News Release

Stock market crashes are predictable, major decline is coming in 2003 and 2004, says UCLA physicist

Peer-Reviewed Publication

University of California - Los Angeles

Successfully predicting stock market swings is as futile as searching for the fountain of youth, some people believe. UCLA physicist and complex-systems theorist Didier Sornette is not among them. Sornette, author of a new book, "Why Stock Markets Crash: Critical Events in Complex Financial Systems" (Princeton University Press), has found patterns that occur in market crashes dating back for centuries. Their statistical signatures are evident long in advance, he concludes.

Sornette has developed algorithms -- based on sophisticated mathematics, statistical modeling techniques and collective behavior theory -- that enable him to analyze more than two dozen stock markets worldwide. Applying techniques of physics to economic data, he has developed a quantitative model that can predict the signatures of a coming stock market crash.

"Economic forecasting is often not effective at predicting changes of direction, but our algorithms are very good at doing so," said Sornette, a professor of earth and space sciences, and a member of UCLA's Institute of Geophysics and Planetary Physics, who is also a research director at the University of Nice's National Center for Scientific Research in his native France.

Sornette cautions that his model allows him to make broad predictions, but is not able to predict where the stock market will be on any particular day or week. He disagrees with Alan Greenspan's Aug. 30, 2002, remark that it is "very difficult to definitively identify a bubble until after the fact," but Sornette is unable to predict whether the result of a particular bubble will be a rapid crash or a prolonged bear market.

Sornette and Wei-Xing Zhou, a postdoctoral scholar in his UCLA laboratory, present bad news for those who hope the bear market is over, in an article in the December issue of Quantitative Finance, a bimonthly journal focusing on the intersection of physics and economics.

"The U.S. stock market is not yet on the verge of recovery," Sornette said. "The bear market that started in July-August 2000 still has a long way to go."

Sornette and Zhou predict the Standard & Poor's 500 (currently above 900) will begin dropping by the second quarter of 2003 and will fall to approximately 700 in the first half of 2004.

The U.S. stock market since 1996 has shown a "remarkable similarity" with Japan's Nikkei index from 1985 to 1992, which may reflect deeper similarities between the fundamentals of the two economies, Sornette and Zhou argue. The S&P 500 has not yet entered a "second phase" of decline, as the Nikkei index did some two-and-a half years into its steep decline.

"Our theory is tailored to identify anomalies, bubbles and their end," Sornette said.

How accurate are Sornette's predictions?

His previous predictions have often been quite accurate, especially his January 1999 prediction (co-authored with Anders Johansen, his former UCLA post-doctoral scholar) that Japan's Nikkei index would rise 50 percent by the end of that year, at a time when other economic forecasters expected the Nikkei to continue to fall, and when Japan's economic indicators were declining. The Nikkei rose more than 49 percent during that time. He also successfully predicted several short-term changes of trends in the Nikkei.

"We have strong supporters, and others who say this is impossible," Sornette said. "Scientists typically do not predict the future, but I'm optimistic. Complex-systems theory is a young science, and the predictions will undoubtedly improve over the next five years. We are not able to predict stock markets with anything close to 100 percent accuracy, but I have confidence in the predictions, and confidence that they will become more accurate as we refine our methods."

How do Sornette's own investments do?

"My research takes all my time; I do not spend even one percent of my time investing in the market," he answered. "However, I have invested with associates, who implemented this system, in particular in hedge-funds. We have done well, and are continuing to do so. This system alone, however, is not sufficient to profit in the stock market with active trading, especially not in the short term and must be complemented with other analyses."

In "Why Stock Markets Crash," based on 10 years of research, Sornette proposes his theory of how, why and when stock markets crash. He uses complex-systems theory to dissect market crashes, and a new set of computational methods capable of searching and comparing patterns.

While most attempts to explain market failures search for triggering mechanisms in the hours, days or weeks before the collapse, Sornette argues that the underlying cause can be found months and even years before -- in the build-up of cooperative speculation. He provides a step-by-step analysis, using cutting-edge statistical modeling techniques, as well as insights from physics.

Sornette analyzes historical precedents, from the decade-long "tulip mania" in the Netherlands that began in 1585, a time of great prosperity, and wilted suddenly in 1637, to the South Sea Bubble that ended with the first huge market crash in England in 1720, to the bubbles and crashes that occurred every decade in the 19th century, to the Great Crash of October 1929 and Black Monday in October 1987. He analyzes herd behavior and the crowd effect, speculative bubbles, and precursory patterns before large crashes, as well as the major crashes that have occurred on the world's major stock markets.

Sornette concludes that most explanations other than "cooperative self-organization" fail to account for the subtle bubbles by which markets lay the foundation for catastrophe.

"Collective behavior theory predicts robust signatures of speculative phases of financial markets, both in accelerating bubbles, as well as decelerating 'antibubbles,'" Sornette said. "These precursory patterns have been documented for essentially all crashes on developed as well as emergent stock markets."

Sornette sees a series of stages, beginning with a market or sector that is successful, with strong fundamentals. Credit expands, and money flows more easily. (Near the peak of Japan's bubble in 1990, Japan's banks were lending money for real estate purchases at more than the value of the property, expecting the value to rise quickly.) As more money is available, prices rise. More investors are drawn in, and expectations for quick profits rise. The bubble expands, and then bursts. (From the early 1970s to 2000, Hong Kong's stock market accelerated and crashed eight times -- a perfect illustration of his theory.)

"Stock market crashes are often unforeseen by most people, especially economists," Sornette added. "One reason why predicting complex systems is difficult is that we have to look at the forest rather than the trees, and almost nobody does that. Our approach tries to avoid that trap. From the tulip mania, where tulips worth tens of thousands of dollars in present U.S. dollars became worthless a few months later, to the U.S. bubble in 2000, the same patterns occur over the centuries. Today we have electronic commerce, but fear and greed remain the same. Humans are endowed with basically the same qualities today as they were in the 17th century."

Sornette, 45, also conducts research on earthquake prediction, which he says is much more difficult than stock market prediction, but which he believes will also be possible. A specialist in the scientific prediction of catastrophes in a wide range of complex systems, he has written or co-written more than 250 papers in scholarly journals.


Sornette's predictions are posted on his Web site:

Note: Sornette, Zhou and UCLA do not assume responsibility for investment decisions made by others.

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.