This paper investigates the credit enhancement value of NYSERDA's on-bill energy efficiency financing program relative to its similar conventional unsecured loan program. In the raw data, while both loan pools perform well relative to credit card lending, the on-bill loans default more often than the unsecured loans. This paper shows that this result persists: on-bill loans default more often, and this finding is not sensitive to model specification. This paper also shows that NYSERDA's alternate underwriting mechanism based on mortgage and utility bill repayment history performs well, and that projected dollar savings from the installed projects do not significantly influence loan performance.