Can bank regulatory technology alleviate financial mismatch? Causal evidence from double-debiased machine learning on bank-firm matched data

Financial mismatch (FM) remains a major challenge for firms, especially amid information asymmetry. The emergence of bank regulatory technology (RegTech) is reshaping regulation and risk management in banking. Utilizing a panel dataset of bank-firm matched loan-level data from 2014 to 2023, we employ double-debiased machine learning to provide empirical evidence that bank RegTech significantly reduces firms’ FM: one-standard-deviation increase in bank RegTech corresponds to at least a 2.29% reduction in the FM. This effect operates through three main channels: improved information transparency, eased financing constraints, and reduced managerial performance pressure. Investor attention amplifies the mitigating impact of bank RegTech on FM. The effects are heterogeneous, with more pronounced impacts observed among non-state-owned enterprises, high-tech firms, firms in less competitive industries, and firms with established bank-firm relationships. Results hold after rigorous robustness validation. Finally, we further demonstrate that reduced FM leads to lower operational risk and a decline in corporate financialization.