The Effects of Underwriting Practices on Loan Losses: Evidence from the FDIC Survey of Bank Lending Practices

Abstract

Purpose – This chapter investigates the influence of bank loan underwriting practices on loan losses and identifies potential determinants of lending practices for five categories of loans: business, consumer, commercial real estate, home equity, and construction and land development loans. Methodology/approach – Using data on the riskiness of lending practices obtained from the U.S. Federal Deposit Insurance Corporation (FDIC) bank examiner surveys from January 1996 to March 2009, I fit a two-step treatment effects model to measure the effects of underwriting practices on loan losses, controlling for the potential endogeneity of lending practices. Findings – In the selection step, I find that for business loans, the likelihood that bank management will adopt low-risk lending practices increases with bank financial performance and management quality hierarchical complexity and decreases with market competition. Results for the selection of lending practices for consumer loans and three categories of real estate loans are similar to those found for business loans but show weaker statistical relationships to all explanatory variables. In the loss determination step, I find that lower (higher) risk underwriting practices are generally associated with lower (higher) gross loan charge-offs (as percentage of gross loans and leases) for five categories of loans: business, consumer, commercial real estate, home equity, and construction and land development loans. Originality/value of chapter – This is the first study to model the determinants of loan underwriting practices with the practices being characterized in terms of their risk to the bank. In addition, this is the first study to consider the effects of the riskiness of lending practices on loan losses, controlling for the endogeneity of practices.