A Theory of Overconfidence, Self-Attribution, and Security Market Under- and Over-reactions

We propose a theory based on investor overconfidence and biased self- attribution to explain several of the securities returns patterns that seem anomalous from the perspective of efficient markets with rational investors. The theory is based on two premises derived from evidence in psychological studies. The first is that individuals are overconfident about their ability to evaluate securities, in the sense that they overestimate the precision of their private information signals. The second is that investors' confidence changes in a biased fashion as a function of their decision outcomes. The first premise implies overreaction to private information arrival and underreaction to public information arrival. This is consistent with (1) post-corporate event and post-earnings announcement stock price 'drift', (2) negative long- lag autocorrelations (long-run 'overreaction'), and (3) excess volatility of asset prices. Adding the second premise leads to (4) positive short-lag autocorrelations ('momentum'), and (5) short-run post-earnings announcement 'drift,' and negative correlation between future stock returns and long-term measures of past accounting performance. The model also offers several untested empirical implications and implications for corporate financial policy.