Here Comes the Sun: An Application of Survival Analysis Techniques to Model Solar Asset-Backed Security Cash Flows

Abstract

The growing popularity of solar photo-voltaic energy to power homes has led to the rise of loans as a homeowner-financing solution and securitization as a developer- financing solution. The SolarCity Corporation, a leading solar panel developer, has issued two securities of these loans, in January 2016 and January 2017, and others are looking to do the same. This thesis tackles the leading challenge facing the industry, the lack of historical data, by applying techniques of survival analysis (the Kaplan- Meier method and the Cox proportional hazard model) to a data set of solar loans. Using the functions of prepayment and default survival fitted by these methods, I model first the cash flows of the pool of loans underlying SolarCity’s FTE Series 2 2017-A security and second the resulting performance of the security in cash flows to its investors. I find that, because the loan terms (30 years) are much longer than the anticipated maturity of the security (6 years), higher prepayment rates are in fact better for the performance of the security. I also find that the pool of loans modeled with these parameters is expected to fail to meet its obligations, which suggests that SolarCity used lower default rates, higher recovery rates, or the addition of fresh loans in its own pricing model.