Financial densities in emerging markets: an application of the multivariate ES density

This paper derives and presents the multivariate Edgeworth–Sargan (ES) density, discusses some of its properties, and estimates it for three exchange rates in emerging markets (Chile, Hungary and Singapore). The ES density fits the data adequately, and the model is estimated simultaneously for all variables. This involves estimating a highly non-linear model with 32 parameters. A multivariate Student's t is also estimated, and both sets of results are compared. The empirical results show that, (a) the ES density applies to emerging markets as well as to more developed economies, as shown in previous research, (b) it is feasible to estimate a multivariate density of large dimensionality, and (c) independent estimation of the marginal densities, although a consistent procedure, yields significantly different results from the multivariate estimation for some parameters.