A Multivariate GARCH Model with Time-Varying Correlations

摘要: In this paper we propose a new multivariate GARCH model with time-varying correlations. We adopt the vech representation based on conditional variances and While each conditional-variance term is assumed to follow univariate formulation, conditional-correlation matrix postulated an autoregressive moving average type of analogue. By imposing some suitable restrictions conditional-correlation-matrix equation, construct MGARCH in which guaranteed be positive definite during optimisation. Thus, our retains intuition interpretation yet satisfies positive-definite condition as found constant-correlation BEKK models. report Monte Carlo results finite-sample distributions MLE varying-correlation model. The applied real data sets. It that extending allow for correlations provides interesting time histories are not available