Abstract

What is the impact of granular credit risk on banks and on the economy? We provide the first causal identification of single-name counterparty exposure risk in bank portfolios by applying a new empirical approach on an administrative matched bank-firm dataset from Norway. Exploiting the fat tail properties of the loan share distribution we use a Gabaix and Koijen (2020a,b) granular instrumental variable strategy to show that idiosyncratic borrower risk survives aggregation in banks portfolios. We also find that this granular credit risk spills over from affected banks to firms, decreases investment, and increases the probability of default of non-granular borrowers, thereby sizably affecting the macroeconomy.