Dynamic patterns and the latent community structure of sectoral volatility and jump risk contagion

Abstract

This study explores the latent driving structure of sectoral risk contagion in terms of volatility and jump through the factor model framework for high-dimensional matrix-valued time series. We find strong inter-sector volatility and jump spillover effects at the mean level and higher effects at extreme tails, while jump spillovers reflect sharper structural shifts with the promotion of major financial or public events. Both volatility and jump are contagious in the form of a community-structure. Our findings provide new evidence about sectoral risk contagion and the importance of accurately identifying the underlying structure of risk contagion for efficiently realizing hierarchical regulation.