Professor François Longin tracks down extreme events in finance

François Longin, Professor of Finance at ESSEC Business School, has published a book co-authored by more than 40 contributors worldwide, entitled “Extreme Events in Finance: A Handbook of Extreme Value Theory and its Applications”. For several years now, the research professor has used “extreme value theory” to improve the measurement of market risks.

The New York stock market crash in 1929, the Y2K internet bubble, the sub primes crisis of 2007… History is scattered with financial and monetary crises having to varying extents strongly impacted the world economy. François Longin has made extreme events the basis of his research work. Since 1994, he has applied extreme value theory to a financial context to measure risks on financial markets. However, François Longin was not especially predestined to become interested in this aspect of the financial world. It took the coming together of two events during his studies to finally end up at this subject of interest. A student at the École Nationale des Ponts et Chaussées, in 1989 he followed a scientific modeling course which he instantly found passionate. ‘This course presented extreme value theory,’ explains François Longin. ‘We used it to define the size of constructions that had to resist extreme events such as earthquakes or tornados. A lot of Dutch researchers were interested in this approach to calculate the dimensions of dykes following the 1953 storm tide that led to nearly two thousand people losing their lives.’   

Several months before studying this theory, he had been confronted with, indeed like the entire world, a stock market crash. The 19^th October 1987, the New York stock exchange lost 20% in only one day. ‘People couldn’t understand how such an event happened,’ explains François Longin. ‘This sparked my interest in studying extreme events in finance.’ François Longin put two and two together and came up with the idea to apply extreme value theory to the world of finance. The bug was never to leave him. In 1993, he published a thesis on the subject. ‘Nobody had done that before. I had to convince more than one. But the timing was right because regulations had become very important in finance and, in the 1990s, people were looking to quantify the risks on the market.’

The possibility, not to predict but to measure risks, appealed to people. ‘At a certain moment in finance, you have to model things. The raw material in this field is risk. It was therefore necessary to calculate it.’ François Longin uses large databases to come up with his statistics. The researcher even dissects semantics in the archives of American newspapers following strong drops in the stock market to obtain a clearer definition of extreme events. Today, the application of extreme value theory does not enable the exact forecasting of the next crash, but its period of return or the probability of its size.An important breakthrough: according to the usual statistics used, an event such as the 1987 should reoccur every billion-billion years or so. By applying extreme value theory, this timeframe become ‘only’ 100 years.

In 2014, François Longin notably brought together numerous experts and researchers on the subject in a conference given at ESSEC on the theme Extreme Events in Finance. Today, he also carries out consulting activities with banks regarding extreme values, market operations, financial regulation and asset management. And, it goes without saying, he is listened to. ‘For a long time, extreme events were rejected by researchers who considered them as anomalies in the data.’ We can, however, deduct many things from them – they provide enormous amounts of information. If people see an interest, they listen. Because if the magic word in asset management is “tax savings”, in finance it’s “stock market crash”.’

Consult François Longin’s articles on ESSEC Knowledge