I study the time series dynamics of commercial mortgage credit risk and the unobservable systematic risk factors underlying those dynamics. The research is conducted within a structural model framework. I modify the first-passage model of Black and Cox (1976) and Longstaff and Schwartz (1995) to capture unique features of commercial mortgage default. First, the first-passage condition is imposed on net operating income (NOI) rather than on property value, which reflects the fact that commercial mortgages mainly rely on the income from underlying properties to service the debt. Second, equilibrium macroeconomic dynamics are linked to commercial mortgage default through the NOI function, based on the observation that commercial property cash flow is largely affected by macroeconomic conditions. Third, I solve the default hazard rate of a representative commercial mortgage as a function of two unobservable state variables, the expected macroeconomic growth and expected property market-specific growth. The solutions of the model provide an estimable system that includes the relationship between systematic risk factors and commercial mortgage credit risk, as well as the dynamics of the risk factors.

My empirical work aims at extracting information about how the systematic risk factors drive the evolutions of commercial mortgage credit risk based on the observed default events of individual loans. I first use a Cox proportional hazard model to control the heterogeneity of commercial mortgage loans and to estimate the default hazard rate time series of a representative mortgage, which is the systematic component of default risk in the commercial mortgage market. The hazard rate time series are then used to estimate my first-passage model in state space form. I estimate the nonlinear state space model using extended Kalman filter. Results show large variations of default probability over time in the commercial mortgage market, and that these variations are well explained by the two risk factors. Both factors are mean-reverting, and the elasticity of commercial mortgage default probability with respect to the expected macroeconomic growth increases when the economy deteriorates. The estimates also reveal a negative relationship between the instantaneous risk free interest rate and commercial mortgage default probability. My model can be used in predicting long term dynamics of commercial mortgage credit risk.