Systemic risk in financial institutions: A multiplex network approach

Abstract

We examine the systemic risk spillovers of Chinese financial institutions using a multiplex network. A multiplex network is based on different dependency measures with dynamic conditional correlations and by a planar maximally filtered graph. We find that more information is included in a multiplex network than in single-layer networks. The systemic importance of institutions varies over time, which would be ignored without consideration of various inter-agent correlations. The results show that an individual institution's contributions to the system significantly diminish as the node strength increases. The greater the closeness centralities of nodes, the more vulnerable the system is to shocks.