Returns Over the Life-Cycles of Open Market Repurchase Programs

Abstract

We construct a large sample of announcements that firms have authorized, suspended, resumed, or completed open market repurchase (OMR) programs. Starting or continuing repurchases is associated with positive average announcement period abnormal returns. Stopping repurchases, either by suspending uncompleted programs or by not immediately following completed programs with new programs, is associated with negative average announcement period abnormal returns. We find that excess returns are not significantly different from zero over the periods between initial authorizations and program completions, i.e., the life-cycles of the OMRs; however, consistent with prior literature, we find excess returns are positive and significant over standard long-run event windows. We then document that the excess returns over standard long-run windows are substantially attributable to positive and significant announcement returns to repurchase authorizations and takeover attempts that occur after the completions of the OMR programs. We confirm that these results generalize to the entire population of OMR programs. Further analysis suggests that the positive post-OMR abnormal returns documented in prior literature are a product of misspecification of the model for expected returns that, once corrected, no longer yields evidence of excess returns. Thus, we conclude that the positive long-run excess returns in prior literature do not represent evidence of market inefficiency.