Glossary
Overconfidence bias
Overestimating one's own knowledge, skill or the precision of one's forecasts, which in investors shows up mainly as trading too much.
Overconfidence is the tendency to believe one knows more, or predicts better, than one does. Psychologists distinguish several forms:
- Overprecision, or miscalibration: ranges that are too narrow. When people are asked for estimates they are 90% sure of, the true answer falls outside far more often than 10% of the time.
- Better-than-average: most drivers rate themselves above average, which cannot be true of most.
- Illusion of control: feeling that one's actions affect outcomes that are mainly chance.
In investing
The best-known evidence comes from Brad Barber and Terrance Odean. In a study of 66,465 US households with accounts at a discount broker from 1991 to 1996, published in 2000, the fifth that traded most earned 11.4% a year after costs, against 17.9% for the market. Their trades were not bad picks on average. The costs of trading so often consumed their returns. A 2001 study by the same authors found that men traded about 45% more than women, and that trading cut men's net returns more.
Other effects
- Concentrated positions, sized as if the forecast were certain.
- Underestimated risk, with too little margin for error.
- Ignoring base rates: believing one's own pick will beat the averages that most active funds fail to beat.
What helps
Writing down predictions with explicit probabilities and checking them later is the most direct test. Comparing results with a simple benchmark, after costs, is another.
Limits
Some confidence is needed to act at all, and studies find the effect varies a great deal between people. Measuring turnover and results is more useful than labeling oneself.
