Risk measures for skew normal mixtures

In this paper we show that linear combinations of multivariate skew normal mixtures can be represented as finite mixtures of univariate skew normals. Based on this result we provide an analytical formula for some well known risk measures.