Network Structural Characteristics of Insurance Institutions and Systemic Risk Contagion

Abstract

The increasing systemic importance of the insurance sector implies that risks originating within it can propagate across industries through complex interconnections, potentially exerting significant impacts on China’s broader economy. This study investigates the transmission and spillover mechanisms of intra-industry risk among insurance-related institutions from a network perspective. Specifically, a DCC-GARCH-CoVaR framework is employed to quantify risk spillovers among six listed financial institutions engaged in insurance-related businesses. Based on forecast error variance decomposition, an insurance network is constructed and further examined using network topology analysis to characterize its structural properties. The findings indicate that, despite gradual improvements in resilience and stability across institutions, network structural characteristics remain critical drivers of the formation and diffusion of systemic risk. Highly connected and central institutions are more likely to act as contagion hubs, facilitating rapid outward risk transmission, particularly in densely connected networks with shorter path lengths. Moreover, a nonlinear relationship exists between network structure and systemic risk, with leverage growth playing a moderating role. Elevated leverage amplifies the risk-enhancing effect of network centrality, while under certain conditions, mitigating adverse effects associated with passive risk exposure. Additionally, capital adequacy conditions shape risk dynamics: well-capitalized institutions can utilize network connections for risk diversification, whereas undercapitalized institutions tend to intensify contagion. Significant heterogeneity is also observed in the effects of coreness, density, and path length across institutional groups.