Published online by Cambridge University Press: 06 April 2009
This paper addresses the bias associated with parametric measurement of timing skill based on monthly timer returns when timers can make daily timing decisions. Simulations suggest that the classic Henriksson-Merton parametric measure of timing skill is weak and biased downward when applied to the monthly returns of a daily timer. The paper proposes an adjustment that mitigates this problem without the need to collect daily timer returns. Four tests of timing skill, carried out on a sample of 558 mutual funds, show that very few funds exhibit statistically significant timing skill. More encompassing, the adjusted-FF3 test (based on the specification that incorporates both the proposed adjustment and the Fama-French three-factor model) is the least biased measure of timing skill among the four—it provides for a sharper inference regarding timing skill and helps mitigate biases associated with the choice of investment style.
All authors, Yale School of Management, 135 Prospect Street, Yale University, New Haven, CT 06510. We thank Don Chance (the referee) for several constructive suggestions. We also thank Steve Shellans and Ed Owens for useful discussions. We are grateful to Ken French, who graciously provided SMB and HML factor returns. Finally, we thank the 1999 WFA conference participants, especially the discussant Mark Grinblatt, for valuable comments.