Measuring the contribution of Chinese financial institutions to systemic risk: an extended asymmetric CoVaR approach

This study proposes an extension of the Asymmetric CoVaR method in Espinosa et al. (J Bank Finance 58: 471–485, 2015) to capture the time-varying asymmetric responses of the financial system to positive and negative shocks to individual institutions. Building on the extended method and considering a set of Chinese financial institutions, we assess the extent to which distress within different institutions contribute to systemic risk. To provide a formal ranking of risk contributions, we implement the significance and dominance tests with bootstrap Kolmogorov–Smirnov statistics. The estimates of the extended Asymmetric CoVaR method reveal an asymmetric pattern that characterizes the tail interdependence in the Chinese financial system and this pattern changes dynamically over time. Particularly, the impact on the system of a fall in individual market value is only slightly larger than that of an increase during tranquil years. However, the entire system becomes extremely sensitive to downside losses than to upside gains during crises. The result also raises concern about privately owned banks in that they are systemically riskier than state-owned banks and other institutions. Using panel regressions, we also find firm characteristics such as institution size and volatility are important predictors of systemic risk contribution.