Integrating the Macroeconomy into Consumer Loan Loss Forecasting

Abstract

The recent subprime mortgage meltdown has forced the consumer credit industry to reevaluate many of its methodologies for assessing credit risk. Given the economy-wide nature of the crisis, and the structural weaknesses it has exposed, a purely microeconomic analysis of the episode will likely prove incomplete. In this article, the authors take a macroeconomic view of credit risk, identifying linkages between mortgage credit performance and the underlying economic, housing, and credit cycles. They identify two distinct cycles, one pertaining to origination quality and underwriting and the other to the performance of legacy accounts. The authors then turn their attention to loan-level data, examining the impact of the business cycle on the distribution of loan delinquencies while pointing out some of the inherent weaknesses associated with such an analysis. Finally, they propose new methods for dealing with these weaknesses by calibrating the loan-level vectors to forecasts generated using the aggregate vintage model. The data used for this analysis were obtained from the liquidator of a large, now defunct, subprime lender. The authors’ findings should therefore be of considerable interest to anyone trying to understand the causes of the crisis.