Robust M-Estimation of Multivariate GARCH models

In empirical work on multivariate financial time series, it is common to postulate a Multivariate GARCH model. We show that the popular Gaussian quasi-maximum likelihood estimator of MGARCH models is very sensitive to outliers in the data. We propose to use robust M-estimators and provide asymptotic theory for M-estimators of MGARCH models. The Monte Carlo study and empirical application document the good robustness properties of the M-estimator with a fat-tailed Student t loss function and volatility models with the property of bounded innovation propagation