European wildfire insured losses have climbed an estimated 8% to 11% a year in real terms since 1970, according to Swiss Re Institute researchers. In that same period, wildfire has remained classified as a secondary peril in reinsurance risk models - alongside floods and hailstorms - a category conventionally used for perils considered frequent but individually less severe than primary events like major hurricanes. Morningstar DBRS says that classification looks increasingly out of date, even as H1 2026 catastrophe losses came in well below their historical average.
The mismatch between how the market prices wildfire and what wildfire is actually costing is the structural tension behind a strong set of H1 2026 reinsurance results - and it is one that softening property catastrophe rates are making more acute rather than less.
Swiss Re Institute put H1 2026 insured natural catastrophe losses at $42 billion, 16% below the 10-year average and the lowest first-half total since 2020. Total insured losses including man-made events reached $48 billion against a $56 billion average. Morningstar DBRS puts global insured catastrophe losses at roughly $46 billion for the half, well under the 10-year H1 average of $64 billion. Economic losses landed around $142 billion against a $159 billion average. The industry has gone five consecutive quarters without a single catastrophe topping $10 billion in insured losses - a run that has not been sustained since the mid-2010s soft market.
Severe convective storms in the US - thunderstorms, hail, tornadoes, derechos - still produced more than $26 billion in insured losses over the same period. Secondary perils as a whole accounted for 92% of insured catastrophe losses worldwide in 2025, with severe convective storms alone generating $51 billion, the third-costliest year on record for that peril. That is the context in which reinsurers are posting their strongest combined ratios in years: a benign primary peril environment masking a secondary peril cost base that is structurally rising.
Monica Ningen, who leads property and casualty reinsurance for Swiss Re in the US, said wildfire and severe convective storms "are no longer 'secondary' in any practical sense." She pointed to population growth in high wildfire-risk areas running at roughly three times the national average and to the fact that around 85% of US wildfires are started by people - a pattern that makes wildfire risk a land-use and behavioural problem as much as a climate one, and therefore both harder to predict and harder to price accurately from historical loss data alone.
"Hotter and drier conditions are making large wildfires more likely, and, with more homes, businesses and infrastructure built in risk-exposed areas, also more costly," said Balz Grollimund, Swiss Re's head of catastrophe perils.
The classification problem runs deeper than semantics. Secondary perils are typically modelled differently from primary perils - with less granular historical data, wider confidence intervals around expected losses, and less developed probabilistic modelling frameworks. If wildfire is now generating losses at a scale and frequency that puts it closer to a primary peril in practice, the models that price treaty programmes and set attachment points are working from assumptions that understate the risk. In a softening market, that gap compounds: rates are falling at precisely the moment the underlying loss trend is rising.
None of that stopped earnings from rising. Eight of the largest global P&C reinsurers - Munich Re, Hannover Re, Swiss Re, SCOR, Everest Re Group, AXIS Capital, Arch Capital and RenaissanceRe - posted combined net income of $14 billion for H1 2026, up 16.7% from $12 billion a year earlier, per Morningstar DBRS. The group's average combined ratio dropped to 79.8% from 87.6% in H1 2025.
At the individual carrier level: Munich Re improved from 71.8% to 67.9%, Swiss Re from 84.5% to 76.7%, Hannover Re from 87.8% to 83.2%, and SCOR from 87.0% to 79.9%. Everest Re improved to 87.8% from 94.3%, AXIS Capital to 93.6% from 92.1%, Arch Capital came in at 76.6%, and RenaissanceRe fell sharply to 72.9% from 103.3% the year before - the largest single-carrier improvement in the group, reflecting California wildfire losses that severely burdened RenaissanceRe's 2025 result.
Property reinsurance pricing continued softening as capital accumulated in the sector and competition sharpened. Guy Carpenter's Global Property Catastrophe Rate-on-Line Index showed pricing down around 16% globally following the mid-year renewals - the steepest annual fall since the late 1990s, and sharper than anything seen during the soft market of the 2010s. "In some cases, risk-adjusted decreases have deepened since January 1, 2026 renewals," the broker noted, "and property catastrophe rate on line remains down globally, around 16%."
Rather than chase volume at softer prices, reinsurers largely reduced participation in business that missed their return thresholds, a stance Morningstar DBRS expects to hold through the rest of 2026. Casualty and specialty lines stayed profitable but faced pressure from social inflation, reserve uncertainty, cyber pricing, and geopolitical risk. Conflict in the Middle East pushed underwriters to watch exposure in marine, aviation, energy, political violence and trade credit more closely, even though losses tied to those events remained manageable in aggregate.
Investment income remained a significant earnings contributor across the group. Swiss Re pointed to higher recurring investment income and better reinvestment yields. Munich Re cited fair-value gains and equity market performance. RenaissanceRe pointed to growth in net investment income, and Arch Capital to a larger invested asset base and gains from equity-method investments. Aggregate invested assets and total investment income for the eight companies both rose from H1 2024 through H1 2026.
Share buybacks stayed a favoured use of capital. Arch Capital ranked among the most active repurchasers of its own stock, and RenaissanceRe continued its own programme.
Morningstar DBRS expects the sector to remain well capitalised through the second half of 2026, with underwriting discipline the real test as competition and available capacity grow. That framing is correct but understates the specific wildfire problem. A market that is simultaneously lowering property catastrophe attachment points, cutting rates at the steepest pace since the 1990s, and underpricing a peril whose annual loss growth rate has run at 8% to 11% in real terms for more than five decades is accumulating a gap between modelled and actual exposure that a single active wildfire season could expose.
For cedants and primary carriers buying property catastrophe reinsurance in the current environment, the practical implication is to review whether treaty wordings specifically address wildfire as a defined peril, whether attachment points reflect updated wildland-urban interface exposure rather than historical loss assumptions, and whether the coverage purchased at declining rates actually provides the protection the programme was originally designed to deliver. In a market where reinsurers are being selective about the business they write, those reviews are best conducted before renewal rather than after a loss demonstrates what the programme does and does not cover.
