Tuesday, April 15, 2008

Traders on Steroids

There is an interesting op-ed in the Financial Times today by John Coates, a former trader at Deutsche Bank who is currently a research fellow at Cambridge University. Coates monitored the testosterone and cortisol levels of a group of male traders in London, and reports some interesting (though unsurprising) findings. Testosterone, which increases confidence and risk-taking, "surges in males as they prepare for a competition, and continues to rise in the winner while falling in the loser." Cortisol, on the other hand, "promotes an anticipatory arousal" when reacting to challenges, particularly "under conditions of novelty and uncertainty." In short, testosterone leads to excessive risk-taking, and cortisol leads to excessive risk-aversion. Here are Coates' findings:

"We found that a trader’s daily testosterone levels were indeed higher when he made an above average profit. We also found that the higher a trader’s morning testosterone, the more money he made that day. This effect was most pronounced in experienced traders. ... We found that cortisol rose with the variance of the trader’s profits. It also rose in lockstep with implied volatility in the options market, raising the intriguing possibility of a biological substrate for the derivatives market. Cortisol was preparing traders for an impending market move; like testosterone, it was highest and most volatile in experienced traders. ... "If testosterone continued to rise it could, by fostering over-confidence and risk-seeking, lead traders to make irresponsible trades. Testosterone is likely to rise in a bull market, increase risk and exaggerate the rally. Chronic cortisol exposure, on the other hand, promotes feelings of anxiety, a selective recall of disturbing memories and a tendency to find danger where none exists. Cortisol is likely to rise in a crash, make traders dramatically and perhaps irrationally risk-averse, and exaggerate the sell-off."
Here's my question: Which way does the causation run? Do high testosterone levels cause traders to take excessive risks, or does excessive risk-taking cause high testosterone levels, which then take over and cause even more risk-taking? It's clear from the op-ed that high testosterone levels cause traders to take excessive risks, but the key question is what induces the high testosterone levels in the first place. If some traders simply have higher testosterone levels than others, then we could identify the irrationally risk-loving traders ex ante. The same goes for cortisol: if cortisol causes excessive risk-aversion, and some traders simply have higher cortisol levels than others, then we could identify the irrationally risk-averse traders ex ante. And here's another question: Does a trader who experiences high testosterone levels also experience high cortisol levels? Or are traders either risk-loving or risk-averse, with no in between? That traders are irrational is neither new nor surprising. But more information on the psychological aspect of asset bubbles is certainly welcome as we wade our way through the wreckage of a popped housing bubble.