When climate risk affects the systemic risk of the insurance sector: ∆CoVaR forecasting

Abstract

Climate change is increasingly recognized as a potential source of financial instability, yet its implications for systemic risk in the insurance sector remain insufficiently understood. This study combines a Copula–DCC–GARCH framework with the ∆CoVaR measure to examine whether climate-related catastrophic events are associated with insurers’ contributions to systemic risk. Using weekly data for the ten largest global insurers over the period 2004–2024, we compare ∆CoVaR estimates across catastrophe and non-catastrophe periods under normal and crisis market conditions. The results indicate that catastrophe occurrence exhibits only a limited association with insurers’ systemic-risk contributions during normal market conditions but is associated with substantially stronger contributions during periods of financial stress. The analysis also reveals considerable heterogeneity across globally significant insurers, indicating that the magnitude of this association differs markedly across companies. To assess the robustness of these findings and provide a forward-looking perspective, we employ Generalized Additive Models for Location, Scale and Shape (GAMLSS). The scenario-based forecasting framework shows how insurers’ expected systemic-risk contributions evolve under alternative market and catastrophe conditions and consistently identifies the crisis–catastrophe scenario as being associated with the strongest deterioration in forecasted ∆CoVaR. Overall, the findings demonstrate that the relationship between climate-related catastrophic events and insurers’ systemic-risk contributions is strongly state-dependent. By combining dynamic systemic-risk measurement with scenario-based forecasting, the proposed framework provides a forward-looking assessment that extends previous descriptive evidence and offers potentially useful insights for climate-sensitive systemic-risk monitoring.