University of L'Aquila, Faculty of Economics, Italy
Measuring the similarity and dissimilarity between models is crucial in many fields of time series analysis. A number of parametric measures of dissimilarity between univariate linear models have been suggested in literature. Up our knowledge, no distance measures between vector autoregressive (VAR) models has yet been proposed. The aim of this paper is to introduce a distance measure between VAR models. This measure can be considered an extension for the multivariate case of the autoregressive distance proposed by Piccolo (1989, 1990). However, it is important to note that the proposed distance is a metric in the class of VAR processes, whereas the autoregressive distance is not a metric in the class of ARMA processes.
Rua do Quelhas 6
Phone: +351-21 3925876