The insurance industry's struggle to accurately model wildfire risk as unseasonal January fires in Los Angeles County contributed to global insured catastrophe losses reaching $80 billion, the second-highest first-half total on record, according to Swiss Re Institute's latest sigma report.
The modeling challenges are particularly acute for wildfires, which Swiss Re describes as fundamentally different from other natural perils.
"Wildfire is a complex peril to model. Human actions influence the occurrence and damage potential of wildfires to a much greater extent than other perils," the report states. "To date traditional risk evaluation methods – such as predictive or probabilistic models – have not been fully captured the complex interactions between drivers of wildfire risk, changes in fire behaviour and inter-annual variability."
In July 2025, the California Department of Insurance completed its review of the Verisk Wildfire Model for the United States—the first catastrophe model ever approved for insurance ratemaking in the state's history.
The agency is also reviewing models from Karen Clark & Company (KCC) and Moody's RMS, with KCC's model having completed the review process in early August 2025. Insurance JournalInsurance Journal This regulatory shift represents a dramatic departure from California's previous approach and comes as insurers have fled the state's wildfire-prone markets, leaving hundreds of thousands of homeowners dependent on the state's insurer of last resort.
The January 2025 Palisades and Eaton wildfires in LA County alone accounted for $40 billion in insured losses, close to 70% of global first-quarter losses and half of the entire first-half total. What made these fires particularly challenging for risk models was their timing: they occurred during California's typically wet winter season, when fire risk is historically lowest.
"Based on sigma records, 99% of the wildfire losses that originate in California occur in the second half of the year," the report notes, underscoring how the January fires defied traditional seasonal patterns that underpin many risk models.
The modeling difficulties extend beyond wildfires to severe convective storms (SCS), another major loss driver that generated $31 billion in first-half insured losses, according to the reinsurer.
While SCS losses were below trend, they still ranked as the fourth costliest on record for a first half-year period, highlighting the persistent challenges in predicting these events.
The report identifies several factors that complicate accurate catastrophe modeling.
For wildfires, rapid exposure growth in wildland-urban interface (WUI) zones has outpaced traditional risk assessments. "Since 1990, exposure growth in the high-risk wildland urban interface (WUI) zones has outpaced exposure growth in non-WUI zones by a factor of 1.8 in the US, and by a factor of 1.9 in California," according to Swiss Re's analysis.
Climate change is adding another layer of complexity to risk models. The report warns that "with changing climates, unexpected and unseasonal weather conditions may occur more frequently, making loss outcomes more volatile and difficult to predict."
This volatility is exemplified by the timing of the LA County fires, which occurred when California should have been experiencing its wet season.
For severe convective storms, post-pandemic inflation has significantly altered the loss landscape in ways that models are still catching up to understand.
Construction costs in the US rose 35.64% from January 2020 to June 2025, directly impacting property claims costs and making historical loss data less reliable for future projections.
The report also highlights how human development patterns are creating new vulnerabilities that traditional models struggle to capture. The expansion of communities into wildland areas increases ignition sources and creates more pathways for fire spread, while new technologies like rooftop solar installations are adding to exposure vulnerability for severe weather events.
Swiss Re says that "wildfire remains a volatile peril and in our view, more and rising associated insured losses can be expected." This assessment reflects the industry's acknowledgment that current modeling approaches may be insufficient for capturing the full scope of evolving wildfire risk.
The implications for the insurance industry are significant. With global catastrophe losses growing at a long-term trend of 5-7% annually, and first-half 2025 losses already exceeding the midpoint of Swiss Re's full-year projection of $150 billion, insurers face mounting pressure to develop more sophisticated modeling approaches that can account for climate-driven changes in risk patterns.
