We propose a model that can capture the typical features of multivariate extreme events observed in financial time series, namely, clustering behaviors in magnitudes and arrival times of multivariate extreme events, and time-varying dependence. The model is developed within the framework of the...