In my view, the clearest price anyone puts on climate risk isn't in a stock, a mortgage or a municipal bond. It's in a catastrophe bond, and that market had its biggest year on record in 2025.

A record year

Investors bought a record $25.6 billion of catastrophe bonds in 2025, according to Artemis, the trade publication that tracks the market and counts both publicly offered and private deals. That was up 45% from just under $17.7 billion in 2024, and it was the first year issuance ever topped $20 billion. By year-end, the amount outstanding had reached a record of just over $61 billion, up 24% in a year.

How a catastrophe bond works

A catastrophe bond works like an insurance policy that investors can buy and sell. An insurer, or sometimes a government, sets one up to protect itself against a disaster such as a major hurricane or earthquake. Investors put up money, which sits in a collateral account invested in high-quality, short-term securities. The sponsor pays investors a premium on top of what that collateral earns. If no qualifying disaster happens before the bond matures, investors generally get their principal back and keep the premium. If one does, and losses or a physical measure such as wind speed cross a threshold agreed in advance, some or all of the principal goes to the sponsor to pay claims.

In effect, investors rent out their capital as a backstop for a disaster, and the premium is the rent.

The loss side is real. In October 2025, Hurricane Melissa triggered a full payout of Jamaica's World Bank-supported catastrophe bond, Artemis reported. The investors' principal went to Jamaica.

Where the risk gets a number

The premium, meaning the spread above the collateral's return, is where the risk gets a price. It pays investors for the chance that a disaster wipes out their principal, and it is set again every time a new bond is sold. In early 2023, after Hurricane Ian, that risk premium averaged roughly 11%. By February 2026 it had fallen to about 5.2%, roughly where it stood before Ian, according to Bloomberg data reported by Insurance Journal.

That drop complicates the story, and it deserves attention. The hazard didn't fall by half between 2023 and 2026. My read is that the main change was in the supply of capital: more investors moved into the market, and more money competing for the same risk pushed the price down. A cat bond spread is a market price for catastrophe risk, but like any market price, it reflects how much capital wants the trade, and not only how dangerous the weather is.

Investors can and do lose principal when a qualifying disaster strikes. The thing that matters for this post is that the price, the spread investors demand to take on the risk, is set up front, before anyone knows the weather.

The same hazard, one step closer to home

Now look at the same hazard from the house's side. A Gulf Coast hurricane that could trigger a cat bond is also, in some form, a risk to the 30-year mortgage on a house in its path, to the owner's equity in that house, and to the municipal bond backed by the local tax base.

None of those markets puts an explicit, separately quoted price on that hazard the way a cat bond does. In The Risk Under the House I wrote that the guarantee fee Fannie Mae and Freddie Mac charge doesn't vary with flood exposure, and that the Congressional Budget Office counts the uncharged flood damage as a subsidy. The muni market is more mixed. As I described in What's Holding Up the Bond, one study found that sea-level-rise exposure began appearing in some municipal yields around 2013, while an industry analysis in 2022 found the broad market didn't yet price physical climate risk.

The comparison isn't one-for-one, and the reasons are worth stating. A cat bond usually covers only the extreme layer of losses above a high threshold, so its premium is concentrated where losses are rare and large. A homeowner's insurance policy sits between the house and the mortgage and absorbs much of the hazard before a lender ever feels it, and that premium is a price too.

My read is that the difference isn't that one risk is real and the other imagined. The hazard is quoted openly in one market and left implicit in the others, where it is folded into insurance, assumed away or simply not measured.

Measured least where it may matter most

Some of the largest banks have said something similar about their own books. In the Federal Reserve's 2023 pilot climate scenario exercise with six of the largest US banks, summarized in May 2024, participants reported "significant data and modeling challenges." Among them was a lack of comprehensive and consistent data on building characteristics and insurance coverage, and in many cases the banks relied on outside vendors to fill the gaps. The pilot was exploratory and carried no capital consequences. In October 2025, the Fed, the FDIC and the OCC withdrew their interagency climate risk-management principles for large banks. As far as I can tell, the data gaps the banks described haven't been closed. The interagency guidance that pushed banks to manage this risk has been withdrawn.

A price, quoted out loud

The catastrophe bond market's record year isn't a forecast about hurricanes or a promise of anything. It is a price, quoted out loud, for a risk that the mortgage, the home equity and much of the muni market still carry without a price of their own.

Sources

Figures current as of October 2026.