Revisiting the Relationship Between CEO Characteristics and Firm Internationalization: Evidence From a Machine Learning Approach

ABSTRACT This study leverages machine learning (ML) techniques to assess the impact of CEO characteristics on the international performance of firms. Analyzing data from Chinese listed companies between 2008 and 2021, this study evaluates 14 ML algorithms and identifies the random forest model as the most effective. Additionally, the SHapley Additive exPlanations (SHAP) algorithm is employed for result interpretation and visualization. The findings indicate that most CEO traits can predict a firm's international success. Notably, international experience, age, and CEO duality emerge as the top predictors. Specifically, both international experience and CEO duality positively influence performance, while the CEO's age exhibits a complex, non‐linear relationship with performance. This study provides a nuanced perspective on how CEO characteristics influence a firm's international success.