Why Investors Stampede

If every trader and investor remembered that aphorism, the financial markets would be far different. But time, experience, and information haven't made people any the wiser. Whatever the spark, when the markets ignite, a flash fire can still easily turn into a firestorm. Traveling up or heading down, the markets habitually take on a momentum of their own, and prices end up "overshooting," or reaching extreme highs or lows before settling back. It's as true today as it was when the stock market crashed in 1987--or 1929.

The panic over Mexico is a case in point. For the past few years, money surged into Mexico, buoying the bolsa and helping to finance short-term borrowing. Political unrest and the assassination of a presidential candidate in 1994 barely dimmed investor enthusiasm. Then, a surprise yearend devaluation stunned the markets. In the following weeks, the peso plunged, dragging the value of Mexican stocks and bonds down. Only a U.S.-organized rescue appeared to stop the carnage. It may be some months before it's clear to what extent prices overshot, but in a few weeks, the peso had lost more than 50% of its value against the dollar, putting it well below purchasing-power parity. How could this happen? How could Mexico go from being an investor darling to dog so quickly--and carry so many other emerging markets down with it?

Easily. At times like these, the markets are propelled largely by psychology. As bandwagons form, investors' herd instincts frequently push prices far beyond where they ought to go. Yet psychology can turn on a dime as new information er news events are absorbed, sending prices off in the opposite direction. This is even truer during a currency crisis than during a market crash, because foreign investors who face massive exchange losses are frequently hostages to indecisive policymakers.

Speculative bubbles and crises of confidence are not new, of course. But the potential for financial and economic dislocation is greater than ever. That's because the growth, sophistication, and leveraging of financial markets around the globe has enabled investor psychology to have swift and powerful effects. A wide array of new securities, from collateralized mortgage obligations to emerging market debt, beckons investors. Pension funds and mutual funds have ventured into new worlds of investment. At the same time, high-speed transmission of information and trades makes it easy for hedge-fund managers and pensioners alike to move in and out of markets at will. In today's global markets, there are no fire walls.

CASCADES. Investor behavior is still a murky area for economic research. By now, though, economists and psychologists have concluded that the most important factor influencing investor behavior is the recent past--recent news, recent earnings or, for that matter, recent trends in the markets. That means earnings prospects or economic fundamentals count for little, while "this week's bad news carries far too much weight," says Richard H. Thaler, economist at Cornell University.

The tendency to see the recent past as prologue is amply demonstrated in surveys of market participants' views before, during, and after the crash of the Nikkei 225 at yearend 1989. Reporting the results in a recent paper, Robert Shiller of Yale University and co-authors Fumiko Kon-Ya of the Japan Securities Research Institute and Yoshiro Tsutsui of Osaka University found that large numbers of investors ignored warnings that the market was becoming more and more overvalued.

Indeed, just before the crash, 43% of the Japanese respondents thought that if prices dropped 3% in one day, the market would nonetheless rise the next day. "This impression of stability for the market may have encouraged the high prices that the Nikkei reached just before the crash," the study says.

The Nikkei, much like the Dow two years before, crashed for no apparent reason--there was no major news event or economic occurrence that precipitated the break in prices or that could adequately explain it. But what economists call "informational cascades" may explain both how investors pile on when markets gain speculative momentum and how bubbles can burst.

Economists Ivo Welch, Sushil Bikhchandani, and David A. Hirshleifer of the University of California at Los Angeles argue that people typically observe the behavior of their peers, whether they are choosing a restaurant or an investment. Even if their own judgment may run contrary, when they learn that their peers favor something, they tend to go with the herd and justify that decision by reasoning that the weight of opinion must be correct. This leads to an informational cascade, with more people observing and following each other. Such lemming-like behavior is understandable. It takes an awfully strong ego to stand one's ground against a tide of opposing opinion. "No one is acting stupidly," says Welch.

Despite their apparent power, these cascades can shift suddenly in the opposite direction with just a few pieces of information--or because just a few people decide to change course. In fact, a look back at the period just before the 1987 crash in the Dow indicates that money in the options market was anticipating a sharp decline. Similarly, when the dollar was still climbing in 1984-85, six-month forward rates pointed to a dollar decline, even though one-month forwards showed strength.

It's generally the smart money, the so-called "stabilizing speculators," who step in when prices have been driven to unexpected troughs or who exit soaring markets before outrageous peaks are scaled. In the case of Mexico, such investors were scarce--too scarce to set off cascades. One small hedge-fund manager, figuring the markets had overreacted in early January, bought short-term dollar-denominated Mexican paper, or tesebonos, in the secondary market to yield 30%--a good bet, given that Mexicans sold such paper at auction with yields of 20% a week later. Some portfolio managers gingerly tested the waters with small purchases of American Depositary Receipts, which had gotten especially hammered. But it took a massive rescue program to turn things around.

By their very size and design, the markets have encouraged fickle behavior. Look at the ease with which capital has flowed across borders and pumped up the growth of new markets. Since 1989, says David D. Hale of Kemper Corp., private capital flows to emerging stock markets in Latin America, Asia, and Africa have expanded from less than $10 billion a year to between $60 billion and $80 billion a year. Meanwhile, the combined stock market capitalization of those regions now exceeds $2.1 trillion, says Hale, compared with only $400 billion 5 years ago and barely $100 billion 10 years ago. And when money flows to emerging markets through unregulated mutual funds rather than through regulated banks, says Albert M. Wojnilower, senior economic adviser at CS First Boston Investment Management Corp., that money is sure to be far more volatile.

"INVOLUNTARY DUMPING." So the January tumble in Mexican securities--and the contagion that spread to other Latin markets--was not merely a matter of psychology. What began as an understandable sell-off picked up speed because scores of portfolio managers at mutual funds and pension funds had no choice. They had to worry about near-term performance and about meeting investment criteria. And mutual fund managers had to worry about the prospect of massive redemptions. "There was a lot of involuntary dumping and forced selling," observes Robert W. Vishny, an economist at the University of Chicago.

In this age of near-perfect information and highly competitive markets, investors aren't supposed to overreact and prices aren't supposed to overshoot. Any "mispricing"--that is, any price that doesn't accurately reflect value--should be quickly traded away by investors seeking to capitalize on the market anomaly. But overshooting and overreaction continue to occur. Given human nature and the vast and sophisticated markets that people can play, that's not likely to change anytime soon.

When Markets Overshoot

Identifying overshooting while it occurs is difficult. But after the fact, periods of jubilation and panic are clearly identifiable. Here are some examples:


The U.S. dollar had been climbing since 1981, buoyed by high real rates of interest. Widening trade and budget deficits should have stalled its ascent, but speculative buyers pushed it higher until the Plaza Accord in September, 1985.


The U.S. bull market started in August, 1982, as the economy rebounded and corporate profits grew. Interest rates started rising in the spring of 1987, with long rates moving toward double digits, but stock prices kept right on climbing.


The Dow plunged 508 points on Oct. 19, 1987. Within

six months, however, prices had stabilized at year-earlier levels, and the market began a steady upward climb. By August, 1989, the Dow had returned to its pre-crash high.


The Nikkei 225 also started climbing in 1982 but withstood the shock from Wall Street. It ascended rapidly, reaching a peak at yearend, 1989. Prices tumbled over the next couple of years

and remain below the 1989 peak.


Mexico was buoyed by a surge of foreign investment in 1992 and 1993. Political unrest worried some investors in early 1994, but it was the surprise devaluation of the peso on Dec. 20, 1994, that spawned a panic.