The Second Leg Down

Chapter 8“Flash Crashes”, Crises and the Limits of Prediction

We have spoken at length about hedging after a drop and even ventured into offensive strategies for a high volatility market. But is it possible to anticipate sharp sell-offs or longer-term crises? The words “predict” and “forecast” are a bit more loaded. These words bring to mind the admonition that the goal of economic forecasting is to lend credence to astrology (1984). Nevertheless, we will wade into the muck and see whether it is possible to say anything about the likelihood of a quick correction or longer-term bear market. We informally describe how a flash crash can develop, using an analogy from mathematical biology. There are attempts to predict crises in the recent literature. We discuss some of these approaches, ending with a cautionary note about the limits of forecasting in financial markets.

LORD OF THE FIREFLIES

Many investors (especially those influenced by Taleb, 2007) vigorously argue that market crises are inherently unpredictable. Portfolio managers sometimes feel that they are operating in a sea of randomness, with a low signal to noise ratio for even their best ideas and systems. Large-scale liquidations are particularly baffling. If conditions had been even slightly different, they may never have occurred. In this context, the claim in Sornette (2009) that extreme moves can be predicted and possibly controlled seems remarkable. It's like trying to solve the hardest problem in a scientific field ...