Dedicated reinsurance capital hits record $663bn as cat budgets shrink

Dedicated reinsurance capital ended 2025 at $663 billion, the highest total on record, according to new analysis by AM Best and Guy Carpenter, while the share of that capital committed to catastrophe risk kept falling. The gap between what reinsurers hold and what the rating agency's model says they need to hold is now wide enough to matter, and it will be priced at the January 1 renewals.

The total rose from $607 billion at the end of 2024, splitting into $540 billion of traditional balance-sheet capital and $123 billion of third-party capital; traditional capital was up from $500 billion, and the ILS and collateralised reinsurance segment from $107 billion. AM Best credits strong underwriting results, investment earnings and investor demand for insurance-linked securities for the build-out, even as catastrophe probable maximum loss exposure remained broadly unchanged during 2025 and required capital stayed broadly flat, extending a multi-year trend.

AM Best's capital utilisation measure for traditional reinsurers — available capital relative to the amount required to maintain the Strongest BCAR of 25% at a 99.6% value-at-risk threshold — fell to 77% at the end of 2025 from 85% in 2024 and 92% in 2023. The agency expects 72% by the end of 2026, a decline driven mainly by rising available capital: catastrophe exposure was flat and required capital crept up only modestly.

The build-out lands at an awkward moment, as Europe's four biggest reinsurers posted a record 21.5% return on equity in the first half and Fitch Ratings has warned that renewal price cuts reaching 25% on natural catastrophe lines at mid-year will erode those gains; capital is arriving just as the market gives back rate. A 77% reading means the industry is carrying a substantial buffer above modelled requirements and the projected slide to 72% widens it, making the utilisation ratio a price-setting fact. The pressure to deploy that capital will show up first in terms — attachment points, coverage breadth, sub-limits — before it shows up in headline rates.

As this publication has argued, the soft cycle is turning from price to terms; the utilisation slide is the mechanism. Cedants with flat catastrophe budgets will arrive at the January 1 renewals with a stronger hand than their PMLs justify, and reinsurers will be tempted to spend the buffer to hold accounts — the spending will take the form of broader coverage rather than cheaper rates. The market can absorb a large loss without a rate response — that is the point of the buffer — but a terms-driven repricing does not need a loss to start.

The $16 billion jump in third-party capital, to $123 billion, suggests investor demand for insurance-linked securities remains strong. More collateralised capital chasing a flat stock of catastrophe risk is spread compression in its simplest form, because the yield for taking cat risk is set by the gap between the capital and the risk — and the gap just got wider. AM Best and Guy Carpenter project dedicated capital will reach $705 billion by the end of 2026; the buffer will keep growing, and the renewal terms signed in January will show whether it is spent on coverage or on price.