The multivariate Edgeworth-Sargan density

The Edgeworth-Sargan density has been shown capable of capturing empirical regularities of financial data (thick tails and asymmetries). When compared to other densities used in applied finance, it has the advantage of its analytical simplicity, and the ability to improve data fits by adding more parameters in a natural way. This paper develops an explicit form for the multivariate Edgeworth-Sargan density and compare its performance to the multivariate Students t. The comparison is carried out with daily financial observations, spanning 25 years of data for several financial variables that include stock markets indices and interest and exchange rates for several countries.JEL Classification: C12, G1The author acknowledges Prof. Ignacio Mauleón, PhD supervisor of the dissertation this article is based on, and also to Esther del Brío, Helen Jones and two anonymous referees for their helpful and constructive comments. This paper is financed by the C.I.C.Y.T. under project SEC98-1112 and the J.C.Y.L. under project SA29/99. Responsibility for the contents, as well as possible deficiencies, remains solely with the author.