If You Think Irishmen Are Red Haired You Won't See the Next Financial Crisis Coming

In standard economic theory financial crises are seen as market responses to catastrophic but very unlikely events, sometimes referred to as black swans. Historically, however, financial crises have occured quite often, particularly after periods of prolonged debt growth. How can crises be at the same time systematic and yet catch investors by surprise?

Nicola Gennaioli (Department of Finance) and co-authors Andrei Shleifer (Harvard University) and Robert Vishny (University of Chicago) in Neglected Risk: The Psychology of Financial Crises (in American Economic Review: Papers & Proceedings 2015, 105(5): 310–314, doi: 10.1257/aer.p20151091) trace this phenomenon to the psychology of investor beliefs, and in particular to the mechanism of representativeness, first highlighted by psychologists Kahneman and Tversky.

"We model representativeness as inducing people to overestimate the probability of outcomes that are only relatively more likely", Professor Gennaioli explains. "For instance, why do we think that many Irishmen are red-haired? The reason is that red hair is more common among Irishmen than among other people. While only one-hundredth of the world population is red-haired, one tenth of the Irish population is. In absolute terms, red hair is uncommon among the Irish too, but we make a mistake because we mix up relative and absolute numerosity."

In this paper, which is only the beginning of a large project studying the role of investor psychology, the authors show that this intuition naturally yields boom-bust financial cycles. After a series of good financial news the probability of other good news is overestimated, even if it's only relatively more likely in light of recently observed data. Investors are too optimistic and debt grows. A few, intermixed bad news do not change investors' mood, because good outcomes are still relatively more likely. Debt continues to grow despite early warnings. But when a string of bad news occurs, the bad outcome becomes sufficiently more likely that the representative scenario changes from boom to bust, leading investors to overreact. There is a liquidity crunch and a panic, which is reinforced by the fact that previous debt growth was excessive. "Bad news episodes", the authors write, "are dangerous because they very quickly change representation, not because of the objective (and unlikely) consequences they bring about".

Financial crises can thus be systematic and yet catch markets by surprise because investors overreact to recent good news, and hence fail to recognize similarities among the different pre-crisis bubbles. Explaining boom and bust cycles purely through volatility in expectations has important policy implications. "Financial innovation", Gennaioli says, "is in principle desirable, but it can lead to bad outcomes if psychological factors cause investors to neglect the risks of new securities. Financial authorities, moreover, should certainly promote financial literacy policies, but they should also pay attention to increasingly available data on the systematic errors embodied in investor beliefs. This data may help monetary policy to stabilize markets against the vagaries of investor sentiment".