In March 2026, North Carolina's insurer of last resort got a $2.07 million check from its own catastrophe bond. Not because a hurricane hit the coast. Because the season stayed quiet.
That money goes to stronger roofs along the Outer Banks and the rest of the coastal book. It is the cleanest live example of a resilience bond, and a useful one. It is also still a bond. It holds collateral against a storm, pays investors a coupon, and pays the sponsor after the loss. Nothing in it keeps the dune line standing, the marsh behind the barrier islands wet, or the river floodplain inland able to take water.
A resilience bond is a bond whose money is tied to reducing disaster risk — either by earmarking what it raises for resilience projects, or by turning lower modeled losses into a rebate that pays for more risk reduction. It can be good paper. It rearranges who holds the risk and who holds the money. It does not, by itself, fund the living system that holds the loss back.
three things called a resilience bond
The term covers at least three structures. They share a name and not much else.
| type | how the money moves | best-known example | what it funds |
|---|---|---|---|
| use-of-proceeds | An ordinary bond. The issuer earmarks proceeds for a resilience portfolio. | EBRD's $700 million climate resilience bond (2019), aligned with the Climate Bonds Initiative's Climate Resilience Principles | Whatever qualifies in the issuer's portfolio: infrastructure, business operations, agriculture and ecological systems |
| rebate (catastrophe-bond based) | A cat bond priced on expected loss. When risk reduction lowers that loss, the savings become a rebate for more risk reduction. | Proposed by re:focus partners' RE.bound program (2015). First issued as the North Carolina Insurance Underwriting Association's Cape Lookout Re 2025-1. | Projects the sponsor picks. In North Carolina, IBHS FORTIFIED roofs. |
| outcome-based | Investors front the cost of the work. Beneficiaries repay as it gets done. | Blue Forest's Yuba I Forest Resilience Bond (2018): $4 million, 15,000 acres of the Tahoe National Forest, all investor capital returned in 2023 | Restoration of a named forest |
A fourth is taking shape. The Milken Institute's September 2026 report with Marsh, Ensuring a Better Future, lists a district-backed resilience bond among five models for pulling private capital into community resilience — alongside an insurer-engaged community plan, a revolving loan fund, a resilience technology fund, and a state policy playbook. Its starting point is that global disaster losses have topped $200 billion every year for a decade, and that today's financing and insurance structures aren't equipped to fund resilience or recovery on their own. We agree with the diagnosis.
how a resilience bond differs from a catastrophe bond
A catastrophe bond is collateral set aside against a named peril. Investors earn a coupon while the trigger stays quiet. If it fires, principal goes to the sponsor. The mechanics, the wider ILS market, and basis risk have their own posts. This one won't repeat them.
The rebate-style resilience bond keeps all of that and adds one channel. In the North Carolina structure, designed by Marsh's reinsurance securities team (GC Securities, part of Guy Carpenter) with Hannover Re as fronting reinsurer, the sponsor pays an extra 0.35% of the bond's original value as a "resilience spread," held in a dedicated account. If paid losses stay below 60% of the level where the bond would start paying out, the spread comes back to the sponsor for FORTIFIED roof programs. If losses run higher, investors keep it. On a $600 million deal, 0.35% is about $2.1 million — roughly the rebate that arrived after the quiet 2025 season. A second deal, Cape Lookout Re 2026-1, kept the feature.
Two things follow from the design, and neither is a complaint. The spread is the sponsor's own money, returned on a condition; in a bad year it becomes part of what investors earn for carrying the risk. And the bond's main job is unchanged. The rebate is a side channel. The note still pays after.
The re:focus authors framed it well in 2015. A cat bond is like life insurance, which doesn't make you healthier. A resilience bond is meant to work more like a health plan that rewards you for quitting smoking. That is the right ambition, and it also names the limit. The bond rewards the healthy choice. It doesn't make it.
what the paper does well
A resilience bond does three things most resilience funding can't.
It makes avoided loss legible. Modeling expected loss with and without the project turns "the storm that didn't wreck the house" into a number an investment committee can read. Marsh describes the goal as giving private capital "a monitorable proxy for a benefit that has historically been almost entirely invisible."
It routes money through someone with a reason to care. The North Carolina Insurance Underwriting Association covers the riskiest coastal property in the state and can't walk away from it. As of March 2026, NRDC reports, it had directly invested $84.7 million to strengthen 9,435 homes in its Beach Area. The bond tops up a program that already worked.
It targets something with an owner. A roof is one parcel, one decision-maker, one premium credit. That is why this structure works on wind. It is also where the structure stops.
what the paper can't hold
Every resilience bond sorts three questions: who holds the paper, who holds the risk, and who holds the living function that decides how big the loss gets. The bond answers the first two well. It rarely touches the third.
A resilience bond rearranges who holds the paper. It does not keep the floodplain wet.
Wind on a roof is a parcel problem. Surge across a barrier island, water spreading into a floodplain, fire running up a ridge — those are landscape problems. They are set by a dune system, a marsh, a forest stand, a river's room to spread. Those landscape-scale determinants have no single owner and no single payer, so no one sponsor's bond reaches them. Flood, for most US homes, isn't on a homeowners or wind policy at all. That isn't a flaw in the North Carolina deal. It is the edge of any deal written by one sponsor for one peril.
The other two structures have their own edges. A use-of-proceeds bond gives you the issuer's credit; the wetland, if there is one, is a line in an allocation report. An outcome bond like the Forest Resilience Bond does fund a living system directly. It is the closest cousin we have, and the beneficiaries who repay it are the right ones. Its shape is project finance: raise capital, finish the treatment, return the capital. That fits a restoration backlog. It is a different shape from holding the condition of a stand year after year.
Horizon matters too. Cat bonds typically run three to five years. A floodplain, a marsh, or a stand works on decades.
cat bond, resilience bond, certificate
| catastrophe bond | resilience bond (rebate) | certificate on a named place | |
|---|---|---|---|
| what you hold | A note collateralized against a named peril | The same note, plus a resilience spread | A position tied one-to-one to one place's account |
| when money moves | After a trigger event | Coupon during the term; rebate to the sponsor in low-loss years | Now, into the place's present condition |
| bad year | Investor principal pays the sponsor | Same, and investors keep the spread | No payout; the place was funded before the event |
| what gets funded | The sponsor's claims | Claims, plus the hardening the sponsor chooses | The living function on that place — the wet meadow, the marsh edge, the stand |
| horizon | Typically 3–5 years | Same | No maturity; funding lands when the position is taken |
| the living system | A boundary condition in the model | A possible project, if the sponsor picks it | The subject |
what ensurance does with the same hunt
Same hunt as the resilience bond: stop the loss before it arrives, and make that worth paying for. Different job.
The watershed, the wetland, and the stand exist whether or not anyone books them as a systemic-risk sleeve or wraps them in a bond. Ensurance funds that living function. It is not a stress-test score, and it is not a coupon.
Here is how the position works. Each place is represented by an agent — an onchain account for that watershed, marsh, or forest. A certificate is tied one-to-one to that agent; buying it funds the place's present condition, and proceeds route to its account. Coins are the broader version, funding protection protocol-wide through trading. The price is a bridge to that funding. It is not a claim that the marsh is worth its modeled avoided loss.
The two fit together. An insurer that sponsors a resilience bond for its roofs can also hold certificates on the landscape functions the bond can't reach. A district issuing resilience debt for pipes and pumps can fund the floodplain upstream the same way. Nature-based insurance and parametric reef covers sit in the same family of cousins, and none of them is replaced by this. Funding the thing that shrinks the loss is the investment case in full.
Our stage, plainly: coins and certificates are live on Base, and volumes are small. We have no rating, no coupon, no return series, and no modeled rebate. A certificate is not a note: no coupon, no principal at risk against a peril, no 144A offering, and no insurance policy behind it. Whether any instrument is a security is a facts-and-jurisdiction question for counsel, not a headline. Nothing here is investment, insurance, or legal advice.
next steps
- Sponsor risk transfer? See how insurers use ensurance alongside the book they already have.
- Want to see a position? Browse certificates on named places, then read what you actually hold.
- Structuring resilience debt and missing the living-system leg? Talk to someone who can help.
- Keep reading: who pays to shrink the systemic loss — the payors, one by one.
frequently asked questions
what is a resilience bond?
A resilience bond is a bond whose money is tied to reducing disaster risk. The term covers use-of-proceeds bonds that earmark funds for resilience projects, catastrophe bonds that rebate part of their cost when risk falls, and outcome bonds repaid by the beneficiaries of restoration work.
who issued the first resilience catastrophe bond?
The North Carolina Insurance Underwriting Association, with Cape Lookout Re 2025-1 — a $600 million named-storm cat bond with a resilience spread that funds FORTIFIED roofs in low-loss years. Hannover Re fronted the deal. It returned $2.07 million in March 2026, and a 2026 deal kept the feature.
how is a resilience bond different from a catastrophe bond?
A catastrophe bond transfers a named peril to investors and pays the sponsor after a trigger event. A rebate-style resilience bond does the same, then adds a channel that returns money to the sponsor for risk reduction when losses stay low. The risk transfer is unchanged.
does a resilience bond fund nature?
Sometimes, and only if someone picks it. Use-of-proceeds bonds can include ecological projects, and the Forest Resilience Bond funds forest restoration directly. The best-known rebate bond funds roofs. The landscape functions that set the size of most losses — floodplains, marshes, dunes, stands — usually have no sponsor.
what do you actually hold?
With a cat bond or resilience bond, you hold a note: collateral at risk against a peril, a coupon, a term. With an ensurance certificate, you hold a position tied to one named place that funds its present condition. It is not a 144A note and not an insurance policy; there is no coupon and no principal at risk against a peril. More here.