Cross-industry contagion of systemic financial risks from the perspective of dynamic tail risk network: Evidence from China

Abstract

Fluctuations in interconnectedness within different industrial sectors are hence critical not only for financial market stability but also for the development of real economy. Thus, this paper contractures the varying network structure and estimates the tail risk spillover linkages at the industry level by applying the TENET (Tail-Event driven NETwork) framework on China’s 24 sectors from 2007 to 2018. There is a higher level of risk spillovers and spatial linkages between sectors during turmoil periods, and the centrality indicators can effectively identify the systemically important industries. Notably, it is more comprehensive and accurate to judge the systemic importance node by considering its location and connection in the network rather than just its internal attributes. Furthermore, the high-order polynomial regressions show that degree and eigenvector centrality have a significantly positive effect on risk contagion, while closeness centrality has a two-regime (“both robust and fragile”) contagion mechanism. From a policy point of view, regulators could formulate early-warning measures based on the impact mechanism and spread paths of systemic risk, especially in financial turmoil periods. Meanwhile, they should pay more attention to preventing the aggregate risks of the most influential sectors.