Measuring Portfolio Gains From Earnings Announcement Trading Signals

Abstract

This study generates out-of-sample predictions from training data to construct investment portfolios that are mean-variance optimized and rebalanced daily to assess gains from incorporating signals based on post-earnings announcement drift (PEAD), the earnings announcement premium (EAP), and firms’ rescheduling of the earnings announcement (RES). We find that portfolios that incorporate each of the signals individually produce higher out-of sample Sharpe ratios than simply holding an equal-weighted benchmark index. The portfolio that incorporates PEAD signals is the preferred portfolio among the three and an investor is no better off by incorporating the EAP and/or the RES signals relative to a portfolio that incorporates the PEAD signal alone. We also find that the gains are concentrated to only a few days surrounding the earnings announcement. This study provides a method that more closely resembles implementable trading strategies and provides new insight into the optimal implementation of the strategies.