Swiss Re Institute: Secondary Perils Caused a Record 92% of 2025's $107 Billion in Insured Catastrophe Losses, and a Trend Year Would Cost $148 Billion

Natural disasters cost insurers $107 billion in 2025, according to the Swiss Re Institute's annual sigma catastrophe study, and almost all of that came from the kinds of events insurers once treated as minor. The Institute reports that secondary perils, a category that includes wildfires, severe convective storms and floods, made up a record 92% of the global insured natural catastrophe losses. This article is based on Swiss Re's press release for the study and on the summary on the Institute's own website. An earlier version of this article described the wildfire growth rate as about 12% a year and gave a figure of $320 billion as a projection. The press release gives 14% for North American wildfire losses and presents $320 billion only as a modelled peak scenario, so the text has been corrected.

The headline figures are as follows. Insured losses from natural catastrophes were $107 billion. Economic losses, which include uninsured damage, were $220 billion, and about 49% of these were insured, the highest insured share in the history of the sigma series. Adding man-made losses of $13 billion brings total insured losses to $120 billion, and total economic losses to $235 billion, compared with $338 billion in 2024 and a previous ten-year average of $280 billion. In other words, 2025 was cheaper than 2024 on every measure, but Swiss Re cautions against reading too much into that.

The reason for caution is stated by Balz Grollimund, the Institute's head of catastrophe perils. He said the relatively low losses reflected favourable variability rather than any easing of underlying risk. The page on the Institute's website adds that the long-term trend would imply about $140 billion for 2025, so the actual outcome came in below it. One important piece of luck was the absence of a major hurricane landfall in the United States in 2025. Tropical cyclones are the largest contributor to long-term average losses, so a year without one produces a lower total.

The perils that did cause losses were secondary perils. The Los Angeles wildfires of January 2025 cost about $40 billion in combined insured losses, the largest insured wildfire loss event in sigma records. Severe convective storms, which include hail, tornadoes and thunderstorm wind, caused $51 billion, the third-costliest year on record for that peril after 2023 and 2024 when measured in 2025 prices. Flood losses were relatively small, at $3.4 billion against a previous five-year average of $15.4 billion. Together, wildfires, severe convective storms and floods caused 88% of the global insured natural catastrophe losses, while the 92% figure covers all secondary perils.

The study looks at why losses keep rising. Swiss Re says long-term insured natural catastrophe losses grow at a rate of 5% to 7% a year. Exposure growth, meaning more and more valuable property in areas at risk, explains more than 80% of the long-term global increase in weather-related insured losses between 1970 and 2025. Severe convective storms account for 38% of the growth in historical losses, wildfires for about 20% despite having the highest growth rates, and floods for roughly 10%. Jérôme Jean Haegeli, who heads the Institute and is the group's chief economist, summarised the problem by saying that more valuable property is being built in harm's way, and rebuilding costs have risen.

Regional patterns differ. In North America, the growth in losses is driven mainly by wildfires and severe convective storms, and the Institute says insured wildfire losses grow at 14% a year, as longer fire seasons and changes in temperature and rainfall add to the threat. In Europe, over half of the growth in insured losses comes from severe convective storms, growing at about 10% a year, and less than half of that increase is explained by exposure growth, which suggests that storm characteristics and vulnerability are changing. In Asia, the growth in secondary-peril losses is dominated by floods. In Oceania, severe convective storms and floods share the growth more evenly, and wildfires play a smaller part.

For 2026, Swiss Re gives two reference points. If losses return to the normal long-term trend, insured losses would be $148 billion. In the modelled peak-loss scenario, which represents a severe year and not a forecast, losses could reach about $320 billion. The Institute stresses that its trend figure is not a prediction of what will happen. Urs Baertschi, who runs property and casualty reinsurance at Swiss Re, noted that the company protects against peak risks by absorbing low-frequency, high-severity events, which is the function of reinsurance.

There is also a gap between damage and protection. In emerging economies, typically 80% to 90% of catastrophe losses are uninsured. When a disaster strikes in a country with low insurance penetration, the cost falls on households, businesses and governments instead of insurers. Swiss Re recommends sustained and well-designed adaptation and risk mitigation to maintain insurability and affordability, identifying the drivers of rising losses so that risks can be managed before losses occur, and pairing stronger risk management with broader and more accessible insurance in emerging economies.

What does this mean for ordinary readers? Property insurance prices reflect the expected cost of events like these, so rising catastrophe losses are one reason premiums for homes in wildfire, hail and flood areas keep increasing. Measures that reduce damage, such as fire-resistant roofing, defensible space around homes, hail-resistant materials and flood barriers, matter for insurers as well as for owners, because they can improve availability and prices. For governments, building codes and land-use decisions determine how much exposure grows. The study makes the case that insurance, adaptation and planning have to work together, because none of them is sufficient alone.

Readers who want the full numbers can consult the complete sigma report on Swiss Re's website. The percentages and totals above are Swiss Re's estimates, and different organisations use different methods and definitions, so figures from other catastrophe analysts may differ.