Common features in volatilities: a new Multiplicative Error Model

Multivariate volatility models could consider the influence of each volatility series on the others (spillover effects). Furthermore, integrating financial markets provides similar dynamics (co-movements). We propose a new model for volatility vectors, belonging to the family of Multiplicative Error Models (MEMs), which incorporates spillover and co-movement effects captured in a separate component. Moreover, to reduce the number of coefficients for high-dimensional datasets, we propose a simple model-based clustering procedure. We apply the model to a set of 29 assets included in the Dow Jones Industrial index, providing the interpretation of spillover effects and co-movement. The proposed parameterization shows a satisfactory performance when compared to other vector MEMs.