Linear credit risk models (Q2282965)

In this extensive paper, the authors present a new linear credit risk model. This approach has the flexibility to deal with the multi-named model associated with negatively correlated default intensities. Within this environment, a single-name model is defined, referred to as the linear hypercube (LHC). A comparison of this is made with the one-factor affine default intensity model. A useful review of the current literature is given. An approximation approach for the pricing of European-styled options on underlying assets that are exposed credit risk is described. This is aided by an application of the Stone-Weierstrass theorem. Interestingly, from this, the price of a CDS option may be approximated by polynomials appearing in the factors. The method may also be used to incorporate CVA in the pricing. The paper continues with an empirical analysis of the LHC model focusing on the efficiency of the approximation techniques.