Jason Zweig’s Intelligent Investor column this week strives to debunk the idea that “[i]f you watch carefully for signs of euphoria, you can sidestep the damage when markets go mad.” In other words, Zweig argues that you can’t reliably predict a financial bubble or even spot a bubble when you’re in one. Zweig gets support from Yale’s Robert Shiller, who correctly forecast both the internet and real estate bubbles: “being right about past bubbles does not automatically ensure that you will be right about the next,” partly because of our inherent overconfidence and partly because the markets are so complex and variable. I have written about the general subject here.
But an academic paper published this month in the SIAM Journal on Financial Mathematics addresses just this issue and claims to have broken the code. In it, the authors seek to answer the question of whether the price increase of a particular asset represents a bubble in real-time by using sophisticated mathematical models. “[W]e show that by using tick data and some statistical techniques, one is able to tell with a large degree of certainty, whether or not a given financial asset (or group of assets) is undergoing bubble pricing,” says one of the authors.
ScienceBlog explains the approach better than I can:
“The key characteristic in determining a bubble is the volatility of an asset’s price, which, in the case of bubbles is very high. The authors estimate the volatility by applying state of the art estimators to real-time tick price data for a given stock. They then obtain the best possible extension of this data for large values using a technique called Reproducing Kernel Hilbert Spaces (RKHS), which is a widely used method for statistical learning.”
Whether this new approach will work remains to be seen, of course, but a systematic, mathematically based approach has a far better chance to succeed than any approach that relies entirely upon our judgment. Our inherent limitations and biases make it too hard for us to make such judgments very often and very well.