Earnings Seasons: Analysis of Inferred and Announced News

Abstract

Earnings announcements are clustered in calendar time and such clusters, referred to as earnings seasons, are characterized by intense arrival of information from firms in an industry. The news arrives not only from earnings announcement by the firm, announced news, but also from pre-emptive inference of news from peer firm announcements, inferred news. We find that announced news appears to be underweighted for firms that announce later in the earnings season and this is not explained by a higher level of inferred news for these firms. This suggests that as the season progresses, investors appear to pay less attention to firm earnings, particularly in the case of good news. Finally, the tone of the season, i.e. nature of news announced on the first day of the season, affects the reaction to announced news of subsequent announcers. Our results highlight the importance of viewing earnings announcements in the context of the earnings season.