Systems in depth
One section per system: its state at evidence close, its exposures and what the operational series will watch.
Contents
- ·Edition home
- ·Executive summary
- 1Protection gaps are shared risk
- 2Key findings
- 3Systems and sectors at baseline
- 4Risks and capability
- 5Large insured loss that meets soft pricing and thin backstops
- 6Readings from the other desks
- 7The 2027 agenda
- 8Systems in depth
- AMethodology
- BData appendix
- CReferences and record
- ·Complete edition, for print
8.1Common causes
One driver can raise loss in many markets at once: rising exposure, secondary perils and low take-up of cover. The shared cause matters more than the line of business.
| Instance | Mechanism in action | Evidence | Status |
|---|---|---|---|
| Secondary perils | wildfire, severe storm and flood made up most insured loss | 1 | Reported |
| Exposure growth | inflation-adjusted exposure grows faster than premiums | 7 | Reported |
| Low take-up | USD 9 billion insured of about USD 73 billion in Asia-Pacific | 2 | Reported |
| Disaster risk trends | global hazard and adaptation context | 2627 | Inferred |
Pathways in this mechanism
- CLI → FIN secondary perils
- FIN → DEV low take-up
- WAT → FIN flood loss
Evidence to bind in 2027
- common cause tagging of insured loss
- take-up data by income
Related editions
- GLB Global Systems
- SR-01 Global Polycrisis and Cascading Systems Risk
- SR-02 Critical Infrastructure and Lifeline Systems
- SR-05 Sovereign, Fiscal and Financial Resilience
- GS-05 Finance and Economy
- GS-08 Disaster and Resilience
8.2Domino effects
A loss in one place removes cover in the next. Nonrenewal sends policies to residual markets, and uncovered flood loss lands on a public scheme or on the household.
| Instance | Mechanism in action | Evidence | Status |
|---|---|---|---|
| Nonrenewal to residual market | higher nonrenewal in high-risk areas; scheme exposure up 252% since 2022 | 128 | Reported |
| Residual market to private market | Citizens fell from about 1.4 million to 439,079 policies through transfers | 9 | Reported |
| Flood to public scheme | most homeowners policies exclude flood | 1110 | Reported |
| Cover to credit | most mortgage lenders require homeowners cover | 23 | Reported |
Pathways in this mechanism
- FIN → FIN nonrenewal
- WAT → INF → FIN flood
- FIN → HLT, DEV uninsured loss
Evidence to bind in 2027
- timing of cover loss after an event
- mortgage default linkage
Related editions
- GLB Global Systems
- SR-01 Global Polycrisis and Cascading Systems Risk
- SR-02 Critical Infrastructure and Lifeline Systems
- SR-05 Sovereign, Fiscal and Financial Resilience
- GS-05 Finance and Economy
- GS-08 Disaster and Resilience
8.3Inter-systemic feedbacks
Cover also shapes the risk it covers. Price signals and scheme terms can reward mitigation, and capital cycles feed back on prices.
year the Principles for Sustainable Insurance were launched20
| Instance | Mechanism in action | Evidence | Status |
|---|---|---|---|
| Pricing and prevention | mitigation discounts in the Australian cyclone pool | 18 | Reported |
| Adaptation | homeowner-level adaptation can reduce insured loss | 1 | Reported |
| Capital cycle | record capital and lower ceded loss pushed rates down | 2 | Reported |
| Disclosure | insurers and others disclose climate risk under IFRS S2 where adopted | 21 | Reported |
Pathways in this mechanism
- FIN → CLI prevention
- FIN ↔ TRD capital cycle
- DAT → FIN cyber growth
Evidence to bind in 2027
- evidence that price signals change building
- effect of disclosure on cover
Related editions
- GLB Global Systems
- SR-01 Global Polycrisis and Cascading Systems Risk
- SR-02 Critical Infrastructure and Lifeline Systems
- SR-05 Sovereign, Fiscal and Financial Resilience
- GS-05 Finance and Economy
- GS-08 Disaster and Resilience
8.4Shared vulnerabilities
Different lines share the same weak points: concentration in a few perils and places, short-dated public schemes and a thin evidence base.
short-term NFIP reauthorisations since FY201710
| Instance | Mechanism in action | Evidence | Status |
|---|---|---|---|
| Concentration in one market | the United States held USD 100 billion of USD 129 billion insured loss | 4 | Reported |
| Concentration in one scheme | one residual market holds USD 773 billion of exposure | 8 | Reported |
| Short-dated public cover | repeated short extensions and two recent lapses | 10 | Reported |
| Cyber accumulation | young market; thin loss data | 19 | Expert input |
Pathways in this mechanism
- CLI → FIN concentration
- INF → FIN scheme deadlines
- DAT → FIN cyber
Evidence to bind in 2027
- accumulation data across lines
- scheme capital after a large loss
Related editions
- GLB Global Systems
- SR-01 Global Polycrisis and Cascading Systems Risk
- SR-02 Critical Infrastructure and Lifeline Systems
- SR-05 Sovereign, Fiscal and Financial Resilience
- GS-05 Finance and Economy
- GS-08 Disaster and Resilience
8.5Thresholds and tipping
A system can absorb loss until a threshold is crossed. Scheme deadlines, borrowing limits and the size of a peak loss are such thresholds.
| Instance | Mechanism in action | Evidence | Status |
|---|---|---|---|
| Peak loss | modelled peak scenario is about twice the normal level | 1 | Estimated |
| Scheme deadline | authorisation to 11 December 2026; limit falls from USD 30.425 billion to USD 1 billion | 10 | Reported |
| Rule change | amended Solvency II rules apply from 30 January 2027 | 22 | Reported |
| Tipping in markets | adverse selection can unravel pooled cover | 3940 | Inferred |
Pathways in this mechanism
- FIN → FIN cover unravelling
- INF → FIN deadline
- CLI → FIN peak loss
Evidence to bind in 2027
- capital thresholds of residual markets
- affordability thresholds for households
Related editions
- GLB Global Systems
- SR-01 Global Polycrisis and Cascading Systems Risk
- SR-02 Critical Infrastructure and Lifeline Systems
- SR-05 Sovereign, Fiscal and Financial Resilience
- GS-05 Finance and Economy
- GS-08 Disaster and Resilience
Notes
Numbers are the edition’s reference numbers, so a note reads the same in every chapter. The full list, with every source and its licence, is in Annex C. Annex C, references
- 1Insurance Journal and Reinsurance News (eds.), reports of Swiss Re Institute, sigma 1/2026, Natural catastrophes in 2025: the persistent rise of wildfire and storm risk (insured losses USD 107 billion; economic losses USD 220 billion; 92% from secondary perils; Los Angeles wildfires USD 40 billion; trend 5% to 7% a year; 2026 normal level USD 148 billion and peak scenario about USD 320 billion), articles of March 2026. The sigma report itself was not opened; the Swiss Re page returned an access error. Observation year 2025. Accessed 2026-11-03.back
- 7Artemis (ed.), report of Swiss Re Institute insurance market outlook, July 2025 (inflation-adjusted natural catastrophe exposure growth of 5% to 7% a year, faster than premiums). The outlook itself was not opened. Observation year 2025. Accessed 2026-11-03.back
- 2Insurance Journal, Insurance Business and Reinsurance News (eds.), reports of Munich Re, natural disaster losses in 2025, 13 January 2026 (overall losses about USD 224 billion; insured about USD 108 billion; Asia-Pacific about USD 73 billion overall and USD 9 billion insured; Africa about USD 3 billion; January 2026 reinsurance rates down 14.7%). One article printed USD 244 billion for overall losses, a typographical conflict with USD 224 billion elsewhere. The Munich Re release was not opened. Observation year 2025. Accessed 2026-11-03.back
- 26UNVERIFIED, not opened: UNDRR (ed.), Global Assessment Report on Disaster Risk Reduction 2025: Resilience Pays, cited from memory; no figure used. United Nations Office for Disaster Risk Reduction, Geneva. Observation year 2025. Accessed 2026-11-03.back
- 27UNVERIFIED, not opened: IPCC, Climate Change 2022: Impacts, Adaptation and Vulnerability, Working Group II contribution to the Sixth Assessment Report, cited from memory; no figure used; the IPCC page returned an access error. Cambridge University Press. Observation year 2022. Accessed 2026-11-03.back
- 12US Department of the Treasury, Federal Insurance Office (ed.), Homeowners insurance report, press release of 16 January 2025 (more than 330 insurers; 246 million policies; 2018 to 2022; ZIP Code level; nonrenewal rates about 80% higher in the highest-risk ZIP Codes). The report itself was not opened. Observation year 2022. Accessed 2026-11-03.back
- 8California FAIR Plan Association (ed.), Key Statistics and Data (September 2026: 698,307 policies in force; exposure USD 773 billion; written premium USD 2.04 billion; policies up 157% and exposure up 252% since September 2022). California FAIR Plan Association, Los Angeles. Observation year 2026. Accessed 2026-11-03.back
- 11FEMA (ed.), FloodSmart (most homeowners insurance does not cover flood damage; almost one third of NFIP claims come from outside high-risk flood areas, stated as 29%). Federal Emergency Management Agency, Washington. Observation year 2026. Accessed 2026-11-03.back
- 9News Service of Florida (ed.), report on Florida Citizens Property Insurance Corporation policy counts, 25 November 2025 (439,079 policies; peak of about 1.4 million in September 2023). Read through a television station republication. Observation year 2025. Accessed 2026-11-03.back
- 10Congressional Research Service (ed.), What Happens If the National Flood Insurance Program (NFIP) Lapses?, Insight IN10835, version of 11 September 2026 (authorised to 11 December 2026; 36 short-term reauthorisations since FY2017; lapses 1 October to 12 November 2025 and 1 to 3 February 2026; borrowing limit USD 30.425 billion; nearly 4.5 million policies; almost USD 1.3 trillion coverage). Library of Congress, Washington. Observation year 2026. Accessed 2026-11-03.back
- 23POST-CLOSE, context only: National Association of Insurance Commissioners (ed.), Homeowners Insurance (most mortgage lenders require homeowners coverage; Homeowners Market Data Call meetings of 14 and 16 October 2026 and a joint meeting set for 16 November 2026). NAIC, Kansas City. Observation year 2026. Accessed 2026-11-03.back
- 19Munich Re (ed.), Global Cyber Risk and Insurance Survey 2026 (premium volume nearly USD 15 billion in 2025; about USD 28 billion forecast for 2030; over 9,500 respondents in 20 countries). Munich Re, Munich. Observation year 2025. Accessed 2026-11-03.back
- 20UNEP Finance Initiative (ed.), Principles for Sustainable Insurance (launched 2012; close to 300 members). United Nations Environment Programme, Geneva. Observation year 2012. Accessed 2026-11-03.back
- 18Australian Reinsurance Pool Corporation (ed.), Cyclone Reinsurance Pool (commenced 1 July 2022; annually reinstated government guarantee of 10 billion, stated as dollars and read here as Australian dollars; mitigation discounts; cost-neutral to government over the longer term). ARPC, Canberra. Observation year 2022. Accessed 2026-11-03.back
- 21IFRS Foundation (ed.), IFRS S2 Climate-related Disclosures (issued June 2023; effective for periods from 1 January 2024; amendments to greenhouse gas emissions disclosures issued December 2025). ISSB, London. Observation year 2023. Accessed 2026-11-03.back
- 4Reinsurance News (ed.), report of Gallagher Re, Natural Catastrophe and Climate Report, full year 2025 (insured losses USD 129 billion; economic losses USD 296 billion; protection gap 56% or USD 167 billion; United States USD 100 billion or 78%). The Gallagher Re report was not opened. Observation year 2025. Accessed 2026-11-03.back
- 22EIOPA (ed.), Solvency II (Directive (EU) 2025/2 amends the framework; new rules apply from 30 January 2027). European Insurance and Occupational Pensions Authority, Frankfurt. Observation year 2025. Accessed 2026-11-03.back
- 39UNVERIFIED, not opened: Akerlof, G.A., The market for lemons: quality uncertainty and the market mechanism, Quarterly Journal of Economics 84(3), 488 to 500, 1970, cited from memory. Observation year 1970. Accessed 2026-11-03.back
- 40UNVERIFIED, not opened: Rothschild, M. and Stiglitz, J., Equilibrium in competitive insurance markets: an essay on the economics of imperfect information, Quarterly Journal of Economics 90(4), 629 to 649, 1976, cited from memory. Observation year 1976. Accessed 2026-11-03.back
