Research

Research Interests

Corporate Finance, Financial Intermediation, Organizational Economics

Publications

Bank Response to Higher Capital Requirements: Evidence from a Quasi-Natural Experiment (with Reint Gropp, Steven Ongena, and Carlo Wix), Review of Financial Studies, 2019, Volume 32(1), 266–299.

Press: HLS

Supranational Rules, National Discretion: Increasing versus Inflating Regulatory Bank Capital? (with Reint Gropp, Steven Ongena, Ines Simac, and Carlo Wix), Journal of Financial and Quantitative Analysis, 2024, Volume 59(2), 830 – 862.

Press: VoxEU

Video: YouTube

Working papers

Who Pays for Payment Fraud? Detection and Liability Rules under Strategic Fraudster Adaptation (With Vimal Balasubramaniam, and Antoine Uettwiller)

We develop a dynamic model of payment fraud detection in which fraudsters strategically adapt their methods in response to detection technologies, causing model performance to decay over time. We characterize the socially optimal detection level and show that in competitive markets full liability induces overinvestment: each PSP ignores how its detection effort accelerates fraudster adaptation marketwide, leading to excessive investment relative to the social optimum. When PSPs retrain models to counter decay, competition creates a ratchet effect-detection accelerates decay, increasing retraining frequency beyond the socially optimal level. Optimal liability allocation sets full reimbursement for both sending and receiving PSPs, complemented by a regulatory fee that internalizes the social cost of fraud and eliminates free-riding in detection efforts. Using Reddit discussion data and bot-detection Twitter data, we document strategic adaptation-fraudsters reallocate effort across scam types following detection, and detection performance deteriorates as fraud becomes more sophisticated, independent of overall fraud rates.

Stung into Action? Cyber Risk Management After a Breach (with Emina Ademović, Emina Ademović and Vasso Ioannidou)

We study how firms adapt their cybersecurity defenses following major cyber incidents. Combining multiple sources of incident disclosures with job posting data for U.S. listed firms between 2010 and 2022, we construct novel firm-level measures of cybersecurity labor demand, software use, and IT architectures. Incidents trigger substantial and persistent increases in cybersecurity hiring and adoption of advanced technologies. Affected firms leapfrog ahead of industry peers. Firms in the same industry and firms with similar IT systems also strengthen defenses. Among financial institutions, larger firms react more strongly, consistent with greater scrutiny and exposure to cyber risk.

Dear Congressmember: Letters to Congress and the shaping of (financial) legislation (with Renée Adams)

Does petitioning Congress work? We examine whether congressmembers appear responsive to letters by consumer groups and finance trade associations concerning 821 bills introduced between 1999 and 2018. We find that positions advanced in both consumer group and trade association letters predict the voting behavior of legislators. However, campaign contributions, membership of the Financial Services Committee and career concerns amplify the responsiveness of legislators to positions by resource-rich trade associations, especially during early stages of the legislative process. Our results suggest that the right to petition works, but legislators’ incentives moderate its ability to provide equal access to competing interests.

Bargaining with a Bank

This paper examines bargaining as a mechanism to resolve information problems. To guide the analysis, I develop a parsimonious model of a credit negotiation between a bank and firms with varying impatience. In equilibrium, impatient firms accept the bank’s offer immediately, while patient firms wait and negotiate price adjustments. I test the empirical predictions using a hand-collected dataset on credit line negotiations. Firms signing the bank’s offer right away draw down their credit line after origination and default more than late signers. Late signers negotiate price adjustments more frequently, and, consistent with the model, these adjustments predict better ex post performance.

Strategic Use of Information and the Allocation of Authority

This paper investigates how agents use their information strategically to affect the allocation of decision-making authority. The empirical analysis uses a unique data set on credit applications of a large commercial bank. Loan officers could use multiple scoring trials to adjust the credit application and game the rules that determine the hierarchical distance between the loan officer and the loan approving officer. By exploiting rule changes which increased the hierarchical distance for a subset of the credit applications, the paper shows that loan officers use their information to alter the allocation of authority in the opposite direction of the imposed rule changes. In addition, the paper shows that loan officer have fewer incentives to game the rules after they lost the authority to approve high quality credit applications. The results show that (re)allocating authority in organizations may be difficult if information collecting agents could use their information strategically to affect the allocation of authority.

Is Loan Officer Discretion Adviced When Viewing Soft Information? (with Hans Degryse, Jose Liberti and Steven Ongena)

We show that the collection of soft information on the activities of small and medium sized enterprises and the exercising of loan officer discretion helps in monitoring these borrowers. We measure loan officer discretion as the deviations in granted loan amounts from the amounts stemming from the bank’s own credit scoring model. Soft information guides discretion, and helps in predicting loan default even when controlling for all available public and private information. Loan officers use soft information when deciding on the loan amount that is being granted: A one standard deviation of more favourable soft information results in the granting of a 16 percent higher loan amount. Beyond using soft information, loan officer discretion per se neither improves nor deteriorates loan outcomes.