Commercial Mortgage Backed Securities: How Much Subordination is Enough?

The commercial mortgage-backed security market has experienced rapid growth in recent years, and is now the second most important source of intermediation to the commercial real estate sector. Despite its growing importance, relatively little academic research has questioned the apparent success of the CMBS capital structure. In this paper, we study whether the recent growth in the CMBS market actually reflects the mitigation of a market imperfection, as some have suggested, or reflects ever thinner subordination levels, perhaps backed up by overly aggressive assumptions about future commercial real estate performance. Our analysis is based on a multi-factor structural model of commercial mortage defaults in which defaults are a function of interest rates and property prices. The model incorporates heterogeneity at the borrower level and transaction frictions, rendering solution of the model highly computationally burdensome task. We solve the model using a parallel alogirithm developed for the Rice Terascale Cluster (RTC), a computational platform capable of terabyte processing and storage. Our preliminary results indicate that, somewhat surprisingly, the optimal levels of CMBS subordination are lower than those observed in recent CMBS deals. One interpretation of these results is that, in the absence of a performance track record for securitized commercial mortgages, the subordination levels of early deals were set too high. Under this interpretation, our results imply that in the future the CMBS market will likely see further reductions in subordination and continued rapid growth. An indirect implication of this conclusion is that the CMBS capital structure does create value. These results have implications not just for commercial real estate, but also other markets that employ securitization structures similar to those in the CMBS market, such as asset-backed commercial paper and CDOs