Test du rapport de vraisemblance pour le rang de cointégration d'un VAR avec des erreurs dépendantes

In this Note we consider the estimation and the test of the co-integration rank of multiple autoregressive time series model with nonindependent innovations. We show the consistency of the estimators and the validity of the likelihood ratio test in the presence of error terms with heteroscedasticity or other forms of dependence. To cite this article: H. Raïssi, C. R. Acad. Sci. Paris, Ser. I 346 (2008).