Print this
Was the Asian Crisis a Contagion?

By Robert Hunter

The term “Asian flu,” bandied about freely in the financial press and elsewhere since late 1997, encapsulates the conventional wisdom on the Asian meltdown—that an economic “bug” spread from country to country, infecting otherwise healthy economies with weakened currencies, sudden credit defaults and extreme stock market volatility.

But now, several months removed from the nadir of the Asian crisis, some investigators wonder whether the Asian contagion was more perceived than real—and whether the very notion of financial contagion is merely a myth.

The chronology of the Asian crisis, on the surface, certainly seems to point to the existence of contagion. The Thai economy, like those of some of the other countries that produced the “Asian Miracle,” had experienced a long period of high domestic demand and trade deficits as its real exchange rate appreciated, along with overexposure to falling prices in its property market. The Thai government for years had contained fluctuations in the baht to a narrow band pegged against the U.S. dollar, largely via capital controls and aggressive forward intervention. Eventually, the pressures became too much to bear, and on July 2, 1997, Thai authorities abandoned the dollar peg. Since neighboring countries with similar economic fundamentals—primarily Indonesia, Malaysia and the Philippines—also pegged their currencies to the dollar, the baht devaluation forced them to accept more exchange rate flexibility, and their currencies marched downward virtually in lockstep with the baht. When Hong Kong and Singapore, countries with strong fiscal and current account positions, faced downward pressure on their currencies—and when the Korean won, which had staved off downward pressures for months, finally plunged in November 1997—the Asian flu seemed to be virulent and spreading. Credit dried up virtually overnight in Asia, and stock markets plunged worldwide, as the notion of equity market diversification seemed laughably out of date.

At no time has an international economic domino effect caused capital flight from countries with healthy economic fundamentals.

But not everyone agrees that contagion, as commonly understood, is possible, and the Asian meltdown has provided ample grist for all viewpoints on the nature of the term.

Information overload?

Most agree that information is at the center of the contagion debate. Graciela Kaminsky, a professor at George Washington University, and Sergio Schmukler, an economist at the World Bank, have examined the kinds of news events that spark the massive market moves characteristic of economic crises. In “What Triggers Market Jitters? A Chronicle of the Asian Crisis,” a World Bank white paper examining East Asia during 1997–98, they argue that market events are caused by both local-country and neighbor-country news. News relating to agreements with international organizations such as the World Bank had the biggest effects on Asian markets during the crisis, moving stock prices by an average of 10 percent to 11 percent—usually upward. News on credit-rating changes had the next-biggest effect on Asian markets, moving them down by an average of 10 percent on days when credit downgrades were announced. News on political issues, capital controls and the economy moved the Asian stock markets by about 7 percent, while monetary policy news moved markets by 3 percent. On average, days in which no news was released witnessed market downturns of 2.5 percent.

Kaminsky and Schmukler also note that good news has a smaller impact on stock prices than bad news, which suggests that “investors may be overreacting to bad news” and may be exhibiting “herding behavior.” Moreover, they assert that “the daily swings in financial markets in days of no news become more pronounced as the crisis deepens, with moody investors driving down prices even in the absence of any fundamental news.”

In a working paper for the International Monetary Fund titled “Financial Market Contagion in the Asian Crisis,” Taimur Baig and Ilan Goldfajn expand on the findings of Kaminsky and Schmukler. First, they examine the correlation between the foreign exchange, equity, interest rate and sovereign debt markets of the various countries. Next, they determine whether the correlations increased significantly during the crisis period compared with historical, “tranquil” periods. When there was no significant increase, they argue, market pressures were mostly a result of a common cause or various spillover effects. But when correlation increased substantially, there was “reason to suspect that sentiments [had] shifted.” They also examine the impact on the stock markets and exchange rates of the countries in question.

As the table below indicates, there were some notable correlations between the various markets between July 2, 1997, and May 18, 1998. On the whole, they found that “there was a clear case of increased correlations in the currency markets [but] this comes with the caveat that the currencies’ movements were minimal prior to the crises due to the existence of [currency] pegs. The evidence is not very clear in the case of the equity markets and the domestic call rates. The spreads on dollar-denominated debt, representing pure default risk, display the most striking degree of correlations and evidence of contagion.”

A country that falls may serve simply as a “wake-up call” to investors, who suddenly examine the fundamentals of other countries.

Baig and Goldfajn examined the impact of news on markets in Indonesia, Thailand, Malaysia, the Philippines, Korea and Japan by performing regression analyses. Good news included successful formation of bailout arrangements; announcement of rescue packages by international organizations; better-than-expected economic news; and specific measures to stabilize the markets. Bad news included collapse of currency regimes or of long-standing financial arrangements; breakdowns in negotiation with multilateral agencies; large-scale bankruptcies or firm closures; credit rating downgrades; worse-than-expected announcements about debt exposures, inflation or growth prospects; confusing policy moves; threats or announcements of capital controls imposition; resignations or firings of high-profile officials; and civil unrest. They also added two other variables—the daily stock market return in the United States (Standard & Poor’s 500) and the yen-dollar exchange rate.

Asian Correlations: A Summary
The Asian markets were correlated in the following ways from July 2, 1997 to May 18, 1998:

Exchange Rates:

  • From October 1997 onward, correlations between Indonesia and Korea increased substantially.
  • Rolling correlations revealed high volatility.
  • Indonesia and Malaysia were consistently correlated.
  • In general, the Korean won was uncorrelated with the rest of the countries (exception: Indonesia).

Stock Markets:

  • The Malaysian and Indonesian stock markets had the highest degree of correlation among all the pairs.
  • In general, correlations were greater than those in the currency markets.

Interest rates:

  • With the exception of the Philippines, all the countries had positive correlations.

Sovereign Spreads:

  • From July 1997 to December 1997, the Philippines spreads were strongly correlated with those of Thailand, Korea and Indonesia.

Source: IMF working paper, “Financial Market Contagion in the Asian Crisis,” by Taimur Baig and Ilan Goldfajn.

Although the authors found that stock market regressions were weaker than currency market regressions, they concluded that “during a period of financial market instability, market participants tend to move together across a range of countries. Shocks originating from one market readily get transmitted to other markets, thus becoming a source of substantial instability. The evidence of contagion in the foreign debt markets reinforces the view that there was an element of financial panic at the onset of the Asian crisis.”


Michael Bordo, an economic historian at Rutgers, asserts that there has never been a true financial contagion: At no time has an international economic domino effect caused capital flight from countries with healthy economic fundamentals. Others have argued similarly that only countries skating on economic thin ice—with a deteriorating current account, a slowdown in growth or deeper recession, bursting real estate and stock price bubbles, and high short-term foreign debt—run the risk of sinking.

But some analysts have argued that a country’s fundamentals do not have to change significantly for a crisis to strike. A country that falls may serve simply as a “wake-up call” to investors, who suddenly examine the fundamentals of other countries and discover, to their chagrin, that they are similar to that of the fallen country. This “demonstration effect” causes investors to flee markets with poor fundamentals, spreading the crisis. Thus in the strictest sense the situation cannot be described as contagion, since the newly stricken countries had already been sick—they had been harboring a low-level economic infection for some time.

The demonstration effect was not a factor in the case of Asia, says Paul Masson in an IMF working paper titled “Contagion: Monsoonal Effects, Spillovers and Jumps Between Multiple Equilibria.” He writes: “The fact that an optimistic view about East Asian economies prevailed for so long (in the face of some reports of banking sector problems), the rapidity of the change in view, and the suddenness and severity of the resulting crisis all argue in favor of the multiple equilibrium story.”

Masson asserts that under certain conditions, changes in the expectations of investors can bring about a change in the equilibrium, thus validating the expectations. For instance, if all investors think that the central bank will devalue, they will attack the currency, provoking a devaluation, while if they think that the currency will remain stable, it’s likely that it will. But only in certain models, and for a certain range of fundamentals, are such self-fulfilling expectations possible. If financial authorities have a lot of reserves, for example, they can defend the currency successfully, even if investors think they can’t. Or, if they aren’t worried about the effect of high interest rates, they can use that instrument to defend the currency. But if reserves are low, or if the central bank cannot jack up interest rates because the banking system is in bad shape, then the possibility of multiple equilibria—and hence contagion—may exist.

Are they all merely splitting hairs, as academics are wont to do? Perhaps. But parsing the definition of contagion and presenting the concepts of herding effects, demonstration effects and multiple equilibria can only help in preventing another regional illness from spreading.

For a copy of the reports, see www.imf.org and www.worldbank.org.

Learning From the Summer of ’98

August and September put risk managers through a torture test. The Russian sovereign defaults resulted in mega losses throughout the financial world, and leveraged positions linked to those exposures only exaggerated the outcomes.

According to a recent survey of risk managers by Capital Market Risk Advisors, 28 percent of respondents admitted that their understanding of the risks involved was not as complete as they desired it to be. The risk management process failed to identify the potential of hedges to fail under defaults, the limited ability to measure spread risk and the degree of leverage undertaken. Respondents also said the most common losses involved the emerging markets, Russia, credit and currency spreads, credit events, and directional trades.

Risk managers said they were particularly concerned about issues involving liquidity, volatility, widening credit spreads, emerging market exposure and settlement risk. To address these concerns, some respondents said they changed the frequency and quality of communication and risk reporting, improved the mechanisms used to measure spread exposure, and began deleveraging positions.

CMRA also found what it believed were significant gaps in stress-testing disciplines. While most respondents claimed to stress-test against parallel market moves, only 40 percent of banks and 60 percent of broker-dealers said they stress-tested parallel shifts in the volatility curve. Neither group reported stress-testing twists in volatility curves and only 30 percent of banks and 20 percent of broker-dealers said they stress-tested correlations during the period.

Leslie Rahl, a principal at CMRA, says she isn’t particularly surprised at the survey’s results. “The only surprise is that we have to keep learning the same lessons. Every time there’s a crisis, we learn that correlations go out the window. And every time the crisis goes away, everyone goes back to looking at historical correlations. I think there’s a new awareness that our lives are probably going to be an endless series of crises. We have to build into risk management more focus on stress in crisis situations where all markets become correlated and bid-offer spreads widen—the things we all know occur, but don’t necessarily build into our processes.”

For more information about the survey, see www.cmra.com.

Was this information valuable?
Subscribe to Derivatives Strategy by clicking here!