Predicting Cross-border Merger and Acquisition Completion through CEO Characteristics: A Machine Learning Approach

The issue of cross-border merger and acquisition (CBMA) completion has garnered significant attention from scholars, yet existing research primarily focuses on macro- and meso-level determinants, with limited exploration of micro-level factors. Drawing on upper echelons theory, this study examines the predictive power of three categories of CEO characteristics-demographic background, professional background, and personality traits-on CBMA completion. Utilizing a dataset of 4,541 CBMA transactions announced by U.S.-listed firms from 1990 to 2021, we employ the LightGBM machine learning model and SHapley Additive exPlanations (SHAP) algorithms to analyze the importance and predictive patterns of these CEO characteristics. Our findings reveal that CEO personality traits, overall, exhibit the strongest predictive power among the three categories. The top five CEO characteristics identified are CEO conscientiousness, CEO promotion focus, CEO external focus, CEO duality, and CEO neuroticism, with many characteristics displaying non-linear relationships with CBMA completion. This study extends the application of upper echelons theory to the domain of CBMA completion and highlights new avenues for future research.