Comovements in Trading activity: A Multivariate Autoregressive Model of Time Series Count Data Using Copulas

This paper introduces the Multivariate Autoregressive Conditional Poisson model to deal with issues of discreteness, overdispersion and both auto- and cross-correlation, arising with multivariate counts. We model counts with a double Poisson and assume that conditionally on past observations the means follow a Vector Autoregression. We resort to copulas to introduce contemporaneous correlation. We advocate the use of our model as a feasible alternative to multivariate duration models and apply it to the study of sector and stock specific news related to the comovements in the number of trades per unit of time of the most important US department stores traded on the New York Stock Exchange. We show that the market leaders inside an specific sector, in terms of more sectorial information conveyed by their trades, are related to their size measured by their market capitalization.