The Death of the Overreaction Anomaly? A Multifactor Explanation of Contrarian Returns

Are the returns accruing to De Bondt and Thaler’s (1985) (DT) much celebrated overreaction anomaly pervasive? Using the CRSP data set used by for the period 1926 through 1982, and, for the first time, an additional two decades of data (1983 through 2003), we provide preliminary support for the original work of DT, reporting that the overreaction anomaly has not only persisted over the past twenty years but has increased, on a risk-unadjusted basis. However, using the three factor model of Fama and French (1993) (FF), we find no statistically significant alpha can be garnered via the overreaction anomaly, with contrarian returns driven by the factors of size and value, not the behavioral biases of investors. It is our conjecture that the anomaly is not robust under the FF framework, with ‘contrarian’ investors following such a scheme simply compensated for the inherent portfolio risk held.