Search for: [Abstrakt = "This paper considers the risk of portfolios which components show concurrence of untypically high losses. Two kinds of models for Value at Risk estimation are discussed\: static models with extreme dependence and stochastic process with varying conditional covariance \(M\-GARCH\). The aim of the investigation is to compare these models and determine whether unconditional extreme dependence may be sufficient for risk measurement purposes even if the true cause of fat tails in the multivariate unconditional distribution is the presence of M\-GARCH effect in data."]
