Working Papers
Mark-to-Market and Bank Risk-Taking: Evidence from Corporate Bond Portfolios, with Stephen A. Karolyi, Arthur Luo, and Ahyan Panjwani [DRAFT COMING SOON]
We study how regulatory capital mark-to-market affects banks’ risk-taking and trading incentives. Using transaction-level data on the universe of large U.S. banks’ corporate bond holdings and quasi-experimental variation in the accumulated other comprehensive income (AOCI) regulatory filter over the 2015-2024 period, we find that banks whose regulatory capital is insulated from mark-to-market take more risk ex-ante, purchasing high-yield bonds with longer duration and higher credit risk. Ex-post, shielded banks are more likely to sell bonds at a gain and less likely to sell them at a loss, a disposition-like effect induced by capital regulation. Shielded banks with large unrealized bond losses subsidize their issuers through the lending book, suggesting unintended spillovers from the regulatory accounting of banks’ securities portfolios.
We investigate how the financial constraints of local governments interact with those of their corporate suppliers to affect outcomes in public procurement. Using auction-level data on Brazilian municipal procurement and population-based discontinuities in federal transfers, we show that higher transfers relieve local municipalities' constraints, allowing larger tenders and faster payments to suppliers. This draws new suppliers into competition, especially small, young, and distant firms that face participation and delivery costs and are themselves constrained. Higher supplier competition is associated with lower procurement prices especially for essential public goods and services, suggesting pass-through effects to local populations.
We show that banks offer higher deposit rates to high-income households: from the bottom to the top income decile, rates increase by 55% of the sample median. These spreads persist independent of banking competition, and seem to arise from banks internalizing participation in nondeposit markets: only participation-related income predicts rates. We identify this mechanism using the 2012 American Taxpayer Relief Act, a capital taxation change whose pass-through varied with state taxes, as a shock to participation incentives. Higher capital taxes lowered participation and, in turn, deposit rates. Our findings highlight lack of participation as a source of deposit market power.
We examine monetary policy transmission when deposit market structure is endogenous. Expansionary monetary policy stimulates bank entry, especially when entry barriers are low. Banks' deposit quantity sensitivities are increasing in entry barriers, but the number of local banks is decreasing in entry barriers, and this channel dominates. Hence, higher entry barriers are associated with reduced monetary policy transmission. We test this prediction using novel, network-based measures of entry barrier shocks stemming from U.S. bank deregulation. Consistent with the model, local establishment and employment growth increase more in response to expansionary monetary policy when entry barriers are lower.