Skip to content

Markets · Issue 01

A quieter catastrophe year is not a safer planet

Insured natural-catastrophe losses fell to $107 billion in 2025. Swiss Re still pencils $148 billion if 2026 merely returns to trend. The gap is luck, not adaptation.

MarketsUpdated 2026-08-317 min read
48.976.6104132160proj.201720192021202220242026Insured losses
Insured nat-cat losses, 2025
$107bn
Economic losses, 2025
$220bn
Los Angeles wildfires, insured
$40bn
2026 if the long-run trend holds
$148bn

2025 was below trend. That is the trap

Swiss Re’s sigma 1/2026 is the print: 190 natural-catastrophe events, $107 billion insured, $220 billion in economic losses. A record 49% of the economic damage was insured — the protection gap narrowed as a share because California wildfires ($40 billion insured) sit inside a deep insurance market. The dollar gap is still vast. First-half 2026, published 11 August, came in at about $42 billion insured, the quietest first half since 2020, against a $66 billion trend. Swiss Re’s own warning is the chart: if 2026 is merely an average year, the full-year bill is $148 billion.

Secondary perils are the product

The old catastrophe was a named hurricane. The new book is wildfire, severe convective storm, flood — high-frequency, high-severity, poorly modelled, and sitting on more valuable houses. That is why a ‘quiet’ year still clears $100 billion. It is also why listed reinsurers can look cheap on a lucky year and still be correctly priced for the decade. Pair this with 1.5 °C, already: three years of a warmer baseline is not a weather story. It is a balance-sheet story for anyone who writes property in California, Florida, Australia or the European flood plain.

The long view is who can still offer a policy

Insurance is how a market says a place is still liveable. Retreat is how it says it is not. The investing lens is not ‘buy climate’. It is the unglamorous stack: reinsurers with real models and real capital, brokers who can still place a layer, adaptation (building codes, defensible space, flood defences) that actually changes the loss curve, and a scepticism toward coastal real-estate models that assume a 2015 cost of capital and a 2015 climate. A below-trend year is a gift to next year’s earnings. It is not a regime change.

Investing lens

Horizon 5–15 years · Educational, not advice

Property catastrophe is a rising-nominal, rising-frequency book. Own balance sheets that can reprice annually and infrastructure that lowers the loss, not a slogan about ‘climate alpha’.

Where the map points

  • Reinsurers and ILS with a demonstrated ability to reprice
  • Insurance brokers in hard markets
  • Adaptation industrials: grid hardening, water, building materials with a code tailwind
  • Be short the assumption that coastal residential always clears

What can break it

  • A lucky multi-year cluster that looks like a new mean
  • State-backed residual markets socialising the tail
  • Model risk on secondary perils

CHART does not recommend securities, funds or trades. Figures can be revised by their publishers. Do your own research and consider regulated advice before allocating capital.

Sources

Every headline number traces to a named publisher. Contextual sources inform the essay, not the key stat.

  1. 01PrimarySwiss Re Institute2026
    sigma 1/2026: Natural catastrophes in 2025 — wildfire and storm risk

    2025: $107 billion insured, $220 billion economic, 190 events. Record 49% insured share. $40 billion from Los Angeles wildfires. 2026 trend $148 billion if the long-run path holds.

  2. 02PrimarySwiss Re Institute2025-04-29
    sigma 1/2025: Natural catastrophes — insured losses on trend to USD 145 billion

    2024 insured natural-catastrophe losses $137 billion.

  3. 03CorroboratedSwiss Re Institute2024-03-26
    sigma 1/2024: Natural catastrophes in 2023

    2023: $280 billion economic, $108 billion insured.

  4. 04CorroboratedSwiss Re Institute2026-08-11
    First-half 2026 insured natural catastrophe losses

    H1 2026 insured losses ~$42 billion, below a $66 billion trend. Lowest first half since 2020.

  5. 05PrimaryCopernicus Climate Change Service / ECMWF2026-01-14
    Copernicus: 2025 was the third hottest year on record

    ERA5: 2025 1.47°C above 1850–1900; 2024 1.60°C (hottest); 2023–2025 average exceeded 1.5°C for the first time.

101 graphics in this issue