Economists at the University of Munich got subjects to trade an artificial asset which paid a dividend at the end of each of 10 trading periods and then became worthless; the advantage of such experimental assets is that they have a clear intrinsic value which can be compared with its actual prices to discover whether market pricing is correct or not.
However, before trading began some subjects were set a test designed to reduce their self-control. The idea here is that our capacity for self-discipline is limited so if we exercise a lot of it in one job we'll have less of it in another context: anyone who's got drunk after a day's hard work will know this.
The economists found that markets where traders had less self-control saw "significantly higher levels of overpricing" than those in which traders had more discipline. This was because traders lacked the discipline to sell overpriced assets.
However, although traders with low self-control earned lower profits, the difference wasn't so great as to drive them out of the market. This was because in trying to exploit the irrational ill-disciplined traders the ones who had more self-control themselves bought overpriced assets in the hope they'd become even more overpriced. This corroborates lots of other evidence which shows that - contrary to simple-minded theories of efficient markets - market forces don't swiftly eliminate irrational traders.
What we have here, therefore, is laboratory evidence to corroborate the Keynesian view that assets can become mispriced because traders' animal spirits sometimes overwhelm their self-discipline.
Does this finding have external validity - truth outside the laboratory?
I suspect so. We know - some of us from personal experience - that hormones can sometimes overwhelm self-control. Researchers at the University of Leicester show that increases in testosterone levels lead to increased risk-seeking in financial markets (let alone in other contexts).
This is not always a mere idiosyncrasy, with some traders having high testosterone and others low. It can be systematic. Alex Krumer at the University of St Gallen shows that men (but not women) are prone to psychological momentum; past successes increase testosterone and willingness to take risks. This is consistent with bubbles feeding on themselves; rising prices and recent profits increase risk-taking behaviour which pushes the prices of already overpriced assets up further. Traders, in effect, gamble more with the house's money.
Also, testosterone increases as the amount of sunlight rises. This is consistent with the tendency for share prices to rise in the spring and fall thereafter: in effect, a lack of self-control in the spring creates mini-bubbles.
It's a cliche that markets are driven by greed and fear. Often, these sentiments are held in check by self-control. Sometimes, though, they are not.
