Short-Run Momentum and Book-to-Market Effects

We present evidence from an event study that runs counter to the notion that the momentum and book-to-market (B/M) effects can be fully explained by time-varying risk premia. We minimize the joint hypothesis problem in market-efficiency tests by examining a relatively short (26-day) window that exhibits both momentum and reversal effects. There is return continuation during the first 17 days but sharp reversal during the last 9 days. The co-existence of strong momentum and reversal over this small event window rules out risk premium or chance as possible causes of momentum, leaving investor misvaluation as the only explanation for this anomaly during this period. Furthermore, several patterns of interaction between B/M and momentum also point to a behavioral interpretation of the B/M effect during the period. The general implication of our evidence is that investor behavioral biases are a necessary ingredient for the explanation of both the momentum and B/M anomalies.