Understanding the Hidden Risks in Climate Change Modeling for Insurers

The insurance industry’s approach to adjusting catastrophe models for climate change, particularly in terms of frequency and severity of extreme weather events, is facing a critical challenge. While efforts are ongoing to adapt these models to account for the changing climate, a significant issue lies in the lack of attention to scenario completeness, especially concerning the most severe impacts expected in the tail of the distribution. This oversight could lead to a significant underestimation of physical risks.

Recent trends have seen insurers modify their catastrophe model outputs to include climate change impacts, driven partly by new regulatory requirements. A key research reference for this is Knutson et al. (2020), which focused on the expected changes in global tropical cyclone activity under a 2°C warming scenario. However, applying these adjustments, insurers have noticed an inconsistency: while short return period losses increase substantially, tail losses experience a relatively smaller increase. This discrepancy is due to the fact that shorter return periods make up the majority of the loss distribution.

The real concern is that traditional models, including catastrophe models, struggle with fat-tailed events, which means they likely miss key aspects of risk, especially in the context of climate change. These include potential tipping points, feedback loops, and systemic risks. There’s a growing realization that insurers need to apply the same rigor in evaluating and quantifying non-modeled risks, such as those under Solvency II, to their climate change adjustments and scenarios.

Moreover, the increased likelihood of co-occurring hazards, such as extreme winds and precipitation, and the potential triggering of tipping points like ice sheet collapse or permafrost melting at a 1.5°C global mean temperature rise, pose significant challenges. These events could have far-reaching consequences, not just for the insurance industry but for society as a whole.

Finally, the indirect effects of climate change, like supply chain disruptions, food insecurity, geopolitical conflicts, and infrastructure failures, are often overlooked. These can lead to systemic effects, stressing global economies. To address these challenges, the concept of "climate loss amplification" (CLA) could be introduced, similar to post-event loss amplification (PLA) used in current catastrophe modeling, to account for complex sources of tail loss, including socio-economic effects.