Winning by Losing: Evidence on Overbidding in Mergers

Do shareholders of acquiring companies profit from acquisitions, or do acquiring CEOs overbidand destroy shareholder value? Answering this question is difficult since the hypotheticalcounterfactual is hard to determine. We exploit merger contests to address the identificationissue. In those cases where, ex ante, at least two bidders had a significant chance at winningthe contest, the post-merger performance of the loser allows calculating the counterfactualperformance of the winner without the merger. In a novel data set of merger contests since1985, we find that the returns of bidders are closely aligned before the merger contest, butdiverge afterwards. In the sample where the loser had a significant chance to win, winnersunderperform losers by 48 percent over the following three years. Our results also imply thatannouncement returns fail to provide an informative estimate of the causal effect of mergersin our sample. Existing measures of long-run abnormal returns tend to underestimate thenegative return implications.