Multivariate Markov Families of Copulas
19 Pages Posted: 8 Oct 2015
Date Written: October 8, 2015
For the Markov property of a multivariate process, a necessary and sufficient condition on the multi-dimensional copula of the finite-dimensional distributions is given. This establishes that the Markov property is solely a property of the copula, i.e., of the dependence structure. This extends results by Darswo et al. (1992) from dimension one to the multivariate case. In addition to the one-dimensional case also the spatial copula between the different dimensions has to be taken into account. Examples are also given.
Keywords: Markov process, copula, Chapman-Kolmogorov equation
JEL Classification: C30, C22
Suggested Citation: Suggested Citation