
Some theories are based on purely economic problems, such as current account deficits, for example, whereas others focus on political problems and constraints, such as the need to create jobs, even in a surplus country, to get votes. Crises may be caused by a weak economy or a weak party or president. Some theories see currency problems as deriving from a nation's weak external economic position, whereas others seek their source in weak internal conditions. Some see currency crises as leading factors, which precede or anticipate a change in fundamental economic conditions, whereas other see these crises as lagging or following their real causes. Some theories focus on strategic behavior as the source of crisis, while others cite competitive forces. Traders and speculators in the private sector are the main actors in some accounts, whereas governments or central banks take the active role in others. Finally, some theories assume that foreign exchange markets have a single stable equilibrium, whereas others explore the possibility of multiple equilibria as the source of rapid exchange rate adjustments.41
Eichengreen, Rose, and Wyplosz concluded that "governments bring currency crises on themselves through the reckless pursuit of excessively expansionary policies" that create current account deficits, push the exchange rate down, and destroy the credibility of government policies.42 If this were the whole story, then we could conclude that currency crises were the just desserts of profligate states. It isn't, however, as the ERM crisis of 1992 showed. Some states bring crises on themselves,
But many other governments whose currencies are attacked do not clearly bring their exchange market difficulties on themselves through the reckless pursuit of expansionary policies. Virtuous behavior, in other words, is no guarantee of immunity from exchange market pressures.... Speculative attacks can occur because markets are uncertain about a government's intentions and test its resolve. Alternatively, speculative attacks can be a symptom of self-fulfilling attacks, in the sense that markets believe that the government will not resist pressure and will shift to more expansionary policies as it abandons the exchange rate commitment in response to the attack itself.43
The problem of currency crisis and speculative attack seems to be an integral part of the system of national currencies. Although states sometimes bring on these problems, it is not clear that they can entirely avoid them either. This does not mean, however, that politicians are entirely blameless. In Mexico, Britain, and recently in Thailand, politicians seeking to protect local interests all created or exacerbated situations in which speculators could make a "one-way bet" against their currencies and thereby encouraged unstable capital flows or speculative attacks.44 The international capital markets punish bad policies, as we have long suspected. Good policies, however, are not always rewarded.
Currency Crises and Globalization
Currency crises are a fact of life in today's international financial markets. Exchange rates are subject to a variety of forces that can cause them to swing sharply, suddenly, and by amounts that are hard to anticipate or hedge against. Some crises, such as Mexico's 1994 collapse, are the result of bubbles, manias, or panics—logical chain reactions gone astray. Other crises are like the ERM breakdown of 1992, the consequence of speculative attacks driven by the iron logic of profit maximization.
It would be one thing if we could say that currency crises are a curse on sinners, a plague on their houses for poor choices and unrealistic economic policy. But, as we have seen, currency crises are driven by their own logics, and it is not just sinners who are punished. The studies cited here suggest that we cannot truly know the sources of currency crises, their timing, their incidence, or their eventual total direct and contagion impacts. This is a lot to not know.
The persistence of currency crises means that exchange rates are subject to tremors that are like earthquakes. The ground shifts suddenly, by an unknown amount, at moments that are very hard to predict. This would matter little if the exchange rate were an insignificant item. Currency crises would then be like earthquakes in Antarctica—of interest only to penguins and academic researchers.
Unfortunately, the exchange rate is a price, and economists think that prices are very important. Prices are the invisible hands that guide the market system. They point to scarcity and to abundance, and they coordinate activities in the most efficient way. When prices are right, resource allocation is efficient. When they are wrong, the wrong signal is sent, and individuals acting in their own self-interest make mistakes. Getting prices right is therefore the most important thing that markets can do.
The exchange rate is the one price that matters most because it is the one price that most directly affects all other prices in a market system. When the U.S. dollar declines in price, for example, everything in the United States is cheaper for foreign buyers or investors, and everything in other countries is more expensive for U.S. citizens. Contagion effects from currency swings are therefore wide, if not also deep. The potential for resource misallocation is tremendous.
For globalization's advocates, the fact of currency crises creates a logical dilemma. If currency crises are unimportant, then prices are unimportant, which means that markets are unimportant. But markets are the driving force of globalization. If currency crises are important, on the other hand, then they create a risk that is associated with global or international activities and that naturally limits the extent of these activities. True globalization, in the sense of a single market operating according to the law of one price, is impossible. Global can only be international or multilocal in this context.
Currency crises are a significant barrier to globalization, but the situation is worse. Global financial markets suffer from both crisis and the potentially more serious property of chaos, which is the subject to which we now turn.
Endnotes
Note 1: Quoted by Moisés Naím, "Latin America the Morning After," Foreign Affairs 74:4 (July/August 1995), p. 45. García Márquez here is speaking to Carlos Fuentes, suggesting they dump their fiction into the sea because reality has outdone their imaginations. Back.
Note 2: This is a rate that Rudiger Dornbusch and others have argued made the peso overvalued. Dornbusch called for Mexico to devalue the peso as early as November 1992. It is significant that Finance Minister Pedro Aspe was a student of Dornbusch when he studied at MIT. Back.
Note 3: This point is made by Jeffrey Sachs, Aaron Tornell, and Andrés Valasco, "The Collapse of the Mexican Peso: What Have We Learned?" Economic Policy 22 (April 1996), pp. 13-64. Back.
Note 4: Zanny Minton-Beddoes, "A Survey of Latin American Finance," The Economist (December 9, 1995), p. 1. Back.
Note 5: Christopher Wood, "A Survey of Mexico," The Economist (February 13, 1993), p. 1. Back.
Note 6: Stephen Fidler, "Survey of Latin American Finance," Financial Times, March 29, 1993, p. 2. Back.
Note 8: Stephen Fidler, "Survey of Latin American Finance," Financial Times, April 11, 1994, p. 2. Back.
Note 9: Zanny Minton-Beddoes, "A Survey of Latin American Finance," The Economist (December 9, 1995), p. 1. Back.
Note 10: The particular problems of a single currency are discussed in Chapter 6. Back.
Note 11: Charles P. Kindleberger, Manias, Panics, and Crashes: A History of Financial Crises (New York: Basic Books, 1978). Back.
Note 12: Walter Bagehot as quoted in Rudiger Dornbusch's, "International Financial Crises," in The Risk of Economic Crisis, ed. Martin Feldstein (Chicago: University of Chicago Press, 1991), p. 117. Back.
Note 13: Walter Bagehot, "What a Panic Is and How It Might Be Mitigated," The Economist (May 12, 1866). Reprinted in The Collected Works of Walter Bagehot, vol. 10, ed. Norman St. John-Stevas (London: The Economist, 1978), pp. 88-89. Back.
Note 14: Paul Krugman, "Financial Crises in the International Economy," in The Risk of Economic Crisis, ed. Martin Feldstein (Chicago: University of Chicago Press, 1991), p. 100. Back.
Note 15: Paul Krugman, "Dutch Tulips and Emerging Markets," Foreign Affairs, 74:4 (July/August 1995), pp. 36-37. Back.
Note 16: Quoted in Naím, "Latin America the Morning After," p. 51. Back.
Note 17: Krugman, "Dutch Tulips," p. 39. Back.
Note 18: Lesley Crawford, "Survey of Latin American Finance and Investment: Only Zedillo Optimistic," Financial Times, March 25, 1996, p. 4. Back.
Note 19: Krugman, "Dutch Tulips," p. 43. Back.
Note 20: Naím, "Latin America the Morning After," p. 51. Back.
Note 21: Kindleberger, Manias, p. 219. Back.
Note 22: Krugman, "Dutch Tulips," p. 28. Back.
Note 23: Naím, "Latin America the Morning After," pp. 49-50. Back.
Note 24: This point is driven home by the success Mexico achieved in meeting its obligations once the crisis had passed. Back.
Note 25: See the discussion in Krugman, "Financial Crises." Back.
Note 26: Kindleberger, Manias, p. 220. Back.
Note 27: This term seems to have been coined by Paul R. Krugman in his 1979 article "A Model of Balance of Payments Crises," reprinted in Paul R. Krugman, Currencies and Crises (Cambridge, MA: MIT Press, 1992), pp. 61-75. Back.
Note 28: Exchange rate stability is only partial because, even though a single currency would fix exchange rates within the European Union, the euro would still vary with respect to the dollar, yen, and other currencies. Back.
Note 29: The decision to exchange ostmarks, the East German currency, for deutschmarks at a rate of one to one, which was far above the market rate for the currency, had the effect of creating substantial purchasing power in eastern Germany without a commensurate rise in output (in fact, output fell). More purchasing power chasing fewer goods and services is a sure recipe for inflation. Back.
Note 30: Paul De Grauwe, The Economics of Monetary Integration (2nd ed.) (New York: Oxford University Press, 1994), p. 124. Back.
Note 31: Sweden, also under attack, raised overnight interest rates to 500 (five hundred!) percent. Back.
Note 32: See Mark P. Taylor, "The Economics of Exchange Rates," Journal of Economic Literature 33 (March 1995), especially pp. 37-39, for an introduction to this literature. Back.
Note 33: Paul De Grauwe, The Economics of Monetary Integration, p. 123. Back.
Note 34: Barry Eichengreen, Andrew K. Rose, and Charles Wyplosz, "Contagious Currency Crises," National Bureau of Economic Research Working Paper no. 5681, July 1996. Back.
Note 36: This idea is discussed in Barry Eichengreen, Andrew K. Rose, and Charles Wyplosz, "Exchange Market Mayhem: The Antecedents and Aftermaths of Speculative Attacks," Economic Policy 21 (October 1995), especially pp. 294-296. Back.
Note 37: This point is made by De Grauwe in The Economics of Monetary Integration, Chapter 5. Back.
Note 38: Eichengreen et al., "Exchange Market Mayhem, pp. 251-252. Back.
Note 40: The authors correctly point out that quarterly data are inconvenient for this purpose, since as much exchange rate variation occurs within quarters as between them. If anything, then, the use of quarterly data is likely to diminish the measured turbulence in exchange rates. Nonetheless, these data are useful and revealing. Back.
Note 41: The possibility of multiple equilibria is well known in foreign exchange economics. It is possible for current demand and supply curves to bend sharply and even slope the wrong way owing to elasticity conditions for the underlying goods, services, and assets, creating a number of equilibria—some stable and others unstable. Back.
Note 42: Eichengreen et al., "Exchange Market Mayhem," p. 294. Back.
Note 43: Ibid., p. 294-295. Back.
Note 44: Thanks to an anonymous reviewer for this point. Back.