DISTRIBUTIONAL TESTS IN MULTIVARIATE DYNAMIC MODELS WITH NORMAL AND STUDENT T INNOVATIONS

We derive specification tests for the null hypotheses of multivariate normal and Student t innovations using the Generalised Hyperbolic distribution as our alternative hypothesis. In both cases, we decompose the corresponding Lagrange Multiplier-type tests into skewness and kurtosis components, from which we obtain more powerful one-sided Kuhn-Tucker versions that are asymptotically equivalent to the Likelihood Ratio test. We conduct detailed Monte Carlo exercises that compare our proposed tests with their competitors in finite samples. Finally, we present an empirical application to ten US sectoral stock returns, which indicates that their conditional distribution is mildly asymmetric and strongly leptokurtic.