Prolepsia forecastYes / no
Will Swiss Re estimate global insured catastrophe losses above $150 billion for 2027?
As of 15 November 2026, Prolepsia puts the chance of yes at 50%.
50%chance of yes
How Prolepsia reasons
The report Prolepsia wrote with this forecast.
How forecasts are made · Track record
TL;DR
The forecast is 49.9% Yes and 50.1% No, conditional on a qualifying estimate being published. Treat this as an even-odds outcome: rising insured exposure puts USD 150 billion within reach, but the loss trend alone does not establish that actual 2027 losses will exceed it.
Key evidence
- Historical experience is balanced after adjustment. A comparison of 30 historical years, adjusted for the changing loss trend and price level, finds 15 threshold exceedances. The underlying annual observations are approximate chart readings from Swiss Re’s March 2026 report, not an exact published numerical dataset (sigma 1/2026).
- The underlying trend is rising. Swiss Re places its loss trend at USD 148 billion in 2026 and USD 186 billion in 2030, both in 2025 prices, with approximately 5–7% annual real growth. These are risk benchmarks, not actual-loss estimates for 2027 (Swiss Re).
- Quiet losses do not establish quiet hazards. First-half 2026 insured losses were USD 42 billion, including USD 28 billion from severe convective storms. US severe-weather reports were 20% above the corresponding historical average, but damaging outbreaks missed highly insured concentrations (first-half assessment).
- Large disasters create a substantial upper tail. Swiss Re’s modeled USD 320 billion peak-loss scenario for 2026 carries a 10% annual probability. Such extremes mean that an elevated average does not translate directly into frequent threshold exceedances (sigma 1/2026).
Uncertainties
The largest unknowns are the occurrence and location of major catastrophes, the distribution of losses around the trend, and changes in insured exposure and reconstruction costs. Current climate guidance does not establish conditions throughout the 2027 hurricane peak.
The reporting rules also matter near the boundary. Exactly USD 150 billion resolves No; the latest qualifying Swiss Re publication through June 30, 2028 governs. The metric is actual worldwide insured natural-catastrophe losses from 2027 events, not economic losses, man-made losses, or a modeled future-loss projection.
Read the full report
TL;DR
The forecast is 49.9% Yes and 50.1% No, conditional on non-annulment. Treat the outcome as an even-odds risk, not as a consequence of Swiss Re’s rising loss trend. The decisive distinction is between a long-term risk benchmark and the actual losses caused by catastrophes during 2027.
Context
The question concerns Swiss Re Institute’s estimate of worldwide insured natural-catastrophe losses from events occurring during January–December 2027. The threshold is strictly greater than USD 150 billion in nominal 2027 dollars. Exactly USD 150 billion resolves No. Uninsured economic damage and losses attributed exclusively to man-made catastrophes do not count.
As of November 15, 2026, the latest comprehensive annual assessment reports USD 107 billion of insured natural-catastrophe losses for 2025, compared with USD 141 billion for 2024, both in 2025 prices (Swiss Re’s annual release). The resolution deadline is June 30, 2028. The latest qualifying publication by that date governs, with the annual sigma report preferred when qualifying estimates share a publication date.
Evidence
The historical backbone is a long record of large departures from the underlying loss trend. Comparing old dollar totals directly with the threshold would understate today’s risk because exposure and prices have risen. A trend- and price-adjusted comparison of the historical record yields 15 exceedances among 30 years. That supports a balanced assessment rather than a strong directional call.
The full chart-derived annual loss history used for that comparison appears below. All values are USD billions in constant 2025 prices. These are approximate readings from the historical chart, not official tabulated values; their decimal places do not establish equivalent accuracy in the underlying loss estimates (sigma 1/2026, Figure 3).
| Year | Insured losses |
|---|---|
| 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 both exceptional loss years and stretches of lower losses. Swiss Re identifies 1999, 2004, 2005, 2011 and 2017 as peak-loss years in its earlier historical assessment. The relatively quiet 2013–2016 interval provides the counterweight: high underlying risk does not prevent several lower-loss years (sigma 1/2025). The inference is that single-year outcomes remain dominated by event variability even as the long-term baseline rises.
The strongest upward pressure comes from Swiss Re’s trend benchmark. It places the trend at USD 148 billion in 2026 and USD 186 billion in 2030, expressed in 2025 prices, and describes annual real growth of approximately 5–7%. Those figures put the question’s nominal threshold close to the underlying scale of risk. They do not constitute an estimate of actual 2027 losses (sigma 1/2026).
The distribution around that trend matters as much as its level. Swiss Re assigns a 10% annual probability to a modeled USD 320 billion peak-loss scenario for 2026. This is a separate tail-risk benchmark, not the ordinary trend and not a realized loss estimate. It shows why a few very costly years can lift average losses without making above-threshold losses the usual annual result (sigma 1/2026). The precise interpretation of the trend—as distinct from an arithmetic mean or a median—is therefore a central uncertainty.
Recent experience supplies a useful warning against extrapolation. First-half 2026 insured natural-catastrophe losses were USD 42 billion. The corresponding ten-year first-half average was USD 50 billion, while the trend benchmark was USD 66 billion; those are different comparisons, not conflicting estimates. Severe convective storms accounted for USD 28 billion. US severe-weather reports through June were 20% above the 2016–2025 average, yet damaging outbreaks missed highly insured concentrations (Swiss Re’s first-half assessment). Low insured losses therefore did not demonstrate a comparable decline in hazard.
An independent exposure-based signal points in the same upward direction. Verisk raised its modeled global insured average annual catastrophe-loss estimate to USD 171 billion, an increase of USD 19 billion. Its assessment attributes 40% of modeled losses to severe thunderstorms and 27% to tropical cyclones, and reports roughly 7% annual property-exposure growth since 2021 (Verisk’s assessment). This supports a high loss baseline without requiring an exceptional hurricane season. It cannot be substituted for Swiss Re’s eventual reported total because the models, coverage and definitions differ.
Prices add another upward influence on a nominal-dollar threshold, but their exact path remains unsettled. Swiss Re’s public inflation projections give 3.3% for 2026 and 2.4% for 2027, while the IMF projections give 3.6% and 2.8%, respectively (Swiss Re; IMF). These are projections rather than realized inflation. Their disagreement matters less than catastrophe variability, and neither permits an independent price conversion at resolution if Swiss Re does not provide enough information.
Climate guidance does not supply a strong directional signal for the full loss year. NOAA’s November outlook gives June–August 2027 probabilities of 33.0% La Niña, 61.0% neutral conditions and 6% El Niño. That window does not cover the entire peak Atlantic hurricane season, and a climate phase is not a forecast of insured landfall losses (NOAA outlook). Current conditions therefore do not justify treating 2027 as a quiet hurricane year.
What's non-obvious
A loss trend above the threshold is not the same as a better-than-even threshold probability. Catastrophe losses are right-skewed. A small number of enormous outcomes can pull the average far above the middle of the distribution. At the same time, Swiss Re’s plotted trend should not automatically be treated as a precisely defined arithmetic mean. Both shortcuts create false confidence. The historical balance and the uncertainty about the distribution explain the near-even forecast.
The other overlooked distinction is between hazard activity and insured damage. A busy storm period can produce modest insured losses when events miss expensive, well-insured assets. That does not mean the hazard disappeared, nor does it mean missed losses are owed in the following year. The first-half evidence supports maintaining a high baseline while leaving the actual geography of future events unresolved (Swiss Re’s assessment).
Uncertainties
The largest uncertainty is where and when major catastrophes occur. Hurricane landfalls, earthquake occurrence, wildfire footprints and severe-storm tracks can move annual insured losses sharply. A growing baseline does not remove that randomness.
The historical evidence also has limits. The numerical series is chart-derived, and the history reflects March 2026 restatements rather than matched first-publication estimates. An exact official annual series, together with a history of revisions, would better connect past experience to the estimate that will actually determine resolution. More detail on the statistical meaning of Swiss Re’s trend would also narrow the uncertainty about ordinary versus extreme years.
Finally, reporting can decide a close result. A published rounded value of USD 150 billion resolves No unless the same publication supplies a qualifying, more precise value above it. Revisions published by the deadline can change the applicable estimate; later revisions cannot. If no qualifying full-year estimate exists, or the nominal 2027 total cannot be identified under the stated price rules, the question is annulled. The forecast remains 49.9% Yes and 50.1% No, conditional on non-annulment.
Forecast history
| Date | Forecast |
|---|---|
| 50% chance of yes |
Sources Prolepsia read
- Wildfires, storms, floods contribute to record 92% of global insured losses in 2025, says Swiss Re Institute | Swiss Rewww.swissre.com
- Global Insured Catastrophe Losses Now Expected to Average $171 Billion Annually, Verisk Finds | Veriskwww.verisk.com
- sigma explorer | sigma research | Homesigma.swissre.com
- World Economic Outlook (October 2026) - Inflation rate, average consumer priceswww.imf.org
- Climate Prediction Center - Official NOAA CPC ENSO Probabilitieswww.cpc.ncep.noaa.gov
- IRI – International Research Institute for Climate and Society | October 2026 Quick Lookiri.columbia.edu
- Excessive Momentum and False Alarms in Late‐Spring ENSO Forecastsrepository.library.noaa.gov
- Swiss Re Institute — sigma No 1/2026legismap.com.br
- Swiss Re TEST |www.finanzwire.com
- Steigende Hitze, mehr Exponierung, sich verändernde Naturgefahren: Ruhigere erste Jahreshälfte 2026 verdeckt steigendes Naturkatastrophenrisiko, zeigt Swiss Re Institute | Corporate - EQS Newswww.eqs-news.com
The question
This question asks whether Swiss Re Institute will estimate that worldwide insured losses attributable to natural catastrophes occurring during the 2027 calendar year exceeded USD 150 billion. The forecast concerns Swiss Re's estimate of actual insured losses for 2027, rather than its projections of expected losses or modeled catastrophe scenarios. The relevant period is January 1 through December 31, 2027, inclusive, and the outcome will be determined using Swiss Re Institute's published estimates, with a resolution deadline of June 30, 2028. As of November 15, 2026, Swiss Re Institute's latest comprehensive annual sigma report, published in March 2026, estimated that global insured natural catastrophe losses totaled USD 107 billion in 2025, compared with USD 141 billion in 2024, expressed in 2025 prices. Swiss Re also estimated that the long-term trend in annual insured natural catastrophe losses would reach approximately USD 148 billion in 2026 and USD 186 billion by 2030, reflecting an underlying annual growth trend of approximately 5–7% in real terms. Its preliminary estimate for the first half of 2026 was USD 42 billion. These figures illustrate both the increasing long-term financial exposure to natural catastrophes and the substantial year-to-year variability in insured losses. Sources: Swiss Re Institute, sigma 1/2026 (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 (https://www.swissre.com/institute/research/topics-and-risk-dialogues/climate-and-natural-catastrophe-risk/first-half-2026-insured-catastrophe-losses.html). The forecast will resolve based on the worldwide annual insured natural catastrophe loss estimate reported by Swiss Re Institute for 2027, using the nominal USD value expressed in 2027 prices where available. The USD 150 billion threshold is strict: an estimate exactly equal to USD 150 billion does not qualify.
How it resolves
The question resolves YES if Swiss Re Institute publishes an estimate of global insured natural catastrophe losses for the full 2027 calendar year strictly greater than USD 150 billion by June 30, 2028. The question resolves NO if the qualifying estimate is USD 150 billion or less. The primary resolution source will be Swiss Re Institute's sigma natural catastrophe report covering the full 2027 calendar year, expected to be published in 2028. If that report is unavailable by June 30, 2028, the most recent publicly available Swiss Re Institute estimate explicitly covering the full 2027 calendar year will be used, including an annual catastrophe-loss press release or sigma research update. If multiple estimates are available, the latest qualifying estimate published on or before June 30, 2028, takes precedence, with the annual sigma report preferred if estimates have the same publication date. The qualifying metric is the worldwide aggregate of insured losses caused by natural catastrophes, as defined and reported by Swiss Re Institute. This includes insured property losses from events such as hurricanes, earthquakes, floods, wildfires, and severe storms. Losses attributed exclusively to man-made catastrophes are excluded. Total economic losses, uninsured losses, and modeled or projected future losses do not qualify. The estimate must represent losses attributable to catastrophes occurring during the 2027 calendar year, rather than insurance claims paid during 2027 for catastrophes in other years. Swiss Re's own event attribution and reporting conventions will determine which losses are included. The reported value will be assessed in nominal US dollars at 2027 price levels. If Swiss Re publishes its 2027 estimate in constant prices from another year, the corresponding 2027-price figure will be used if Swiss Re explicitly provides it. If Swiss Re provides only inflation-adjusted figures in another year's prices, without sufficient information to identify the nominal 2027 total, the question will be annulled rather than relying on an independently constructed conversion. If Swiss Re Institute has not published any qualifying estimate for full-year 2027 losses by June 30, 2028, the question will be annulled. Estimates published after the resolution deadline and subsequent revisions will not alter the outcome.
Fine print
The USD 150 billion threshold is exclusive: USD 150.0 billion resolves NO, while any explicitly reported value greater than USD 150 billion resolves YES. If Swiss Re reports a rounded figure, the published rounded figure will determine resolution unless the same publication provides a more precise underlying figure. Preliminary estimates may be used if no subsequent qualifying estimate is published before the resolution deadline. The forecast does not require Swiss Re's estimates to match those of other organizations, such as Munich Re, Aon, or Gallagher Re. The question measures Swiss Re's reported estimate of actual 2027 insured natural catastrophe losses, not whether its long-term modeled loss trend exceeds USD 150 billion. A projection made before or during 2027 does not qualify as a resolution estimate, even if it exceeds the threshold. The assessment is global and is not restricted to any particular country or geographic region. For resolution purposes, the publication date of the relevant Swiss Re estimate governs, rather than the date on which the underlying calculations were completed.
More public forecasts
Will global insured natural catastrophe losses in 2027 exceed $120 billion, as reported by Swiss Re?
73%chance of yes
Prolepsia forecast
Will US severe convective storm insured losses in 2027 exceed $50 billion, as estimated by Gallagher Re?
57%chance of yes
Prolepsia forecast
Will at least one US Atlantic hurricane cause more than 20 billion dollars of insured losses in 2027?
27%chance of yes
Prolepsia forecast
Will a single cyber event cause insured losses above USD 1 billion in 2027?
19%chance of yes
Prolepsia forecast
Will global cyber insurance premiums exceed $20 billion in 2027, as estimated by Munich Re?
10%chance of yes
Prolepsia forecast
Will Eurostat's first estimate show euro area GDP growth above 1.2% for 2027?
47%chance of yes
Prolepsia forecast