your liabilities are 30 years out. so is a forest.

A 30-year gilt is not a 30-year asset.

At a 5% coupon and a 5% yield, its Macaulay duration is about 16 years. If you run a closed defined-benefit book, your liability duration probably sits somewhere between 15 and 25. The paper that looks like a perfect match on the maturity ladder can leave you a third short on the number that actually moves your funded status.

Every liability-driven investing desk knows this. It is worth restating anyway, because the same confusion is about to be made about ecological assets — and this piece is about whether a forest, a wetland, or a restored river reach can ever be a liability-matching claim. The honest answer has three parts: the duration math is real, the instrument mostly does not exist yet, and the gap between those two facts is narrower than you would guess.

duration is not maturity

Maturity is the date of the last cash flow. Macaulay duration is the weighted-average time to all of them. High coupons pull duration in. Zeros push it out to the maturity date. Level payment streams front-load present value, which shortens them a lot more than most term sheets admit.

At a flat 5%:

That last row is the one worth memorizing. A perpetual claim at a 5% yield has a Macaulay duration of about 21 years — finite, even though it never matures, because the duration of a perpetuity is (1+y)/y. At 6% it is about 17.7 years. Forever is shorter than it sounds.

The level-annuity row uses the same arithmetic, and it is the row that matters for anything sold as a multi-decade payment path:

D = (1+r)/r − N / ((1+r)^N − 1)

So when a counterparty tells you an instrument has a 25-year term, they have told you almost nothing about its duration.

liability-driven investing inverts the definition of risk

LDI builds the portfolio to track the liabilities, not a market benchmark. That single move rewires every risk label on the desk.

In an asset-liability frame, cash is the risky position: it has roughly zero duration against a liability that has twenty. This is why the same instrument is "too volatile" for a retail bond fund and "the hedge" for a closed scheme. Anything pitched into a matching portfolio has to lead with surplus and funded-status stability, not with a Sharpe ratio.

And it is not one number. Real hedges are built across key-rate buckets — sensitivity at the 10-year, 20-year, and 30-year points separately — because a liability curve has a shape, not a scalar. Any single long claim contributes to one or two buckets at best. Keep that in mind before anyone describes an ecological sleeve as a hedge. It is a bucket contribution.

the two ways the long end runs out of road

Physical scarcity. There is chronically less 30-to-50-year-plus paper than the pension and insurer books that want it. Century issuance is episodic rather than structural — sovereigns, universities, charities, utilities, and in 2026 Alphabet's sterling and dollar century bonds reviving a 1990s corporate precedent. Buying duration at the long end means competing for a thin float against everyone else with the same problem.

Synthetic completion. So desks complete duration with interest-rate swaps, gilt repo, and STRIPS. It is capital-efficient and entirely standard. It is also leveraged and collateralized daily, and that combination has a specific failure mode.

After the UK fiscal statement of 23 September 2022, long gilt yields spiked. Leveraged LDI hedges — repo and swap positions used to buy duration without buying the whole bond — faced collateral calls. Schemes could not post cash fast enough. Managers sold gilts to raise it. The selling pushed yields higher, which triggered more calls. The Bank of England intervened on 28 September with a temporary long-dated gilt purchase facility announced at up to £65 billion, though actual purchases came in far below the envelope. Regulators then pushed schemes toward much larger collateral buffers; headroom for a yield move of roughly 250 basis points is now the common reference point.

The lesson is not "don't hedge." It is narrower and more durable: when your duration comes from leverage, stress liquidity is part of the product, not an externality. That is the bar any new claim calling itself duration-matched should be held to — including ours.

a wetland is also a long-duration engine

Here is the turn, and it is a cash-flow argument rather than an ecological one.

Ecological systems run on clocks that look a great deal like infrastructure clocks. Hydrologic function in a restored reach can respond within a few seasons. Soil carbon, riparian structure, and stand composition run decades. Structural maturity in a temperate forest runs a century or more. And a functioning watershed has no maturity date at all — it keeps producing water regulation, flood attenuation, and habitat until something breaks it.

That is a long-duration production profile. It has been financed for a long time in one narrow form: timberland and farmland institutionalized biological hold periods, but only where extraction monetizes the biology. The harvest is the coupon. Everything else nature produces has been unpriced, which is why the duration was never visible to a matching portfolio.

Ecological duration is the discipline of terming a claim to the clock of the place rather than to the fund life of the vehicle. It requires two clocks to clear at once:

A claim that matches the liability clock but fails the ecological clock is greenwash with a coupon. A restoration that matches the ecology but offers no duration-clear claim never clears the investment committee. Both clocks are underwriting inputs, and most instruments in this market so far have optimized exactly one of them.

the ladder, and its honest duration

Ensurance prices protection as a rate card: several duration slices against a single asset, the way an issuer sells 5s, 15s, and 25s against one project. Below is the ladder on a real underwritten asset — 83 acres of mixed hardwood swamp in the southeastern US, carried at a protection cost of $267,500 against roughly $1.45 million a year in measured ecosystem-service flows.

The investor position here is the cost side: front the cost of bringing the asset into protection, receive the serviced payment stream. The term tiers price that stream at a 10% target yield; the perpetual tier is struck off a 5% baseline.

Read the right-hand column again, because it contains the problem rather than the pitch. The tier that terminates in permanence has the shortest duration. The tier with liability-matching duration is the one that never terminates.

That is not a marketing wrinkle, it is arithmetic. Level payments front-load present value, so a 25-year path carries roughly 10.5 years of duration at a 5% discount rate and about 8.5 at 10%. A 25-year ENTRUST path is, in duration terms, an intermediate-credit instrument wearing a long label. If you want 18 years of duration out of an ecological claim, the schedule has to be built for it — deferred start, escalating payments, or a ladder that pairs a perpetual ENSURED position with terminating ENTRUST paths. Term is marketing. The payment schedule is the duration.

Two things this claim has that a gilt does not:

A condition feedback loop. The stream is underwritten on RealValue — ecosystem-service flow value scaled by measured condition — so underperformance can reduce payment. A gilt does not care whether the state is functioning. This claim is contractually tied to whether the thing works.

An end state. ENTRUST retires the claim: the asset graduates to permanent commons, zero residual, no lien, nothing to refinance. Sovereign debt rolls forever; this terminates. For a matching portfolio that cuts both ways — you are not exposed to perpetual reissuance risk, and you have to model the runoff rather than assume a rollover.

One thing it does not have: a curve. No ecological yield curve, no rating, no index, no key-rate bucket. More on that below.

availability payments are the closest traditional twin

If you already underwrite core infrastructure, you have priced this cash-flow shape before.

An availability-payment concession pays the private partner for the asset being available and meeting performance standards — not for usage. Payment is for readiness and quality. Deductions apply when standards slip. Contracts typically run 25 to 35 years. The private partner carries construction and operating risk; the authority keeps demand risk. That predictability is exactly what makes the asset financeable.

An ensurance premium is an availability payment for ecological function: paid for the system being available and at standard, not for anyone's use of it. The payors are the parties who depend on the function — a water utility whose treatment costs move with watershed condition, a city carrying flood exposure, a data center underwriting its own water balance, an insurer with concentrated wildfire loss. Payment is sized against the avoided cost of the gray alternative and reduced when measured condition slips.

This is not hypothetical structuring. The Forest Resilience Bond is already, structurally, a natural-infrastructure concession: an SPV channels capital, an implementation partner performs the restoration, and beneficiaries pay against value received. What is missing is scale, standardization, and tenor — not the mechanism.

the constraint, stated plainly

What "manufactured" means, item by item:

1. The payor. Availability payments work because a government covenants to pay. Ecological premiums work when a utility, city, insurer, or dependent corporate signs a multi-year contract. Those payors exist and some are paying. What does not yet exist is a deep, diversified, ratable book of them.
2. The schedule. See the table above. Most current terms are level annuities, which means most current claims are 3-to-10-year duration instruments carrying 25-year labels. Longer duration requires deferred or escalating structures that nobody has priced at scale.
3. The residual. Zero at entrust. That is a feature for the commons and a modeling requirement for you: there is no terminal value, so the entire return sits in the payment stream. Familiar next to amortizing infrastructure debt. Unfamiliar next to a lease or a ground lease, where the residual does much of the work.
4. The regulatory door. Post-Solvency II reforms in the UK opened a limited allowance for matching-adjustment portfolios to hold assets with "highly predictable" rather than strictly fixed cash flows. A condition-linked ecological premium is precisely the kind of asset that would have to walk through that door, and no one has walked it yet. In a pension, the realistic first home is an existing real-assets, natural-capital, or private-credit line — not a new asset-class vote.
5. The credit analog. Natural cap rate, condition score, and evidence have to become as skimmable as BBB. Without a rating and an index, duration clarity still does not get you a mandate line.

Two claims we are deliberately not making.

We are not claiming low volatility. Illiquid, infrequently marked assets report smooth returns, and smooth reported returns are not the same thing as low risk — Cliff Asness's "volatility laundering" critique of private markets applies here as much as anywhere. And this is the horizon sleeve, not the tipping-point risk sleeve; repricing nonlinear nature risk inside your existing book is a different piece of work with a different mandate and a different answer. What is genuinely on offer is duration and the purpose of the capital, not a better Sharpe ratio.

one pension has already taken the first step

West Yorkshire Pension Fund — roughly £20 billion in assets, part of the £60 billion-plus Northern LGPS pool — took a 25% equity stake in the UK natural-capital manager Rebalance Earth and committed £25 million as cornerstone investor to its RE(DARWIN) portfolio. The vehicle targets £150 million and a stated 8–12% IRR over 15 years, underwriting peatlands, rivers, wetlands, and oyster reefs against contracted ecosystem-service payments.

Read that precisely, because the precise version is more useful than the headline. It is a 15-year fund IRR in an alternatives allocation, not a duration-matched claim in a matching portfolio. What it proves is narrower: place-based natural infrastructure with contracted cash flows can clear a real investment committee, occupy a real mandate line, and be sponsored by a pension as an owner rather than merely an LP. That is the harder gate, and it is already open. The duration engineering comes after.

five questions to ask before underwriting an ecological duration claim

1. What is the Macaulay duration of the payment stream — not the term? Ask for the number computed at your own discount rate, and ask for the key-rate contribution if they have it.
2. Who is the payor, what is their credit, and what is the remedy if they stop? Named and contracted, or it is a grant with better branding.
3. What triggers a deduction, who measures it, and how often? Condition feedback is a feature only if it is independently verifiable.
4. What is the end state and the residual? If the claim terminates, model the runoff. If it rolls, say so and price the reissuance.
5. Which bucket does this sit in today — real assets, private credit, natural capital, infrastructure — and what would it take to get a line without an asset-class vote?

If a counterparty cannot answer 1 and 2 in writing, everything else is decoration.

frequently asked questions

what is liability-driven investing?

Liability-driven investing (LDI) constructs a portfolio to track a scheme's liabilities rather than a market benchmark. Risk is measured as surplus and funded-status volatility instead of tracking error, which inverts the usual labels: cash becomes the risky position, and long-dated bonds become the hedge.

does duration equal maturity?

No. Maturity is the date of the final cash flow; Macaulay duration is the weighted-average time to all of them. A 30-year par bond at 5% has about 16 years of duration, a 30-year zero has 30, and a perpetual claim at 5% has about 21 — computed as (1+y)/y.

can a pension hold protected nature as a duration hedge today?

Not as a hedge. What exists today are contracted-cash-flow vehicles sitting in alternatives and real-asset sleeves, with payment schedules mostly shorter than liability duration. Duration-matched ecological claims are being designed, not traded.

why is a long payment stream shorter than its term?

Because level payments front-load present value. A 25-year level stream carries roughly 10.5 years of Macaulay duration at a 5% discount rate. Extending duration requires reshaping the schedule — deferring the start or escalating the payments — not extending the label.

what to do next

If you want to see how a claim is structured against a single named natural asset — term, premium, condition, end state — start with specific ensurance, where certificates are issued against individual places rather than a pooled fund. The protocol view of how those payments route is at ensurance.

Read next: who actually holds nature for decades for the patient-capital pillar behind this post, how to underwrite natural infrastructure like an asset class for the credit-analog work, and the investment value of ecosystem condition for how RealValue turns measured condition into the number the premium is priced off.