I argue that academic research often inadequately accounts for alpha decay. As an anomaly's alpha (i.e., the risk-adjusted expected excess return) and realized returns are negatively related, alpha decay coincides with positive realized returns. If the alpha decays at publication, observers may...