XVA sensitivities

Summary

An important family of CCR metrics is the XVA family associated with a portfolio of trades. XVA values represent the price of credit risk, funding costs and capital costs in books of derivatives. This chapter discusses different ways in which the Composition Technique can be used to accelerate the computation of XVA sensitivities. A direct and accurate way of computing sensitivities using the pricing functions of a risk engine involves using finite differences. In terms of accuracy, the state‐of‐the‐art technique to compute sensitivities is Adjoint Algorithmic Differentiation. With it, XVA sensitivities can be obtained, to a high degree of precision, in about 10 times what it takes to compute a single XVA value. The increased computation capability for XVA sensitivities assumes the building of proxy pricers every time the XVA function is called.