Record reinsurance capital is not closing the catastrophe gap

Aon estimates global reinsurer capital reached a record $785 billion at the end of 2025, up almost 10% on the year, while the price of deploying it keeps falling; Guy Carpenter's rate-on-line index moved from a 12% decline at January 1 to 16% at the July 2026 renewals, the steepest midyear slide in decades, per AM Best. Cat bond issuance set a record $24.9 billion in the twelve months to June 30, leaving $63.4 billion of cat bonds outstanding and total ILS capital at $144.5 billion.

None of it reached the losses. Insurance covered less than half of global catastrophe losses in every year from 2015 through 2025, Moody's analysis reported by Insurance Business America finds, with 57.8% of losses over the period going uninsured. In the US, about 84% of earthquake exposure carries no coverage, a figure the analysis attributes to optional coverage, cost and relatively low take-up; under a one-in-200-year aggregate US catastrophe, Moody's estimates more than $400 billion of insured losses against more than $1.1 trillion of total losses to insurable property, leaving a gap of roughly $700 billion.

The reflex at a moment like this is to read the gap as a capacity problem and the ILS market as the answer. The scale does not cooperate: total ILS capital sits at $144.5 billion and outstanding cat bonds at $63.4 billion, against a $700 billion hole from one severe US event. Those are different scopes, global investor capital against a single country's tail, and the comparison is rough on purpose because the market described as the fix for the protection gap is smaller than the gap one bad year in one country would open. ILS was built to let the industry finance its own tail, while the exposures it declines to write require a different buyer and a different business.

The asset class is doing what it was built to do, which is to let institutional investors take a funded share of catastrophe exposure without insurers and reinsurers carrying all of it on their own balance sheets — a transfer of risk already underwritten, not an extension of coverage to risk never written. The 84% of US earthquake exposure that goes uncovered is uncovered because the coverage is optional, expensive relative to what it protects, and bought by too few households; reinsurers have not run out of retro, and capacity at the top of the tower never reaches that decision.

The correlation makes it harder still, because Moody's finds the gap is widest for risks capable of producing severe losses across large numbers of policyholders at once — the non-diversifiable end of the catastrophe spectrum and precisely the exposure that offers an ILS investor the least diversification. The geography-specific repricing already visible this year, when a Hong Kong fire, a Colombia quake and a Midwest derecho each pulled their own terms out of the market, reads the same way from the other side of the desk: capital prices what it can model and spread, and the largest gaps sit where it can do neither cleanly.

The capital is landing in structure rather than rate, with Guy Carpenter reporting cedents exploring parametric solutions and sidecars alongside traditional protection and abundant capacity supporting broader coverage options. That is the terms capitulation, as this publication has argued since January: reinsurers and the capital-markets funds competing with them give away attachment points, breadth, recovery speed and trigger design rather than headline price, because structure is where margin lives. It is also where basis risk lives — a sidecar or a parametric trigger moves the distance between the modelled loss and the actual one onto the investor's book, and that trade reads as cheap right up until the trigger does not fire.

The buyers are not waiting to be sold it: Liberty Mutual placed capital-markets hires on both sides of the policy within a week in September, building a structuring capability for large commercial risk, with reporting lines that indicate the carrier means to quote the structure alongside the paper. A large cedent holding the pen on its own risk-transfer architecture says something about where advantage sits in a soft market: it accrues to the balance sheet willing to hold volatility a fund cannot, more than to whoever posts the cheapest quote.

The mechanism pulling outside money in has nothing to do with the protection gap: global bonds and equities totaled about $319 trillion at the end of 2025, per Moody's, a scale at which a $1.1 trillion worst-case US event looks substantial against insurance-sector capital and modest against the capital markets. Institutional investors seeking a diversifying return do not need the gap to narrow any time soon.

So watch January 1. If the rate-on-line index falls again while parametric attachments and sidecar capacity keep widening, the market will have shown it has more to concede on terms than on price, and the pressure moves to the reinsurers holding the volatility. The 84% will not move either way. Closing it takes something on the demand side, whether a subsidy, a mandate or a product cheap enough that households actually buy it, and a record capital year is not that thing.