Work in Progress

Covid Economics Working Paper Series (2020) [BibTeX] [COVID Economics WP] [CEPR WP] [VOX ] [Slides]

Banks with a greater deposit footprint in areas hit by COVID experience a significantly greater deterioration of their loan portfolios. While such lenders provide paycheck protection grants in affected areas more than other lenders, we observe negative effects on regular small business bank loans and on loan terms in the market for large syndicated loans.

Swiss Finance Institute Research Paper (2020) [BibTeX]

We explore which explanations for the reduction of bank branches are supported by the data. While technology is associated with de-branching of banks across countries, economic factors like growth or bank fragility and consolidation of banks are more robustly linked to de-branching in US counties and on the branch level in the US.

Main Publications

Journal of Financial and Quantitative Analysis (2020) [WP] [BibTeX] [Slides]

Changes in banking regulation have unintended and undocumented effects on the market for corporate credit. Bank branching deregulation following the Riegle-Neal Interstate Branching and Banking Efficiency Act of 1994 decreased syndicated loan issuance but spurred bilateral lending to corporations. This shift is also reflected in interest rate spreads, pointing to a supply-driven substitution effect. Results suggest that changes to banking regulation can affect not just the amount but also type of credit in the economy.

Industrial and Corporate Change (2019) [WP] [BibTeX] [Data] [Slides]

There is no evidence that concentration has any type of positive effect on long-run profitability differences - testing linear relations, interactions with critical concentration levels and with mobility barriers. Introducing business segments data in this analysis, a new IV and a novel natural experiment I obtain results that point into the opposite direction, towards statistically and economically significant negative causal effects.

Journal of Corporate Finance (2017) [WP] [BibTeX] [Data]

Using Compustat to measure industry concentration is problematic. Popular Herfindahl Index approximations have vanishingly low correlations with the more comprehensive Census metric. As a result, major variables of interest in corporate finance correlate markedly different with these indicators. I show that this can lead to a breakdown of regression results.