This study examines the importance of the self-selection problem when evaluating returns to bidder firms around announcement events. Takeover announcements are not random because managers decide rationally whether to bid or not, which indicates announcements are timed; consequently, in the presence of the sample selection problem, standard ordinary least square estimates are biased. Using a conditional model, the results indicate that after controlling for the self-selection bias effect, shareholders of bidder firms make normal returns. In sum, failing to account for sample selection bias may lead to erroneous conclusions about a bidder’s true economic wealth effects around an announcement event.
- sample selection bias