Title: On testing for independence between the generalized errors of several time series
Authors: Bruno Remillard - HEC Montreal (Canada) [presenting]
Abstract: Test statistics are proposed for checking the independence between the generalized errors of several univariate time series models. These models include volatility models as well as regime-switching models. In order to obtain consistent tests, the statistics are constructed from lagged empirical processes whose asymptotic distribution is studied. Examples of applications from financial time series are given.