Assessing the Determinants of Corporate Risk-Taking Using Machine Learning Algorithms

Given that risk-taking is an essential channel for companies to obtain high returns and realize value enhancement, the goal of this study holistically explore determinants corporate using various machine learning algorithms. Based on data from Chinese listed between 2010 2019, we document adaptive boosting (AdaBoost) model makes better predictions risk-taking. We further visualize importance influence firm basic characteristics, performance, chief executive officer (CEO) characteristics discover in AdaBoost model, performance factors, such as firm’s fixed asset investments, size, return equity, are important predicting risk-taking, while CEO less important. Finally, role variables varies among large small enterprises. Overall, our findings deepen comprehension what drives provide a potential way real-world firms seeking adjust their level.