A holistic approach to merger models with an emphasis on heterogeneity

Though mergers and acquisitions comprise a significant share of aggregate economic activity in the US, researchers disagree on the effect and importance of M&A determinants. Such disagreement stems from adherence to disparate theoretical models that preclude the use of significant explanatory variables. Our paper contributes to this scattered literature by demonstrating the presence of significant bias in the most common approaches to this question. We do so by comparing the Neoclassical, behavioral, and macro approaches to explaining M&A activity. Finding that the assumptions of these respective paradigms have inhibited individual researchers from including important determinants of M&A, we show that these constraints have generated misleading point estimates. By constructing a more holistic model of M&A activity, we produce more accurate estimates of merger determinants and reveal the bias of previous estimation techniques. Additionally, we relax the common, yet unrealistic, assumption that firms are homogeneous, showing that the determinants of M&A vary by firm size, region, and sector. These results demonstrate that an aggregated approach to merger determinants conceals important variation.