Yongzhe Wang, University of North Carolina, Chapel Hill
This study reveals that commercial banking backgrounds of corporate executives negatively impact corporate innovation by reducing patent applications, researcher employment, and R&D expenses. The effects intensify as the number of executives possessing such backgrounds increases. A novel machine learning framework, ML-DID, is outlined to mitigate endogeneity concerns, including omitted variable bias and selection bias in the setting of “Panel Data with Multiple Treatment Timing.” Its robustness is tested against various specifications using the random forest algorithm. Additional analysis demonstrates that the impact of a banking background is strongest among firms with financial constraints and in less competitive industries.
Read the full paper here.