November 1, 2026
Marsh Re’s Randy Fuller, Josh Darr and Kim Roberts explain how new data sources are augmenting vendor model outputs and creating more refined views of risk
In an era of increasing climate volatility and rising insured losses, carriers face a critical question: are out-of-the-box catastrophe models telling the whole story about their exposure? Increasingly, the answer is no. While catastrophe modeling remains essential to risk evaluation, new data sources, advanced analytics and industry experience now make it possible — and necessary — to build a more refined view of risk.
Evaluating catastrophe risk is a complex process that requires a granular understanding of both the structures being insured and the hazards that make them subject to loss. The major catastrophe model vendors have added tremendous value over the years, but the market now has access to more data and more tools that can supplement model output and help companies develop a clearer picture of their portfolios.
At Marsh Re, we see three key areas where carriers can strengthen their understanding of both hazard and risk characteristics.
First, catastrophe model adjustments can help refine out-of-the-box views of risk. Industry data or company experience can be used to adjust vendor model output when there is evidence of a disconnect between modeled results and observed trends or performance. We have developed recommended adjustments for certain perils where those disconnects appear meaningful, including recent vendor model releases for US hurricane, severe convective storm and wildfire.
Second, hazard scoring by peril can create a more granular view of risk than vendor models typically provide. That added precision can support underwriting selection and pricing, by leveraging more up-to-date and granular weather data than is typically included in the catastrophe models.
We have seen companies optimize portfolios by incorporating hazard scores into underwriting and multivariate pricing models. Demonstrating the power of hazard scoring, Marsh Re’s wildfire scores successfully identified the majority of structures exposed to the Los Angeles wildfires as being within the top 5% of risk for wildfire, with many in the top 1%.
Third, granular structure details available through data enrichment are materially improving portfolio insight. Advances in AI have made data on physical property features — such as roof age, roof shape, tree overhang and effective year built — more accessible than ever before. That increased granularity is improving data accuracy across the industry by reducing reliance on insureds and agents to provide precise property details. It also creates efficiencies across the value chain, from application experience to underwriting selection and pricing, risk management, modeling and reinsurance purchasing.
Marsh Re helps companies bring these elements together to build and present a view of risk that better describes their portfolio to the market. In a marketplace that increasingly values clarity, carriers need more than a single standard model output — they need a view of risk that is evidence-based, defensible and aligned with their actual business.
Now is the time to reassess whether your current view of risk is telling the full story. A more refined approach can help companies better understand, manage and communicate catastrophe exposure.