Prolepsia forecastYes / no

Will global insured natural catastrophe losses in 2027 exceed $120 billion, as reported by Swiss Re?

As of 15 November 2026, Prolepsia puts the chance of yes at 73%.

73%chance of yes

Yes 73%No 27%
Prolepsia forecast made

How Prolepsia reasons

The report Prolepsia wrote with this forecast.

How forecasts are made · Track record

TL;DR

The forecast is 73% Yes and 27% No that global insured natural-catastrophe losses in 2027 exceed USD 120 billion in Swiss Re’s controlling annual assessment. Structural growth favors Yes, but large annual swings leave a substantial chance of losses at or below the threshold.

Key evidence

  • The underlying loss level is above the threshold. Swiss Re places the 2026 trend at USD 148 billion in 2025 prices and describes long-term growth of 5–7% annually in real terms. Growing asset concentrations, rising values and reconstruction costs support higher losses, although a trend is not an annual floor (sigma 1/2026).
  • Recent annual results show the variability. Swiss Re reports USD 107 billion of natural-catastrophe insured losses for 2025, compared with USD 141 billion for 2024 when restated in 2025 prices. The USD 120 billion figure associated with 2025 includes man-made disasters and is not the question’s metric (annual assessment).
  • A quiet period does not establish lower underlying risk. First-half 2026 insured losses were USD 42 billion, versus a USD 50 billion corresponding historical average. Severe-weather activity was above average, but major outbreaks largely missed the most exposed locations (first-half assessment).
  • Losses are not just a hurricane story. Secondary perils accounted for 92% of 2025 insured natural-catastrophe losses. Storms, floods and wildfires provide several routes to a costly year (Swiss Re release).

Uncertainties

The largest uncertainty is where disasters strike relative to insured assets. Continued structural growth does not prevent a quiet hurricane year, limited damage from earthquakes, or storms that miss major population centers. Early climate outlooks do not settle those questions.

Historical comparisons also contain price-basis changes and later loss revisions. The question instead uses the first comprehensive annual assessment for 2027, expected in 2028, without additional price adjustments. Exactly USD 120 billion resolves No unless the same report supplies a more precise figure above it. These reporting details matter near the boundary, but event severity and geography remain the larger risks to the forecast.

Read the full report

TL;DR

The forecast is 73% Yes and 27% No that Swiss Re’s controlling annual assessment reports global insured natural-catastrophe losses strictly above USD 120 billion for 2027. Treat exceedance as the planning case, not as a certainty. The decisive distinction is between a rising underlying loss level and a realized annual total that still depends heavily on where disasters strike.

Context

This question concerns worldwide insurance-industry losses from natural catastrophes occurring during calendar 2027. It excludes uninsured economic damage and man-made disasters; it does not ask about Swiss Re’s own claims. The first comprehensive annual sigma report for 2027, expected in 2028, controls the result. Exactly USD 120 billion is No unless that report provides a more precise figure above the threshold.

As of November 15, 2026, the latest comprehensive annual assessment reports USD 107 billion for 2025. The available first-half 2026 assessment reports USD 42 billion. Both observations sit below their respective trend benchmarks, but Swiss Re continues to identify structural growth in insured catastrophe losses (annual report; first-half assessment).

Evidence

The historical backbone is a rising loss level with large deviations around it. A fixed-dollar threshold becomes easier to cross as insured exposure grows. Yet the historical record also contains quiet years between costly ones. Neither the latest low observation nor a single extreme year is a sufficient guide.

The table gives the full 1996–2025 historical series used in this assessment. These are approximate readings from Swiss Re’s published chart, in USD billions at 2025 prices—not official unrounded annual totals. Small differences between the chart readings and rounded headline figures should not be treated as meaningful precision (historical chart, sigma 1/2026).

Year Insured natural-catastrophe losses, USD billions at 2025 prices
1996 20.3
1997 10.7
1998 32.2
1999 57.7
2000 16.7
2001 20.9
2002 26.3
2003 32.1
2004 80.7
2005 170.0
2006 21.3
2007 37.5
2008 67.1
2009 34.6
2010 68.3
2011 174.4
2012 92.7
2013 52.0
2014 40.2
2015 38.5
2016 63.7
2017 190.2
2018 106.3
2019 67.9
2020 106.0
2021 133.5
2022 142.0
2023 118.4
2024 141.4
2025 107.4

The record shows why simple extrapolation fails. Very costly years are followed by much lower totals, while the general level rises over time. Comparing all these years directly with the question’s threshold would also be misleading: the table uses a common historical price basis, and older years had less insured exposure. The relevant reference class is annual variability around the growing underlying risk level, not an unadjusted count of past threshold crossings.

Swiss Re’s current anchor favors exceedance. Its annual report places the 2026 trend at USD 148 billion in 2025 prices and describes annual real growth of 5–7%. The drivers include concentrations of insured assets in exposed locations, higher asset values, reconstruction costs and changing hazards. I read this as the strongest evidence for Yes: the threshold is below the published underlying trend even before moving into the target year (sigma 1/2026).

That does not turn the trend into a floor. Swiss Re reports actual 2025 natural-catastrophe insured losses of USD 107 billion, below its trend. It also estimates an approximately one-in-ten-year 2026 loss scenario of USD 320 billion in 2025 prices. Together, those figures describe a broad, right-skewed loss environment: many years fall short of the underlying trend, while a few years produce exceptionally large losses. The extreme scenario establishes tail risk; it does not by itself establish the chance of crossing the much lower threshold (annual report).

The latest annual comparison also requires care. Swiss Re’s current report gives USD 141 billion for 2024 in 2025 prices, while the preceding annual report gave USD 137 billion in 2024 prices. Those figures use different price bases. They should not be presented as contradictory observations or as evidence of a sudden change in physical risk (current assessment). The question uses the reporting-year price basis Swiss Re adopts for 2027; it does not call for a separate inflation adjustment to the published outcome.

The quiet first half of 2026 is genuine counterevidence to an expectation of uniformly high annual losses, but weak evidence of a lasting reduction in risk. Swiss Re estimated USD 42 billion for January–June, compared with a USD 50 billion first-half average over 2016–2025 and a USD 66 billion first-half trend. Severe-weather activity was approximately 20% above the corresponding historical average, while insured severe-convective-storm losses were USD 28 billion. Swiss Re attributed the favorable loss result partly to major outbreaks missing the most exposed places (first-half assessment).

This distinction between activity and damage matters. A quiet insured-loss period can result from favorable geography rather than weaker hazards or reduced exposure. It therefore supports keeping substantial downside uncertainty, not assuming that the same luck carries into 2027.

The peril mix broadens the routes to exceedance. Secondary perils accounted for 92% of insured natural-catastrophe losses in 2025. A forecast centered only on Atlantic hurricanes would miss the contribution from storms, floods and wildfires (Swiss Re’s annual release). A major hurricane landfall can dominate a year, but several other perils can also build a large global total.

Climate information adds context rather than a decisive annual-loss signal. NOAA’s November outlook describes strong El Niño conditions through the early part of 2027 followed by weakening. It does not establish conditions throughout the target year’s hurricane season, much less insured landfall losses (NOAA discussion). Carrying an early-year climate label through the full calendar year would give that signal more weight than the evidence supports.

What's non-obvious

Low insured losses do not necessarily mean low hazard activity. The first-half 2026 evidence makes that concrete: above-average severe-weather activity coexisted with below-average insured losses because location mattered. This cuts both ways. It explains why recent quiet losses do not erase structural growth, and why structural growth does not guarantee a costly target year.

There is also a measurement trap. Swiss Re’s USD 120 billion total associated with 2025 includes man-made disasters; the natural-catastrophe figure was USD 107 billion. Using the broader total would falsely suggest that the latest year had already reached the question’s benchmark (sigma 1/2026). The case for Yes rests on the underlying loss environment, not on redefining the metric or treating a threshold-level rounded figure as sufficient.

Uncertainties

The largest uncertainty is event geography. Hurricane landfalls, earthquake locations, wildfire spread and storm tracks determine how much insured property is hit. A year with substantial catastrophe activity can still resolve No if the damage avoids heavily insured assets. Conversely, clustering across several perils can produce a costly year without one dominant event.

The next uncertainty is whether the historical relationship between growth and variability remains representative. Exposure, coverage and peril composition change. Recent quieter volatility does not establish a permanent new regime, and the long-term growth estimate is not a promise about any particular year. More complete exposure and coverage data would help distinguish durable risk growth from changes in what is insured.

The historical record also differs from the resolution target. Swiss Re updates its loss database, while this question is governed by the first comprehensive annual assessment. A complete history of first-report totals and later revisions would clarify that gap; the available evidence does not quantify it (methodology in the annual report).

Finally, reporting precision matters near USD 120 billion. A rounded total of exactly that amount resolves No unless a qualifying precise figure in the same report exceeds it. Later revisions do not change the outcome. These details can decide a boundary case, but the main uncertainty remains the physical and insured severity of the catastrophe year.

Forecast history

How Prolepsia’s forecast has movedOne forecast so far: 73% on 15 Nov 2026.
DateForecast
73% chance of yes

Sources Prolepsia read

  1. annual reportwww.swissre.com
  2. Climate Prediction Center: ENSO Diagnostic Discussioncpc.ncep.noaa.gov
  3. Consumer Price Index News Release - 2026 M10 Resultswww.bls.gov
  4. Consumer Price Index for All Urban Consumers: All Items in U.S. City Average (CPIAUCNS) | FRED | St. Louis Fedfred.stlouisfed.org

The question

This question asks whether total worldwide insured losses resulting from natural catastrophes occurring during the calendar year 2027 (January 1 through December 31, 2027) will exceed USD 120 billion, according to Swiss Re Institute's annual natural catastrophe loss estimates. The forecast covers insured losses globally, including those arising from hurricanes, tropical cyclones, severe convective storms, floods, wildfires, earthquakes, winter storms, and other natural catastrophes included in Swiss Re's reporting methodology. It excludes uninsured economic losses and losses attributable to man-made disasters. As of November 15, 2026, Swiss Re Institute's sigma 1/2026 report estimates that global insured natural catastrophe losses totaled USD 107 billion in 2025, compared with USD 141 billion in 2024. Swiss Re's August 2026 assessment estimates that insured natural catastrophe losses reached USD 42 billion during the first half of 2026, approximately 16% below the corresponding 10-year average. Despite this relatively quiet period, Swiss Re identifies a continuing long-term upward trend in insured catastrophe losses, driven by increasing concentrations of insured assets in hazard-prone areas, rising asset values, reconstruction costs, and changing hazard patterns. Swiss Re estimates that structural factors could contribute to long-term insured loss growth of approximately 5–7% annually, although realized losses fluctuate substantially between years. The USD 120 billion threshold therefore represents a meaningful benchmark for assessing whether 2027 experiences elevated insured catastrophe losses. The question will be evaluated using Swiss Re Institute's annual sigma natural catastrophe statistics for 2027, expected to be published in 2028. Background sources: Swiss Re Institute, sigma 1/2026, Natural catastrophes in 2025: the persistent rise of wildfire and storm risk (https://www.swissre.com/institute/research/sigma-research/sigma-2026-01-natcat-2025-wildfire-storm-risk.html); Swiss Re Institute, First-half 2026 insured catastrophe losses: below trend, rising risks (https://www.swissre.com/institute/research/topics-and-risk-dialogues/climate-and-natural-catastrophe-risk/first-half-2026-insured-catastrophe-losses.html).

How it resolves

The question resolves YES if Swiss Re Institute's first comprehensive annual sigma report covering natural catastrophes in calendar year 2027 reports total global insured natural catastrophe losses strictly greater than USD 120 billion. It resolves NO if that report places the corresponding total at or below USD 120 billion. The primary resolution source will be Swiss Re Institute's sigma natural catastrophe report for 2027, expected to be published during 2028. The relevant metric is the worldwide annual total for insured losses attributable to natural catastrophes, expressed in billions of US dollars, rather than total economic losses or combined natural and man-made catastrophe losses. The reported figure for 2027 will be used in the price basis adopted by Swiss Re for that reporting year, without additional inflation adjustments or currency conversions. The first comprehensive annual sigma report for 2027 will take precedence over Swiss Re's preliminary year-end 2027 estimates and any subsequent historical revisions. If the report provides an explicit global insured natural catastrophe loss total, that figure will determine resolution. If it reports a rounded total of exactly USD 120 billion, the question resolves NO unless the same report explicitly provides a more precise figure exceeding USD 120 billion. If Swiss Re changes its reporting format but publishes a clearly equivalent global annual insured natural catastrophe loss figure, that figure may be used. If Swiss Re does not publish its comprehensive annual sigma report for 2027 by December 31, 2028, an equivalent official annual estimate published by Swiss Re Institute by that date may be used, with preference given to the most comprehensive available assessment. If no qualifying Swiss Re estimate is available by December 31, 2028, the question will be annulled rather than resolved using estimates from another organization. Swiss Re's classification of natural catastrophes and methodology for insured-loss aggregation will govern the outcome.

Fine print

The threshold is strictly greater than USD 120 billion: a reported total of exactly USD 120 billion does not qualify. All figures are nominal US-dollar amounts on the price basis used in Swiss Re's report for the relevant year, not inflation-adjusted to a common historical base. Losses are attributed to 2027 according to Swiss Re Institute's reporting conventions, even if related insurance claims are settled in subsequent years. The metric includes insured losses borne by the insurance and reinsurance industry worldwide, as defined by Swiss Re, rather than losses incurred specifically by Swiss Re itself. Swiss Re's published global total will be authoritative even if it differs from figures reported by Munich Re, Aon, Gallagher Re, or other catastrophe-loss tracking organizations. Preliminary estimates may differ from the eventual annual sigma assessment, but the designated annual report will be used for resolution. Subsequent revisions published after the question has resolved will not change the outcome. If Swiss Re publishes an unrounded figure and a rounded summary figure, the unrounded figure takes precedence. If the selected report explicitly presents conflicting figures for the same metric that cannot be reconciled, the question will be annulled unless Swiss Re issues a clarification by December 31, 2028.