The market

What reinsurance is

Reinsurance is insurance purchased by insurance companies. When an insurer writes a policy (say, homeowner’s insurance in Florida), it takes on the risk that a hurricane could destroy thousands of homes simultaneously. That concentrated exposure can threaten the insurer’s solvency. Reinsurance lets the insurer transfer some of that risk to another company, the reinsurer, in exchange for a share of premium.

The mechanism is straightforward: the insurer (called the cedent) pays a premium to the reinsurer, and the reinsurer agrees to cover some portion of the cedent’s losses under specified conditions. Those conditions are defined in a contract, and the precise structure of that contract is where the analytical complexity begins.

Two flows run in opposite directions: premium moves from policyholder to cedent to reinsurer, and when a catastrophe strikes, recoveries move back the other way.

Key Insight

Reinsurance is not a different business from insurance. It is the same business operating at a different level. The fundamental problem is identical: pricing risk and managing capital. The difference is scale, concentration, and the sophistication of the counterparties.

An insurer needs that transfer because some losses arrive together, and pooling does not smooth them.

Why reinsurance exists

Insurance works because of the law of large numbers: an insurer pools thousands of independent risks (individual homes, cars, businesses), and the aggregate loss becomes predictable. But some risks are not independent. A single hurricane can damage 100,000 homes on the same day. An earthquake can trigger claims across an entire region simultaneously.

These correlated losses, events that affect many policyholders at once, are called catastrophes. The law of large numbers still holds; what a catastrophe violates is the independence assumption that makes pooling effective.¹ When one event drives thousands of claims at once, aggregation stops smoothing the volatility, and that single event can generate losses that exceed the insurer’s capital.

	Independent risks (fire, theft, motor)	A catastrophe (hurricane, earthquake)
Claims from one cause	One policy	Thousands of policies at once
Dependence between policies	Negligible	Every exposed policy moves together
Aggregate loss	Predictable: the law of large numbers	Set by single events; pooling does not smooth it
What protects the insurer’s capital	Its own pool of premiums	Reinsurance

Reinsurance exists to solve this problem. By transferring catastrophe risk to reinsurers, insurers can:

1. Write more business: without reinsurance, an insurer’s capacity is limited to its own capital
2. Stabilise earnings: reinsurance smooths the volatility of catastrophe years
3. Meet regulatory requirements: regulators require insurers to demonstrate they can survive extreme events²
4. Access expertise: reinsurers specialise in understanding, pricing, and managing catastrophe risk

Where those transfers take place, and between whom, is the subject of Where reinsurance operates.

A brief history

Reinsurance is older than the insurance companies that buy it, by about three centuries.³

The milestones that shaped today's reinsurance market, from the first known contract to the events that made catastrophe modelling and alternative capital standard.

Origins

The earliest known reinsurance contract dates to July 1370, written on a marine cargo voyage from Genoa via Cadiz to Sluys, the port serving Bruges. The reinsurer took the riskier Cadiz-to-Sluys leg.³ For several centuries, reinsurance remained informal: ad hoc agreements between merchants sharing risk on specific voyages.

Formalisation (19th century)

The Great Fire of Hamburg (1842), which destroyed about a third of the old town,⁴ demonstrated that insurers needed systematic risk transfer. It led directly to the first dedicated reinsurance company: Cologne Re, chartered on April 8, 1846, which wrote its first reinsurance treaty in 1852.³ Swiss Re followed in 1863, Munich Re in 1880.³ Cologne Re proved the model, but Swiss Re and Munich Re became the lasting leaders: both still rank at the top of the industry. Cologne Re survives inside Gen Re, which Berkshire Hathaway has owned since 1998.⁵

Catastrophe era (1990s–2000s)

Two events transformed the modern market:

• Hurricane Andrew (1992): about $15 billion in insured losses at the time,⁶ or $27.3 billion in 2017 dollars.⁷ Published insurer-failure counts differ: the Insurance Information Institute (III) attributes at least 16 insurer failures in 1992–1993 to Andrew,⁷ while an analysis published by the National Association of Insurance Commissioners (NAIC) counts 11 direct insolvencies.⁶ Either way, the industry realised its catastrophe models were inadequate and its capital reserves were insufficient.
• September 11, 2001: about $32.5 billion in insured losses at the time, or $42.9 billion in 2013 dollars.⁸ Demonstrated that catastrophe risk was not limited to natural perils. Triggered a sharp hardening of the reinsurance market and a large inflow of capital.

After these events, catastrophe modelling, capital regulation, and alternative capital reshaped the business:

1. Catastrophe modelling became the basis of the business: rating agencies and regulators began asking for modelled loss estimates, and historical loss data on its own no longer sufficed⁹
2. Capital adequacy requirements were formalised by regulators. The NAIC’s risk-based capital standards, prompted by the insurer insolvencies of the 1980s, were adopted for property and casualty insurers in 1993, and Solvency II applied in Europe from 2016¹⁰ (Market participants names today’s frameworks)
3. Alternative capital entered the market through insurance-linked securities (ILS)¹¹

The forces that have shaped the market since (climate, exposure growth, alternative capital, analytics, consolidation, regulation) are the subject of How the market is evolving below.

Where reinsurance operates

The market is a network of interconnected risk transfer, not a simple chain. Multiple policyholders aggregate into a single insurer. Each insurer spreads its risk across multiple reinsurers. Reinsurers in turn buy protection from retrocessionaires, and a retrocessionaire can itself be another reinsurer, standing on both sides of the market (Reinsurer P below). The diagram traces this many-to-many structure.

Geographically, that network is concentrated in a handful of cities.

Major hubs

City	Significance
Zurich	Headquarters of Swiss Re, the largest reinsurer by year-end 2024 reinsurance revenue; on AM Best’s year-end 2025 edition the two places swap on exchange rates alone12
Munich	Headquarters of Munich Re, the second-largest on the same year-end 2024 basis, the long-time leader on earlier rankings, and first again at year-end 202512
London	Lloyd’s of London, tracing its origin to Edward Lloyd’s coffee house in the 1680s and, on its own description, the world’s leading market for specialist insurance13
Bermuda	Domicile for many reinsurers and ILS funds, and where $8.5 billion of new reinsurance capital was raised after 9/11; historically a no-income-tax jurisdiction, though a 15% corporate income tax has applied to large multinational groups since 202514
Singapore	Growing hub for Asian reinsurance business15
New York / Hartford	US domestic reinsurance market; Hartford is the historical US insurance capital16

Market size

The International Association of Insurance Supervisors (IAIS) reports gross reinsurance premiums of roughly $900 billion by the end of 2023, a total that spans both life and non-life reinsurance.¹⁷ It is a concentrated market: the ten largest groups together reported about $215 billion of reinsurance business for 2024, the top three about $96 billion.¹²

#	Company	2024 volume	Measure
1	Swiss Re	$36.2B	Reinsurance revenue (IFRS 17)
2	Munich Re	$32.6B	Reinsurance revenue (IFRS 17)
3	Hannover Re	$27.5B	Reinsurance revenue (IFRS 17)
4	Berkshire Hathaway	$26.9B	Gross premiums written
5	Lloyd's (marketplace)	$23.5B	Gross premiums written
6	SCOR	$16.8B	Reinsurance revenue (IFRS 17)
7	Reinsurance Group of America	$15.6B	Gross premiums written
8	Everest Group	$12.9B	Gross premiums written
9	RenaissanceRe	$11.7B	Gross premiums written
10	Arch Capital	$11.1B	Gross premiums written

The top three carry close to half of the top ten’s volume. AM Best publishes two lists (one per reporting basis) and declines to compare across them; the single order in the table and the chart is this site’s merge by volume, which is why the chart colours each bar by the basis its company reports on.

The ten largest reinsurers by 2024 volume, from AM Best's year-end 2024 data. Slate-blue bars are reinsurance revenue under IFRS 17; ochre bars are gross premiums written. AM Best does not rank across the two measures; the top-to-bottom order is this site's merge by volume, so the order across colours is approximate. Hover a bar for its volume and measure.

Source: AM Best, “World’s 50 Largest Reinsurers,” August 2025, year-end 2024 data.¹² AM Best ranks International Financial Reporting Standard 17 (IFRS 17) reporters by reinsurance revenue and other reporters by gross premiums written; the two measures are not directly comparable, so the interleaved ordering is approximate. S&P Global Ratings’ 2025 top-40, ranked on year-end 2024 gross premiums written throughout, also places Swiss Re first, ahead of Munich Re and Hannover Re.¹⁸ Lloyd’s is a marketplace, not a single entity.

Beyond the top tier, the market includes dozens of mid-sized and smaller reinsurers. Concentration is not increasing at the top. The next twelve largest grew from 30% of the market in 2015 to 38% in 2023, while the top three fell from 34% to 27%.¹⁹ The smaller players below them (including specialty firms) often concentrate on niches where the large incumbents are less competitive. Some, in this site’s assessment, also adopt technology faster, because they have less existing technology to maintain. Neither trait is guaranteed by size. It varies company by company. What the mid-tier reliably provides is balance in a market that might otherwise consolidate around a few dominant players.

Lines of business

Reinsurance covers many lines of business, but this site focuses on property catastrophe reinsurance, the segment whose pricing runs directly on simulated loss distributions. Even within property catastrophe, the market distinguishes between sub-segments: residential, commercial, industrial, and others, each with its own exposure characteristics and modelling approaches.²⁰ Property cat reinsurance covers losses from natural catastrophes (hurricanes, earthquakes, floods, wildfires), and it is where, in this site’s assessment:

• Losses are volatile and heavy-tailed
• The catastrophe model carries the loss estimate
• The contract structures compose in the most ways
• The analytics are the hardest to engineer

Watch Out

“Reinsurance” does not mean only catastrophe risk. Reinsurers also write casualty, specialty, life, and health business. However, property catastrophe reinsurance is where the analytical complexity (and the content of this site) is concentrated. When we say “reinsurance analytics” throughout, we mean property catastrophe analytics unless stated otherwise. The site’s scope statement and conventions index draw the full boundary.

Climate, exposure growth, and alternative capital are changing this segment, and the analytics with it.

How the market is evolving

Several forces are reshaping reinsurance, and each one changes something specific about the analytics systems reinsurers build:

Force	What it changes	What it asks of the analytics
Climate change	The frequency and severity of some weather perils	Several views of risk held at once, and a discipline for reconciling model versions
Exposure growth and urbanisation	The insured value standing in catastrophe-prone areas	Current exposure data behind every loss set
Alternative capital	Who carries the risk, and how the capital behind a contract is backed	One computation that serves traditional and collateralised forms alike
Real-time analytics	Portfolio analysis from a quarterly exercise to an on-demand one	The roll-up re-run as the portfolio changes, so a new contract’s marginal impact is available on demand
Model proliferation	One model per peril to several, each in several versions	Runs that multiply, and a reconciliation between the views they produce
Consolidation	Fewer independent reinsurers, and narrower panels to place with	Larger portfolios and more programs per system
Regulation and accounting	What must be computed, reported, and audited	Solvency computations and reported figures produced on a supervisor’s schedule

Climate, exposure growth, and consolidation

Climate change is shifting the frequency and severity of some weather perils: the observed signal is strongest for heat extremes and heavy precipitation, and weaker or mixed for others, such as tropical cyclone frequency.²¹ Growth in insured losses is also driven by exposure growth, urbanisation, and inflation:²² urbanisation concentrates more insured value in catastrophe-prone areas, so the same event costs more each decade. On the supply side, mergers have reduced the number of independent reinsurers, and cedents have tended to rationalise the panels they buy from,¹⁹ though the largest three have been losing share to the tier below them rather than gaining it. The market remains cyclical: prices harden after a major catastrophe and soften as new capital arrives; Market participants sets out the stages of that cycle and what each does to risk appetite and pricing.

Convergence of traditional and alternative capital

Traditional reinsurance and the capital markets increasingly overlap. Capital market instruments such as catastrophe bonds and insurance-linked securities, along with collateralised reinsurance vehicles, now supply about 17% of global reinsurer capital: Aon estimates $136 billion of alternative (third-party) capital against $785 billion of total reinsurer capital at year-end 2025.²³ Note the denominator: that is a share of the industry’s capital base, not of the coverage limit it sells. This means analytics systems must handle both traditional contract structures and these alternative capital forms; the vehicles include funds, cat bonds, sidecars, and collateralised reinsurance.

Real-time analytics

Historically, portfolio analysis was a quarterly or annual exercise, and the tooling was spreadsheet-based modelling.²⁴ The cadence is shortening, and what the market is working toward is pricing a new contract and seeing its marginal impact on the portfolio in the same sitting.

Model proliferation

Relying on a single catastrophe model from a single vendor is no longer the norm, and a company that does is expected to justify the view it uses.²⁵ Multiple models, multiple views of risk, and internal adjustments create a combinatorial explosion of analytical runs. Systems must be designed to handle this complexity without becoming unmaintainable. Vendors also release new model versions, and each one moves the loss numbers downstream. Reconciling the old view of risk with the new (model change analysis) is a discipline of its own and out of scope for this site.

Regulatory evolution

Two distinct kinds of regime are increasing the analytical requirements, and they should not be conflated. Prudential capital frameworks, Solvency II in the EU²⁶ and the Swiss Solvency Test (SST) in Switzerland²⁷, govern how much capital a (re)insurer must hold to survive a bad year and require it to demonstrate solvency to its supervisor. Which frameworks apply where, and which risk measure each computes, is set out under Market participants.

IFRS 17, by contrast, is an accounting standard that sets out how insurance contracts are recognised, measured, presented, and disclosed in financial statements.²⁸ Both demand more rigorous and more frequent computation (regulatory capital, solvency monitoring, and financial reporting each depend on the same analytics), so those systems are no longer optional: a supervisor can require the solvency computation, and an auditor the reported figures.

---

The market’s participants start with cedents, reinsurers, and brokers, and extend to the capital and the model vendors behind them.

Footnotes

1. Paul Embrechts, Alexander McNeil, and Daniel Straumann, “Correlation and Dependence in Risk Management: Properties and Pitfalls”, ETH Zürich, 1999: “insurance has traditionally been built on the assumption of independence and the law of large numbers has governed the determination of premiums.” ↩

2. EIOPA, “Calculation of the Solvency Capital Requirement”, Solvency II Single Rulebook (Directive 2009/138/EC, Article 101). The Solvency Capital Requirement is the Value at Risk of basic own funds at a 99.5% confidence level over one year, a one-in-two-hundred-year loss the undertaking must be able to absorb. Lloyd’s, “Realistic Disaster Scenarios”. Every syndicate reports estimated losses against a set of compulsory catastrophe scenarios, plus two events it defines for itself. ↩

3. David M. Holland, “A Brief History of Reinsurance”, Reinsurance News, Issue 65, Society of Actuaries, February 2009 (the SOA has since moved the file; the link is the Internet Archive’s April 2026 capture). The first known reinsurance contract, on a Genoa cargo, dates from 1370, three centuries before the first insurance companies. Invitations to found a reinsurer in Cologne went out in December 1842, statutes were drafted in 1843, Cologne Re was founded on April 8, 1846, and its first treaty was written in 1852. ↩ ↩² ↩³ ↩⁴

4. Wikipedia, “Great fire of Hamburg”. About a third of the buildings in the old town destroyed, some 1,700 residential buildings among them. ↩

5. Gen Re, “About Us”. Cologne Re (1846) and General Re (1921, acquired by Berkshire Hathaway in 1998) have marketed globally as Gen Re since 2003. ↩

6. Jack E. Nicholson, Karen Clark, and Glen Daraskevich, “The Florida Insurance Market: An Analysis of Vulnerabilities to Future Hurricane Losses”, Journal of Insurance Regulation, Vol. 37, No. 3, NAIC. Andrew “caused $15 billion in insured losses and resulted in 11 insurer insolvencies.” ↩ ↩²

7. Insurance Information Institute, “Hurricane Andrew Fact Sheet”. Insured losses of $27.3 billion in 2017 dollars; “at least 16 insurer failures in 1992 and 1993.” ↩ ↩²

8. Insurance Information Institute, “Terrorism and Insurance: 13 Years After 9/11 the Threat of Terrorist Attack Remains Real”, September 9, 2014, as distributed by PR Newswire (the institute’s own copy has moved): “insured losses of about $32.5 billion, or $42.9 billion in 2013 dollars.” ↩

9. AM Best, “Catastrophe Analysis in AM Best Ratings”, Best’s Methodology and Criteria, 2021. AM Best calculates the catastrophe capital requirement in its capital adequacy ratio from company-provided modelled loss estimates, notes that “most insurers use sophisticated catastrophe modeling tools”, and warns that models “cannot be solely relied upon for the management of maximum exposures”. Florida Statutes, § 627.0628(3)(d). For residential property rate filings an insurer “shall employ and may not modify or adjust” methods and models the Florida Commission on Hurricane Loss Projection Methodology has found accurate or reliable. Modelling is required in that narrow sense and expected in the broader one; no source supports a general obligation. ↩

10. NAIC, “Risk-Based Capital”. RBC is “a statutory minimum level of capital that is based on two factors: 1) an insurance company’s size; and 2) the inherent riskiness of its financial assets and operations”; the life formula was adopted in 1992 and implemented in 1993, the property/casualty model act was adopted in 1993, and “a large number of insurer insolvencies in the 1980s was the driving force for the NAIC’s RBC standard”. European Commission, “Solvency II Overview — Frequently Asked Questions”, MEMO/15/3120, January 2015, which dates the directive to 2009 and its application to January 2016. ↩

11. Andy Polacek, “Catastrophe Bonds: A Primer and Retrospective”, Chicago Fed Letter No. 405, Federal Reserve Bank of Chicago, 2018. Eight insurers failed after Hurricane Andrew, leaving demand that “new capital had to flow into reinsurance” to meet, which is what the catastrophe bond was created for; the first were issued in 1997. ↩

12. AM Best, “World’s 50 Largest Reinsurers”, Market Segment Report, August 18, 2025. Year-end 2024 rankings (Exhibit 1). ↩ ↩² ↩³ ↩⁴

13. Lloyd’s, “History”. Lloyd’s traces its origin to Edward Lloyd’s coffee house in the 1680s and describes itself as “the world’s leading market for specialist insurance” and “the only insurance marketplace of its kind in the world”. It does not claim to be the oldest, and older insurance institutions exist. ↩

14. Association of Bermuda Insurers and Reinsurers, “World’s climate risk capital: a natural progression for Bermuda”. Bermuda’s “Class of 1993” raised $4.8 billion after Hurricane Andrew and the “Class of 2001” $8.5 billion after the September 11 attacks. PwC, “Bermuda: Taxes on corporate income”, Worldwide Tax Summaries. Bermuda historically imposed no tax on profits, income, dividends, or capital gains; the Corporate Income Tax Act 2023 applies a 15% rate from 2025 to Bermuda entities of multinational groups with revenue of EUR 750 million or more. ↩

15. Gan Kim Yong, “Driving Asia’s Growth: The Role of Reinsurance in Sustainable Development”, keynote address at the 20th Singapore International Reinsurance Conference, Monetary Authority of Singapore, November 2024. Singapore’s reinsurance premiums grew 31% year on year to S$27.6 billion in 2023, about 21% of Asia’s reinsurance market, and 16 of the world’s top 25 reinsurers run their regional operations from there. The speaker is the regulator’s chairman, so the ambition in the speech is the regulator’s own; the figures are MAS statistics. ↩

16. Connecticut Humanities, “Aetna Helps Make Hartford ‘The Insurance Capital of the World’”. Hartford earned the nickname during the 19th century, as insurers including Hartford Fire in 1810 and Aetna in 1819 established their headquarters there. ↩

17. International Association of Insurance Supervisors, Global Insurance Market Report 2024, December 2024: “reported gross reinsurance premiums reaching $900 billion by the end of 2023.” ↩

18. S&P Global Ratings’ 2025 top-40 global reinsurers ranking, based on year-end 2024 gross premiums written, as reported in “Swiss Re, Munich Re and Hannover Re Top List of 40 Largest Reinsurers in 2025: S&P”, Insurance Journal, September 9, 2025. ↩

19. Swiss Re Institute, M&A in insurance: start of a new wave?, sigma No 3/2015. Cedents have tended to “rationalise the set of reinsurers they use”, and some classes “can often only be underwritten by reinsurers with large capital buffers”, pressure that falls hardest on small and mid-tier firms. Raphael Troitzsch, Nadine Moore, and Gianpaolo Caramellino, “Where’s the Growth for Reinsurers? How to Outperform”, Boston Consulting Group, May 2025. The top three reinsurers’ combined share fell from 34% in 2015 to 27% in 2023, while “the combined share of the 12 next-largest companies grew steadily, from 12% in 1980 to 30% in 2015 to 38% in 2023.” ↩ ↩²

20. Paul Sigona, Catastrophe Modeling Overview, Casualty Actuarial Society Reinsurance Boot Camp, August 2019. Catastrophe models treat personal lines property (the site’s “residential”), commercial property, industrial property, and builders risk as distinct modelled lines of business, each with its own occupancy and vulnerability assumptions. ↩

21. IPCC, AR6 Working Group I, Chapter 11: “Weather and Climate Extreme Events in a Changing Climate”, 2021. Increases in heat extremes are virtually certain and increases in heavy precipitation likely over most land regions, while trends in other extremes, such as tropical cyclone frequency, carry lower confidence. ↩

22. Swiss Re Institute, “Hurricanes and earthquakes could lead to global insured losses of USD 300 billion in a peak year”, press release, April 29, 2025. Underlying risk has grown with “economic and population growth as well as urban sprawl”; a repeat of Hurricane Andrew’s 1992 track today would cost roughly three times as much. ↩

23. Aon, Reinsurance Market Dynamics, April 2026. Global reinsurer capital of $785 billion and third-party capital of $136 billion at December 31, 2025. ↩

24. Verisk, “4 Reasons for Rolling-Up Reinsurance Portfolios in Real Time”, March 2018. Roll-ups “were done annually because of the time and resources needed”, and in a 2017 poll of the vendor’s own European user-conference attendees only 4% still rolled up yearly, 65% had moved to quarterly, and the rest went more often. It is a model vendor describing demand for its own product, and the poll has no stated sample size; it is the only source found that states the before-and-after cadence in numbers. ↩

25. AM Best, “Catastrophe Analysis in AM Best Ratings”, Best’s Methodology and Criteria, 2021: “The availability of several specialized tools for modeling catastrophes allows for a range of perspectives on a company’s loss exposure”, and companies are expected to “explain why the outputs selected are the most appropriate for capturing their catastrophe exposures”. NAIC, “Catastrophe Models (Property)”, Center for Insurance Policy and Research, March 2024, which records internal company models, vendor models with adjustments or different weights, and derivative models built on vendor models as the forms in regulatory use. The run counts that follow are this site’s inference, not the regulator’s. ↩

26. European Commission, “Solvency II Overview — Frequently Asked Questions”, MEMO/15/3120, January 2015. ↩

27. FINMA, “Swiss Solvency Test (SST)”. ↩

28. IFRS Foundation, “IFRS 17 Insurance Contracts”. Effective for annual reporting periods beginning on or after January 1, 2023. ↩