Gallagher Re: Reinsurers' 19.9% Return Is 13.8% Underlying, and 2027 Is Priced Off the 19.9%

Michael van Wegen, head of international for Gallagher Re's Global Strategic Advisory team, framed the first half as evidence that "the reinsurance industry remains in a position of exceptional financial strength" (Reinsurance News, September 2026). His own report quantifies how much of that strength repeats. The Gallagher Reinsurance Composite returned 19.9% for half-year 2026 and 13.8% on an underlying basis, a 6.1-point wedge built entirely from items that do not arrive on a schedule.

The composite is the twelve large Bermudian and Big Four European reinsurers that disclose enough to decompose. Its undiscounted combined ratio reached a record-low 85.8%, improved from 92.1%. Catastrophe losses alone improved that ratio by 6.8 points. The ratio as a whole improved by 6.3. Everything other than catastrophes therefore got worse.

Key Takeaways

• 6.1 points separate the reported 19.9% ROE from the 13.8% underlying return. Below-normal catastrophes account for 3.4 points and prior-year reserve releases 1.5, with the remaining 1.2 from IFRS 17 discounting and investment gains.
• 94.2% underlying combined ratio, deteriorated by 0.9 points year over year, sits behind the record-low 85.8% headline. Strip catastrophes and reserve development entirely and the accident-year ratio deteriorated 1.1 points to 84.7%.
• $46 billion of insured catastrophe losses on Gallagher Re's count, 28% below the $64 billion ten-year average, and the composite absorbed only 3.8% of them against 6.6% a year earlier. Primary carriers retained the rest.
• $688 billion of dedicated capital, up 5%, against capital demand down 0.2% on a revenue proxy. Supply grew at twenty-five times the rate the thing it supports shrank.
• Revenue fell 6.1% while shareholders' equity kept building, which is the second reason underlying ROE fell. The denominator is outrunning the numerator even before margin erosion.

What the 6.1 Points Are Made Of

The reported combined ratio improved 7.4 points, from 86.3% to 78.9%. Gallagher Re strips the IFRS 17 discounting benefit before comparing across time and reporting regimes, and that benefit itself widened from 5.8% to 7.0%. So 1.2 points of the reported improvement is the interest-rate environment applied to reserve duration, not underwriting.

On the undiscounted basis the improvement is 6.3 points, from 92.1% to 85.8%. Against that, the catastrophe load fell 6.8 points, from 10.4% to 3.6%, as the California and Los Angeles wildfires dropped out of the comparison. Favourable prior-year development added a further 0.6 points of benefit, to 2.5%.

Remove both and the arithmetic inverts. The accident-year ratio excluding catastrophes and development deteriorated by 1.1 points, to 84.7%. Replace actual catastrophes with the report's five-year normalised load and the underlying combined ratio deteriorated 0.9 points, to 94.2%.

Combined ratio bridge, Gallagher Reinsurance Composite	2025 HY	2026 HY	Change
Reported combined ratio	86.3%	78.9%	Improved 7.4 pts
IFRS 17 discounting benefit removed	5.8%	7.0%	Benefit up 1.2 pts
Undiscounted combined ratio	92.1%	85.8%	Improved 6.3 pts
Natural catastrophe loss impact	10.4%	3.6%	Improved 6.8 pts
Prior-year development benefit	1.9%	2.5%	Benefit up 0.6 pts
Accident-year ratio, ex-cat and ex-development	83.6%	84.7%	Deteriorated 1.1 pts
Underlying combined ratio (normalised cat)	93.3%	94.2%	Deteriorated 0.9 pts

The ROE waterfall runs the same way. Gallagher Re attributes 3.4 points of the headline return to catastrophes landing below normal and 1.5 points to prior-year development, the highest development contribution in three years. Discounting and investment gains supply the remaining 1.2. Underlying underwriting margin, the component that actually reflects the price of risk, fell from 4.3% to 3.2%.

The Load an Actuary Prices Off

Renewal negotiations run on calendar-year disclosures, so a cedent's broker opens January discussions with the published 85.8% and 19.9%, and both of those numbers are true. Neither is a statement about the price of risk.

The report supplies the correction. Its ex-catastrophe accident-year ratio is 84.7% and its underlying ratio 94.2%, so the normalised catastrophe load embedded in the underlying figure is 9.5 points. Actual first-half catastrophe cost was 3.6 points. An actuary who reprices off the reported result is carrying a catastrophe load 5.9 points light before a single point of rate reduction is agreed.

That is not a rounding difference. AM Best recorded property reinsurance rates falling 10% to 20% at January 2026, and Moody's survey of 40 primary insurers found a growing share expecting further declines into January 2027, with property catastrophe prices already down more than 20% in the 18 months since 2024. Aon has guided to roughly another 10% at January 1. The catastrophe-load error is comparable in size to the entire rate movement being negotiated.

Some reinsurers are refusing to let the benign half earn through. Hannover Re books catastrophe losses at the higher of actual or budget, so its reported first-half figure reflects the EUR800 million budget against EUR436 million of actual losses. Gallagher Re notes Swiss Re, Hannover Re and SCOR all signalling buffer building through reserve strengthening despite the favourable environment, an approach the site tracked when Swiss Re raised loss assumptions 4.2% while net prices fell 5.3%. That is the discipline the composite average conceals: reserve releases still reached 2.5 points of the combined ratio, roughly $1.2 billion and up 21% year over year, concentrated in short-tail and property lines.

Capital Buffer Is Not Price Adequacy

Gallagher Re's headline reassurance is that the sector could absorb a $50 billion to $75 billion insured event on top of normal second-half catastrophes and still earn its 11.4% cost of equity, and that it would take $150 billion or more to erase the excess capital build. Both are solvency statements. They answer whether the industry survives a bad year, not whether it is charging enough for the risk it is writing.

The distinction shows up in how little the capital return is achieving. Dedicated capital rose 5% to $688 billion while capital demand fell 0.2%. Among the pure reinsurance groups, net income of $16 billion was met with $15 billion of buybacks and dividends, a 92% payout that still left capital down only 1%. Hannover Re, Munich Re and Swiss Re each returned slightly more than 100% of first-half profit, around 13% of opening shareholders' equity apiece. Arch Capital repurchased almost $2 billion against roughly $360 million a year earlier.

None of it reached the fastest-growing part of supply. Non-life alternative capital rose 9% in the half to a record $147 billion (Artemis, September 2026), and Aon Securities put catastrophe bond issuance at $24.9 billion for the 12 months to 30 June, 15% above the previous record, with $63.4 billion outstanding. ILS investors do not conduct buybacks. The one lever management holds cannot reach the capital that is setting the marginal price, and alternative capital is now extending into casualty as well.

Meanwhile the buffer being celebrated is partly an artefact of the same benign half. Swiss Re Institute put global insured catastrophe losses at $42 billion, the lowest first half since 2020 and well under the $66 billion trend, with only 11 billion-dollar events against a ten-year average of 16. The premium-to-capital ratio for reinsurance groups has fallen to 86% from 95% to 97% in prior years, which is what excess capital looks like from the inside.

Gallagher Re estimates it would take 6 to 7 points of incremental combined ratio deterioration sustained for three years, beyond the earn-through of 2026 softening, to consume the $13 billion of cumulative excess profit the composite has banked since 2017. The underlying ratio moved 0.9 points in a single half, and the 2026 rate reductions have not earned through yet.

Further Reading on actuary.info

• Gallagher Re's Record $648 Billion of Dedicated Capital - the full-year 2025 edition of this same report, and the base the $688 billion is measured from.
• AM Best's 6.7-Point Spread and the Hurdle That Moved Underneath It - the cost-of-equity side of the same comparison, where 11.4% and 13.8% leave a 2.4-point cushion.
• Fitch: Big Four Reinsurers' Record 21.5% ROE Masks Price Cuts Still to Earn Through - the same half read at the European peer level, with the IFRS 17 discount effect isolated.
• Lloyd's H1 2026: The Underlying Combined Ratio Behind the Headline - the identical headline-versus-underlying split in the Lloyd's market.
• AM Best: Record Capital Tests Reinsurance's 87% Combined Ratio - the reserve-cushion question across the wider US, Bermuda and European peer set.
• Swiss Re Raised Loss Assumptions 4.2% Even as Net Prices Fell 5.3% - what repricing against updated loss picks rather than reported margin looks like.
• AM Best on Third-Party Capital Growth - the alternative-capital side of the supply build that buybacks cannot reach.