Abstract
Drawing on behavioral sufficient statistic theory, we propose a reduced-form approach to examine the influence of cognitive biases in the decision-making process involving mergers and acquisitions (M&As). We re-conceptualize managerial ability not merely as an efficiency metric, but as a proxy for instrumental rationality and cognitive resource availability. Thus, our reduced-form model considers managerial ability as a sufficient statistic to investigate cognitive biases. We assume that more (less) able managers are less (more) susceptible to the influence of cognitive biases. Using large-scale archival financial data, we conduct two sets of tests. First, we validate our proxy by adapting the empirical frameworks of prior studies and substituting their cognitive-bias measures with the managerial ability score. We find robust evidence that managerial ability functions as a strong proxy for cognitive biases. Second, we test our proxy in the context of M&As. The main results reveal that a positive and statistically significant association between the premium paid on M&As-accounting goodwill-and firm returns emerges only in the top decile of managerial ability (where decision-making is less afflicted by behavioral distortions). Our findings indicate an M&A failure rate of 87%, suggesting that the majority of premiums paid in M&A deals represents the materialization of judgment errors rather than economic substance.