Once they have observed information, hindsight biased agents fail to remember how ignorant they were initially. This bias has been amply documented by the psychological literature. To study its economic consequences, we formulate a theoretical model of the hindsight bias, providing a foundation for previous empirical measures. We show theoretically that hindsight biased agents overreact to information signals, conduct incorrect learning about volatility, and engage in suboptimal investment decisions. To test empirically the hypothesis that the hindsight bias reduces performance in important economic situations, we collect psychometric and performance data from a sample of investment bankers in London and Frankfurt. We find strong evidence that these agents are hindsight biased, and that more biased agents have lower performance. These findings are robust to differences in location, information and experience. Many thanks for thoughtful comments to participants in the Toulouse and Amsterdam Universities seminars, the Warwick Conference on Behavioral Finance (2006) and the Munich Summer School on Economics and Psychology (2006), in particular Denis Hilton, Luis Garicano, Robin Hogarth, Florencio Lopez de Silanes, Thomas Mariotti, Paola sapienza, Jean Tirole and Luigi Zingales.
Stored only when both papers are in the corpus — 17 citations exist upstream.