Reinsurance Illiquidity: Why Record Capital Still Costs Too Much

Global reinsurer capital hit a record $785 billion in April 2026, and property catastrophe rates fell 14.7% in January, another 16% in April, and up to 25% at June 1. Classical economics says the cost of that capital should be falling with it.

It is not. Howden Re's David Flandro attributes the floor to structural illiquidity: a reinsurance contract cannot be traded, transferred or exited before maturity, and that lock-up carries a premium no amount of supply removes.

Key Takeaways

• $785 billion of total global reinsurer capital, $649 billion traditional and $136 billion alternative, on roughly 10% year-over-year growth with alternative capital up more than 18%.
• Weighted-average rate-on-line fell up to 25% at the June 1 renewal on a capacity ratio of 1.6, with Florida programs down 17.5% to 20%.
• The ROE spread is halving: a 19.3% composite return in 2025 against an 8% to 10% cost of capital becomes a projected 14% to 15% in 2026, cutting the margin from roughly 10 points to 5 or 6.
• Munich Re cut its April book 18.5% and Swiss Re reduced nat cat volumes 11% in Q1. Both are decisions not to renew, not exits, which is the constraint in one line.
• 272 trades in a single month was the busiest cat bond secondary month on TRACE record, which is extraordinary for the asset class and trivial against any comparable bond market.

The Capital That Will Not Cheapen

Aon's April 2026 figure of $785 billion breaks into $649 billion traditional and $136 billion alternative, growing roughly 10% year over year, with traditional capital up more than 8% and alternative capital up more than 18%. Gallagher Re's narrower dedicated-capital measure came to $648 billion, an 11% increase and the second-strongest growth year in more than a decade.

The pricing side is behaving as that supply predicts. Property catastrophe declined 14.7% in January, 16% in April and up to 25% at June 1, on a June capacity ratio of 1.6. Florida programs renewed down 17.5% to 20%. Retrocession fell mid-teen percentages, and even cyber reinsurance dropped as much as 25%.

The return side is not. Gallagher Re put the 2025 composite ROE at 19.3%, among the highest since the series began, with 2026 projected at 14% to 15%. Fitch estimates 15.5% for 2026 against a cost-of-capital range of 8% to 10%. The spread compresses from roughly 10 percentage points to 5 or 6 in a single year.

What is anomalous is not the compression, which any softening market produces. It is that the cost-of-capital floor is not moving down with the supply increase. In most markets a capital surge lowers the return required to attract it. Here the required return stays anchored.

What the Lock-Up Does to the Spread

The mechanism is the contract itself. A treaty is a bespoke bilateral agreement with no standard specification, no exchange, no clearinghouse and no way for the reinsurer to exit before expiry. Capital deployed on January 1 is immobilized for the contract period whatever the market does afterward.

That asymmetry showed clearly this spring. Munich Re cut its April renewal book 18.5%, and Swiss Re reduced natural catastrophe volumes 11% in Q1. Neither sold anything, because neither could. Both declined new business while the existing book ran to maturity at its original terms. Ark's Outrigger Re sidecar returned 30% on invested capital across the 2023 to 2025 underwriting years for exactly the same reason in reverse: that vintage is locked at hard-market pricing and cannot be competed away.

The arithmetic is where it starts to bind. Take a 7% cost of equity for a comparable liquid investment and an illiquidity premium of 2 to 3 points, and the effective cost of reinsurance capital is 9% to 10%. Against a 19.3% ROE the premium is invisible inside a 10-point spread. Against a projected 14% to 15%, it consumes roughly 40% of the excess return.

Two actuarial consequences follow from that, and neither is visible in a rate quote.

Capital models generally release allocated capital when a treaty is placed, treating cession as a liquid mitigation. If the position cannot be traded or exited, that relief overstates real flexibility, and an illiquidity haircut on ceded capital credit, of the kind already applied to illiquid assets, would give a more accurate adequacy picture.

Pricing carries the same distortion in the other direction. A catastrophe load built on the current spot treaty rate understates the sustainable long-run cost of that capital, because the spot rate reflects a soft renewal while the cost of capital behind it has not fallen. That is the specific version of the bind we described when cheaper reinsurance started reaching primary pricing actuaries, and it is why Fitch declined to extend its cost-of-capital assurance past 2026.

The Cat Bond Exception Marks the Ceiling

Catastrophe bonds are the one place the industry built something tradeable, and their limits show how far the structure can be pushed.

Outstanding cat bonds reached $69.1 billion by May 2026, with H1 2026 issuance projected at $16.3 billion across roughly 72 transactions. Rule 144A adoption among Southeast U.S. cedents has reached 80%, inverting a ratio that sat near 20% five years ago. Non-life alternative capital in aggregate is $135 billion to $136 billion.

They trade over the counter with TRACE reporting, and the busiest month on record, April 2024, saw 272 trades. May and June 2024 each exceeded 250. Those are remarkable numbers for the asset class and negligible against corporate or even municipal bond volumes. Swiss Re described 2025 secondary turnover as subdued for large parts of the year, with cash building up across investor portfolios and conditions bid-heavy.

Below cat bonds the spectrum gets worse rather than better. Schroders Capital describes collateralized reinsurance and private ILS as effectively locked until maturity with no secondary market, and collateral can be trapped outright, including through maturity extensions that pay a spread while denying redeployment on schedule. Mark Gibson framed the choice as "how much liquidity and transparency they are prepared to forego in pursuit of additional spread."

This is why alternative capital growing from roughly $38 billion in 2012 to $136 billion has not arbitraged the premium down. It carries its own illiquidity cost. Guy Carpenter names five frictions that persist across the alternative market: modelling limits that create information asymmetry, trapped collateral, softening returns, cultural distance between capital markets and reinsurance practitioners, and opacity in bespoke risk allocation.

The result is that even institutional allocators building ILS books price the same premium the treaty market does. Liquidity in reinsurance is not a spectrum that ends in a liquid market. It ends at episodic.