 |
 |
Valuation Commentary - Feb. '08
How to use Credit OAS: an ABS Valuation Case Study for the Illiquid Market
by Alex Levin
In a number of Pipeline articles and at AD&Co’s Annual Conference in June 2007, we introduced the concept of Credit OAS as a sound analytical approach to modeling non-agency MBS. The basic idea is similar to the traditional OAS method, but it revolves around coupled simulations of interest rates and home prices. It also requires the use of a model for defaults and losses (e.g., the LoanDynamics™ Model, LDM), which delivers projected vectors of prepayment, default, loss severity and delinquency rates for every market scenario. These vectors are then submitted to a cashflow generator; the resultant cashflows are discounted using the usual method, but without a credit-inflated spread. The discount spread, which we denote crOAS, accounts for liquidity risk, but not for losses, which are now modeled explicitly. The crOAS metric levels the playing field among ABS tranches of different supports and credit ratings. In a liquidity nirvana, all ABS tranches are expected to have zero crOAS to credit-perfect benchmarks.
Last month we mentioned that an application of the Credit OAS approach may result in a totally different exposure to interest rates from that of the traditional OAS method. We brought up two arguments: A) the loss stream may not necessarily look like an IO nor be equivalent to an inflated discounting, and B) a home price dependence on interest rates may play a key role.
Let us illustrate the application of the Credit OAS method to a case study of the SAS06BC4 subprime deal. The date of this analysis is September 28, 2007, by which the deal had accumulated 5.0% delinquencies and 9.5% severe delinquencies. We analyzed the deal’s collateral and 4 tranches, A5, M2, M5 and M8, differing in credit protection. Each tranche’s protection had gone up since origination, but this fact points merely to the very low level of already accumulated collateral losses (0.22%). Obviously, fairly large losses are expected to occur in the future given the above mentioned impaired pool composition and falling or stagnated home prices. Actual pricing quotes were available for all tranches; these prices should be viewed with caution, however, because of the illiquid market. Nevertheless, we show how the Credit OAS method can help us analyze the deal from many practical angles.
