Earnings-Announcement Timing Factors

Abstract

Contrarian anomalies have earned no premiums since 2003. Earnings announcement-timed versions of these factors, however, continue to earn substantial premiums. These factors condition on firms’ locations in their earnings calendars; they are profitable because mispricings first expand and then contract from one earnings announcement to the next. We show that this return pattern aligns with an optimism-pessimism cycle in analysts’ recommendations and that a risk-based explanation for this effect is unlikely. Our results suggest that the disappearance of unconditional anomalies does not imply that the market has become more ecient.

Keywords: Anomalies, analysts, market efficiency

JEL Classification: G12, G14, G40