OUR WORLD IS AN UNCERTAIN PLACE WHERE BIOLOGICAL SYSTEMS AND FINANCIAL MARKETS CAN COLLAPSE IN AN INSTANT.
POWERFUL PREDICTIVE MODELS FUELED BY SMARTER DATA SETS ARE THE TOOLS THAT WILL ALLOW US TO KNOW SOONER ANDADAPT MORE QUICKLY TO THE PROBLEMS THAT DEFINE OUR COMPLEX AGE.
On July 10, 2009, Swiss econophysicist Didier Sornette published a paper in the online journal arXiv.org bearing the provocative title, “The Chinese Equity Bubble: Ready to Burst.” The title implied that Sornette had accomplished a seemingly impossible feat: building a model of financial markets that was able to identify bubbles and predict when they would burst.
Sornette and his team asserted that they had found a bubble in the Shanghai Composite Index, and more boldly, that this bubble would end between July 17 and 27. Few outside of Sornette’s group saw reason to believe the prediction, because constructing a model of a financial market and making accurate predictions about its future behavior have been a long-sought holy grail of economics.
Perhaps not surprisingly, the 27th came and went, and the index continued to climb. Sornette seemed to have failed. But then on August 4 the market changed course. The index dropped sharply. Over the course of the next two weeks it fell almost 20 percent. Sornette’s prediction was correct.
There is still much debate as to whether Sornette’s team actually succeeded in forecasting the decline in the Shanghai Index. Some critics claim that the prediction itself may have influenced investor behavior and caused the market to drop, and many others simply question the idea that we are capable of making such accurate predictions at all about systems as complex as the Chinese stock market. But Sornette’s team is not alone. His is one of a number of pioneering groups of multidisciplinary researchers seeking to identify reliable, generic early warning signs that can be used to forecast the behavior of a wide variety of social, planetary, and biological systems.
This wasn’t Sornette’s first experience with the vagaries of a dynamic system. Before turning his attention to finance, Sornette studied rupture points in biological systems, which led to predicting earthquake eruptions. “But natural systems don’t fight back,” Sornette says of his decision to attempt predictive models of more complex systems. “In social systems such as markets, the theories become the engines that modify the structure of the system.” To date, his model stands out as one of the only potentially successful attempts to predict the behavior of a real-world complex system. But more may soon be on the way. Researchers have spent the past few decades laying the theoretical groundwork necessary to build powerful predictive models. Now smarter technologies capable of collecting and parsing more robust data sets than ever before may begin fueling these models with the information they need to address some of the world’s most pressing problems.