Vybrané přístupy pro zpracování mnohorozměrných časových řad ve financích

Abstract

In the present work, we study ARMA model at the beginning, then we write about one-dimensional and multivariate ARCH and GARCH model, further we move on to the multivariate GARCH model. At the end, the principal component decomposition is introduced, it is a procedure to reduce the number of parameters involved in a multivariate GARCH model. The theory is explicated rst on a basic ARMA model, afterwards it is modi ed step by step for the one-dimensional and the multivariate GARCH model. There are solved examples for multivariate ARCH and GARCH model and nancial data are analyzed by means of these models.