Post-Earnings-Announcement Drift: Expected Growth Risk or Limits-to-Arbitrage?

To explain post-earnings-announcement drift (PEAD), we suggest expected growth risk, which is measured as covariance between stock returns and expected future real GDP growth rates. We find that both expected growth rates and expected growth risk increase with standardized unexpected earnings, and expected growth risk is significantly priced in the cross-section of returns. The model including expected growth risk alone explains PEAD satisfactorily. We also find that after adjustment for expected growth risk, the systematic relation of PEAD with the degree of limits-to-arbitrage disappears. This indicates that the empirical evidence supporting the mispricing hypothesis due to limits-to-arbitrage is a consequence of the failure in incorporating appropriate risk and the drift is a manifestation of expected growth risk.