Spatial Multivariate GARCH Models and Financial Spillovers

We estimate the risk spillover among European banks from equity log-return data via Conditional Value at Risk (CoVaR). The joint dynamic of returns is modeled with a spatial DCC-GARCH which allows conditional variance log-returns each bank to depend on past volatility shocks other and their squared in parsimonious way. backtesting resulting measures provides evidence that (i) multivariate GARCH model Student’s t distribution more accurate than both standard Gaussian Filtered Historical Simulation (FHS), (ii) introduction component improves assessment profiles market spillovers.