Modelling the Spillover Effects of Systemic Financial Risk based on Dynamic CoVaR Method

Abstract

Systemic financial risk spillover has become a core concern in global financial stability and macroprudential regulation, as the interconnectedness of financial institutions and markets amplifies the transmission of extreme tail risks across the entire financial system. Traditional Value-at-Risk (VaR) only measures individual entity risk and fails to capture risk interdependence and spillover effects, while the static Conditional VaR (CoVaR) model cannot reflect the time-varying characteristics of financial risk spillovers under volatile market conditions. This paper constructs a dynamic CoVaR model integrating GARCH volatility clustering and dynamic conditional correlation (DCC) to quantify the time-varying spillover effects of systemic financial risk, with a focus on measuring the marginal risk contribution of individual financial institutions to the overall system. Using daily return data of 12 major listed financial institutions in China (covering banking, securities, and insurance sectors) from January 2018 to December 2023, this paper conducts empirical analysis, calculates dynamic ΔCoVaR to measure spillover intensity, and compares risk spillover differences across financial subsectors. Results show that large state-owned commercial banks have the strongest systemic risk spillover effects, followed by securities companies, while insurance institutions have relatively weaker spillovers; risk spillovers surge significantly during crisis periods such as market turbulence and liquidity shocks, verifying the asymmetric and time-varying nature of systemic risk transmission. The dynamic CoVaR model outperforms static models in fitting accuracy and early warning efficiency, providing a reliable quantitative tool for macroprudential supervision and risk prevention.