Wildfires and severe convective storms (SCS) dominated North America’s natural catastrophe losses in 2025. Total insured losses surpassed USD 90 billion, below recent years but still reflecting a rising underlying risk trend. According to sigma data, wildfires alone caused a record USD 40 billion in insured losses, while SCS generated USD 46 billion, making 2025 the third-costliest year on record for this peril. Together, these two perils produced the highest annual aggregate insured losses ever recorded from secondary perils in the region. Losses for all other perils were muted. The overall insured share of losses was unusually high at 71%, reflecting the composition of events, with wildfire and SCS typically having high insurance penetration in the US. In North America, wildfire losses are rising much faster than economic growth, with exposures increasingly growing in high-risk areas and underlying risk factors also trending upward.

North America – comprising the USA and Canada – recorded natural catastrophe-related insured losses of USD 90 billion in 2025, down from USD 117 billion in 2024 and below the previous five-year average of USD 97 billion. The below-trend result in 2025 does not signal a reduction in underlying risk. Year-to-year volatility reflects natural variability in weather conditions, and the underlying trajectory of annual insured losses remains upward. The trend reflects structural changes in that the baseline loss level is rising as loss pressure extends beyond the classic US tropical cyclone and earthquake corridors.

Tropical cyclones continue to define the region’s most severe tail-loss scenarios, but secondary perils are broadening the range of hazards with significant loss potential and pushing annual losses higher. In the absence of a major US hurricane landfall event, wildfires and SCS collectively were the main loss-generating perils last year. This came despite an active North Atlantic hurricane season: the year 2025 marked only the second time on record (after 2005, the year of Hurricane Katrina), that there were three or more Category 5 hurricanes in the basin.

Insured losses by peril in 2025 and for previous past year averages,(%) in North America

North America is a vast continent and has a great diversity of climatic regimes. There is high exposure to a wide range of loss-inducing natural hazards. Given high insurance penetration and high-value assets in the US, most of last year's global insured losses originated from there (USD 88.9 billion). It was a relatively benign year in Canada: last year's insured losses of USD 1.5 billion were less than half of the previous five-year average of USD 3.2 billion, and well down on the record annual high of USD 6.5 billion in 2024.

The largest loss-making disasters of the year in North America were the Palisades and Eaton wildfires in Los Angeles County. Combined, the said wildfires resulted in insured losses of USD 40 billion, by far the largest global insured wildfire loss event on sigma records. SCS-induced insured losses were third-highest for the region. Together, these events lifted the share of secondary perils in total losses to an all-time high of 99.9%.

The USD 40 billion insured losses from the January wildfires in southern California were driven by the burning of some of the highest-value and most densely built wildland-urban interface (WUI) areas in the US. The wildland-urban interface is the zone where built-up areas meet or mix with natural vegetation. Because structures and wildland fuels are in close proximity, these areas face elevated wildfire risk. Strong Santa Ana winds spread embers into tightly packed neighborhoods, turning wildland fires into urban conflagrations. The biggest of the fire outbreaks burned close to40 000 acres in Los Angeles County. The fires destroyed more than 16 000 structures in the Palisades and Eaton neighborhoods, nearly three times the average annual number since 2016, even though 2025’s burned area in California was only about one-third of the post-2016 average.¹

Over the past 55 years, insured wildfire losses in North America have grown about twice as fast as exposure.² Exposure-related factors (construction cost, population at risk, and economic growth per capita) together explain a little more than one-third of wildfire loss growth. The remaining roughly 60% points to changes in the underlying risk landscape. In the case of wildfire risk, this is driven by the strong increase of high-value assets in high-risk regions, changes in weather conditions (drought and strong winds) and sources of fuel (flammable structures linked by dry grassland that provide a pathway for fire).

Insured loss growth from wildfire in North America since 1970, decomposed into key exposure drivers, insurance effects and residuals

The single biggest driver of insured losses caused by wildfire is exposure growth in high fire-hazard regions, in particular the WUI. In relative terms, population growth since 1975 in high wildfire-risk areas of the US has been three times higher than overall across the country.³ Although in absolute terms most population growth has been in moderate fire risk zones, about 10% was in regions of high to extreme wildfire risk. This matters because faster expansion in the areas most at risk of wildfires leads to a disproportionately large increase in overall risk.

With population growth comes exposure growth, that is accumulation of property and other assets that need insuring. Since 1990, exposure growth in WUI zones has outpaced that in non-WUI zones by a factor of 1.8 in the US, and by a factor of 1.9 in California.⁴ As of 2020, about one in three Californians resides in the WUI. The January 2025 fires burned exclusively in Los Angeles' WUI.⁵

In the case of wildfire risk, population expansion brings another consideration with respect to exposure growth. Humans cause about 85% of US wildfire ignitions,⁶ creating a self-reinforcing feedback loop. As populations in the WUI increase, both the exposure and the hazard (ignition risk) are higher. This human-based component poses challenges for traditional catastrophe modelling on account of the complex interactions between hazard and exposure.

At the same time as exposure has increased, shifts in seasonal conditions and rainfall patterns have contributed to drier, more flammable fuels, increasing ignition and fire spread potential. Fire-conducive weather characterized by low humidity and strong winds creates conditions that favor wildfire ignition and rapid spread.⁷ Historically, California has had wet winters. However, in recent years, the fire seasons have been longer with greater overlap between the period of peak fuel dryness and the peak Santa Ana wind season in December-January. The January 2025 fires were unusual in their timing as California’s largest wildfire losses typically occur in late summer/autumn. The winds played a significant role in driving the losses higher.

The magnitude of the residual suggests wildfire is the peril with the clearest implied hazard-driven signal in historical insured loss data. The residual reflects both increases in fire-conducive conditions and disproportionate exposure growth in high-risk areas. The January 2025 wildfires and observed loss trends highlight the growing importance of mitigation and adaptation in supporting long-term resilience in the years ahead.

SCS-insured losses reached USD 46 billion in 2025, extending the run of years with losses above USD 45 billion. This was the third-costliest SCS year after 2023 and 2024 (USD 59 billion and USD 52 billion, respectively, at 2025 prices). This peril is the second fastest growing after wildfires. The number of SCS events causing average annual losses of USD 1 billion or more in the five years to 2025 was 59% higher than in the five years to 2020.

From an industry perspective, the term SCS is used to describe damage caused by hail, tornadoes and straight-line winds. The US is particularly exposed to SCS because its geography provides favorable conditions for their formation. Severe storms develop when key atmospheric ingredients such as moisture, instability, upward motion of air, and wind shear come together. In the US, these ingredients frequently align as warm, moist air from the south interacts with colder and drier air from the north and west, creating unstable conditions.

Breakdown of growth in SCS-insured losses in North America since 1970, by key exposure drivers, insurance effects and residuals

Insured losses from SCS in North America have risen by 7% annually since 1970. The drivers of rising SCS insured losses growth in North America are structural and imply increasing average annual losses, as well as a materially higher annual loss in an extreme season. The increase is largely explained by exposure growth, including rising asset values and expanding development in exposed areas (80% of the SCS-associated loss growth). In the US, urbanization and sprawl are increasingly concentrating high-value, damage-prone assets in hail belts. Changes in insurance coverage only play a minor role. The remaining residual, albeit small, may reflect a combination of changes in vulnerability and hazard intensification. For instance, ageing housing stock increases susceptibility. The median age of homes nationally has risen over the past decades.⁸ Older homes are typically more vulnerable to hail and wind damage, with roof damage being the main contributor to insured losses. Vulnerability has risen with rooftop solar and other exterior equipment. US property reconstruction costs, though flat since March 2025, are still 37% above December 2019,⁹ impacting property claims costs.¹⁰

Floods (excluding tropical cyclone-induced flooding) contribute comparatively modestly to overall natural catastrophe driven insured loss growth in North America. Current estimates suggest that flood-related losses account for around 3% of all-peril insured losses for the region, with annual growth rates of approximately 4% in real terms, much less than SCS and wildfires. Most of this growth is down to exposure factors, that is rising asset values and continued expansion into flood-prone areas. The residual component capturing changes in hazard and vulnerability is near zero or slightly negative. This pattern suggests that existing protection and mitigation measures like flood defenses help offset underlying risk and contain losses. As a result, compared with other secondary perils in North America, flood appears to be more effectively managed from a loss-growth perspective.

Looking ahead, hazard intensification in the US is projected to increase average annual flood economic losses by 1% per year to 2050, with coastal and pluvial (surface water) flooding to be most impacted. Fathom, a Swiss Re-owned company specializing in water risk, projects the 1-in-250-year tail loss to grow approximately 33% more slowly than average annual flood losses.¹¹

Break down of insured loss growth from floods in North America since 1970, by key exposure drivers,  insurance effects and residuals

Adaptation and mitigation measures are crucial to reducing loss potential and maintaining insurability in exposed regions. Our findings indicate that adaptation can materially offset loss growth when implemented effectively. Evidence from the US shows that targeted adaptation can deliver measurable benefits. For example, in Alabama a coordinated strategy combining the Insurance Institute for Business & Home Safety’s (IBHS) Fortified Building Code Standard with premium discounts, public retrofit grants and independent third-party verification, has led to materially lower insurance losses. Homes built or retrofitted to resilience standards such as the FORTIFIED program have seen claims frequency decline 55–70% and severity by 14–40%, with loss ratios up to 72% lower than conventional construction.^12,13

Well-defined adaptation measures can translate directly into insurance benefits. In Louisiana, insurers offer a 7‒24% discount on annual homeowners insurance premiums for homes retrofitted to the IBHS Fortified Building Code Standard demonstrating that risk-based pricing can reward resilience measures. Retrofitted homes in higher-risk coastal areas typically receive a larger discount than homes in interior regions.¹⁴ Such measures show that when mitigation is measurable, enforceable and transparent, insurers can incorporate it into pricing, aligning incentives with societal risk reduction. Institutional measures also matter recent legal and regulatory reforms in Florida have reduced loss amplification and supported market stability.