The January 2025 wildfires in Los Angeles turned into an unplanned stress test for how insurers measure wildfire risk, according to analysis from Delos Insurance Solutions.
Insurers have paid more than $22.4 bn on wildfire claims tied to the Los Angeles wildfires that erupted on Jan. 7 last year, according to figures released by the California Department of Insurance.
Two major fires, Palisades and Eaton, destroyed more than 16,000 structures during intense Santa Ana wind conditions. A third blaze, Sunset, ignited nearby under nearly the same regional weather pattern.
It burned roughly 40 acres and was contained within hours. Same season. Same region. Very different outcomes.
Delos says the divergence had little to do with chance. Weather mattered, but it wasn’t decisive. The difference came from fire physics and landscape conditions.
Palisades and Eaton ignited inside large, continuous fuel blocks on steep terrain that hadn’t burned in decades. Mature vegetation produced heavy ember loads. Fire lines expanded faster than crews could contain them.
Kevin Stein, CEO of Delos Insurance Solutions, said the lesson was blunt. Wildfire risk isn’t defined by whether ignition happens. It’s defined by whether the landscape allows a fire to accelerate beyond control.
Stein said Delos built its wildfire risk model to separate those outcomes, a distinction he said matters for communities, insurers, and coverage access.
Delos argues risk assessment needs to move away from ignition probability and parcel-level snapshots. The real question sits upstream. Does the surrounding landscape support fires that escape initial attack and turn into urban disasters.
According to Beinsure, that framing aligns more closely with loss behavior seen in recent California fire seasons.
Two other Los Angeles County fires reinforced the point. Kenneth and Hurst ignited under difficult conditions and still caused no structure losses.
Both burned through terrain already scarred by multiple fires over the past 20 years. Reduced shrub height and thinner tree cover limited ember generation and gave suppression crews time to work. Prior fire history, in this case, acted as a brake rather than a predictor.
Delos says these outcomes expose blind spots in many wildfire models used across the insurance market. Traditional approaches often lean on historical burn frequency or property-level characteristics.
