Working papers

Banks and Non-Banks: Competitors in Loan Origination, Partners in Regulatory Arbitrage

Abstract: This paper studies how interactions between banks and non-banks can exacerbate regulatory arbitrage and affect competition in the loan market. I develop a two-stage game in which a bank and a non-bank first compete for loan origination and then contract on the performance of the resulting portfolios. The bank differs from the non-bank as it can issue government-insured deposits and is subject to capital requirements. Capital requirements are proportional to expected losses and apply to individual exposures, while the deposit guarantee operates at the balance-sheet level. This regulatory asymmetry creates scope for arbitrage. The optimal contract reallocates cash flows away from states in which the rest of the balance sheet performs poorly toward states in which it performs well. I show that contracts such as Synthetic Risk Transfers (SRT), which appear to off-load risk from the banking sector, can instead increase the value of the implicit subsidy from the deposit guarantee. Finally, I show that the optimal contract can expand credit supply, though at the cost of greater rent extraction.
Related policy discussion (with Frederic Malherbe) on the growth of private markets in the UK following reforms introduced after 2008, opened by the Financial Services Regulation Committee (UK Parliament link)
Non-technical summary (UCL Econ Brief)

Cournot Competition in the Loan Market: Microfoundations and Limitations (submitted)
the paper previously circulated as "How to model bank competition: the Case for Cournot"

Abstract: When firms choose capacity and then compete à la Bertrand, the equilibrium corresponds to the Cournot outcome (Kreps and Scheinkman, 1983). In banking, regulatory capital requirements constrain a bank’s lending capacity, making capital choices analogous to capacity choices. This paper establishes the conditions under which the Bertrand-Cournot equivalence extends to banks with risky loans and asymmetric information. The equivalence holds for a broad class of information frictions, including moral hazard and adverse selection, but breaks down under screening and relationship banking. These microfoundations clarify when Cournot competition is a valid and tractable modeling choice for the banking sector.

Publications

Bank Capital Requirements and Bank Lending: From Theory to Empirics to Policy (with Saleem Bahaj and Frederic Malherbe) 2024. Forthcoming in the Handbook of Macroprudential Regulation. (also CEPR DP19942)

Abstract: Macroprudential regulation is often viewed as a trade-off between banking system stability and aggregate credit supply. In this paper, we provide a comprehensive analysis of how changes in capital requirements affect bank lending. We use a theoretical framework to assess and nuance the trade-off. We show that imperfect competition, general equilibrium effects, and asset heterogeneity among banks result in lending responses that are complex and difficult to estimate. Armed with these theoretical insights, we assess existing strategies in the empirical literature and provide guidance for future research.