Three Essays on Technology and Competition in Bank Lending and Corporate Finance

Abstract

This thesis consists of three empirical studies on bank lending and corporate finance, addressing two central policy-relevant issues for financial market dynamism and technology development. The first study explores whether and how banks develop specialized knowledge about corporate technology through their lending activities. We find that loans to firms sharing similar technologies with banks' prior borrowers obtain lower loan spreads. In the second study, we develop a novel AI-driven Co- Lending Graph Neural Network (CoLGNN) model to capture risk spillovers in syndicated lending markets. Our approach integrates bank characteristics, loan attributes, and network topology to generate a robust and novel spillover risk measure (CLN score) that predicts bank distress and profitability up to two years ahead. Leveraging natural experiments involving credit rating downgrades and the unexpected collapse of Lehman Brothers, we provide causal evidence of risk transmission across financial institutions. In the third study, we swift to a corporate finance to investigate how antitrust lawsuits affect corporations. We focus on government procurement cases, which we identify using large language models. Using a difference-in-differences design, we document that non-defendant firms gain federal contracts, expand employment and sales, and experience positive abnormal stock returns following antitrust filings. We provide evidence that these effects arise from the effective exclusion of defendant firms from the market. While antitrust lawsuits enhance market competition and reduce concentration, we document that the benefits are not evenly distributed: larger, well-established firms benefit more than smaller, financially constrained businesses. We also find no significant impact on government acquisition costs.