Predicting the Likelihood of Lodging CMBS Loan Default

Abstract

This study investigates whether the traditional underwriting loan metrics, loan-to-value (LTV), debt coverage ratio (DSCR), and debt yield (DY) ratio, can predict lodging commercial mortgage-backed securities (CMBS) loan defaults. Using a data set of 5,266 fixed-rate lodging whole loans that were securitized into CMBS between 1996 and 2015, the results of the study provide evidence of significant relationships between all three metrics and the likelihood of default. The LTV varies positively with default, while the DSCR and DY are negatively related to default. These results hold for a subsample analysis of loans originated prior to the global financial crisis (GFC). Finally, the results show the DY spreads to be adequate proxies for the DY ratio.