New findings on repurchase anomaly - The first-month effect

Prior studies find positive abnormal returns following share repurchase announcements. We examine the association between announcement month and the excess returns. We find that the average excess return is consistently higher for repurchases announced in the first month of a fiscal quarter than for those announced in the other two months. Interestingly, 1st-month and non-1st-month announcers are highly comparable in firm characteristics, pre-announcement returns, and disclosed motives. The magnitude of the first-month effect barely changes after we switch to multivariate regressions, and it remains large under firm-fixed effects. Investment strategies based on BTM, firm size, and pre-announcement return are all improved by a first-month strategy. The first-month effect extends well beyond the first year, but the market does not seem to realize it. We propose an explanation for the first-month effect based on the conjecture that managers receive firm information in an uneven manner throughout a fiscal period. This explanation is supported by our empirical tests. Additional empirical analyses provide results that fail to support several plausible alternative explanations.