An Evaluation of the Base Correlation Framework for Synthetic CDOs

Abstract

In April 2004, JP Morgan introduced the notion of base correlations, a novel approach to quoting correlations for synthetic CDO tranches, which facilitates a simple relative valuation of off-market tranches. Using a simple intensity based credit risk model we generate 'true' tranche spreads and examine the behavior of the base correlations and the merits of the relative valuation approach. We reach four conclusions: First, even if the true default correlation increases, base correlations for some tranches may actually decrease. Second, we uncover a structural uniqueness problem with the definition of base correlations. Third, in the relative valuation framework expected losses can go negative for steep correlation skews. Fourth, the relative spread errors are small in some segments but can be very large, +/- 10% for mezzanine tranches, the errors changing sign from tranche to tranche.