Inferring Correlations of Asset Values and Distances-to-Default from CDS Spreads: A Structural Model Approach

Abstract

Using structural credit risk models to estimate default dependence requires estimates of correlations of changes in distance-to-default. We present a structural model that yields simple relations between asset value, distance-to-default, and CDS spreads, allowing the correlations to be estimated from CDS spreads. We generalize the model to include a randomly varying default boundary; in this version the distance-to-default dynamics also depend on the movement of the default boundary. The CDS spread correlations we estimate exceed equity correlations, consistent with a randomly varying default boundary. We also present evidence that variations in funding liquidity affect the correlations, consistent with recent models.