Financing the Synergy

Executive Summary

The case for combining adaptation and mitigation is settled. What is not settled is who writes the cheque for the half of the asset that earns nothing.

An earlier report in this series argued that adaptation and mitigation are inputs to each other in physical terms — protection runs on electricity, and clean supply must itself be protected. That argument is taken as given here. This report is about what happens when a project built on it reaches a financing committee.

The answer is visible in the aggregates. Of the tracked adaptation finance recorded for 2023 — in the region of US$63–65 billion depending on the tracker's cut — only about US$5.7 billion, roughly nine per cent, came from private actors. Mitigation, by contrast, is financed overwhelmingly by commercial capital. That gap is not a failure of conviction. It is what happens when one category of investment produces a revenue line and the other produces a counterfactual.

Section 1 sets out the cash-flow asymmetry precisely. Section 2 shows how compounding risk prices itself into the sovereign's whole debt stock, creating a fiscal case for resilience that no project appraisal captures. Section 3 examines the five instruments that convert avoided losses into cash flows, with the transactions that prove each. Section 4 gives the operative rule — finance the joint asset once — and Section 5 asks what the new finance goal changes. Sections 6 and 7 give recommendations and conclusions.

1. The Shape of the Cash Flow

Finance does not sort investments by their purpose. It sorts them by what they pay, to whom, and when — and on that test adaptation and mitigation are different asset classes.

1.1 Three shapes

Climate investments come in three cash-flow shapes, and the shape determines the instrument:

• Revenue-generating. The asset sells something. A solar farm sells electricity under a contract; a transmission line earns a regulated return. There is a receivable, it can be assigned, and the asset can be project-financed at a leverage and tenor set by the strength of the offtake.
• Cost-avoiding. The asset reduces an expense that was actually being paid. An efficient chiller cuts the electricity bill; a solar pump removes the diesel bill. There is no receivable, but there is a verifiable saving against a historical baseline, which is enough to support energy-service contracts and on-bill repayment.
• Loss-avoiding. The asset reduces damage from an event that may not occur. A raised substation, a drainage channel, a cyclone-rated roof. There is no receivable and no historical baseline — only a modelled counterfactual — and this is the shape that finance cannot hold.

Most mitigation sits in the first two shapes. Most adaptation sits in the third. That is the entire asymmetry, and it explains the aggregate.

1.2 Why splitting the project is the default

Consider a port authority in a cyclone-exposed country planning a terminal upgrade: electrified cranes and a solar canopy on one side of the drawing, raised quay levels and a stormwater system on the other. The first set has a tariff behind it and will attract commercial lenders. The second has a benefit-cost ratio in the appraisal and no payer.

What happens next is institutional rather than ideological. The commercial half goes to a bank; the protective half goes into a queue for concessional finance or a national budget line. The two proceed on different timetables, and the protective half — being the one with no revenue and no deadline imposed by an offtaker — is the one that slips. The asset is built exposed, and the resilience is scheduled for a later phase that often does not come.

2. Compounding Risk and the Vulnerability Premium

Physical risk does not only raise the cost of the project exposed to it. It raises the price of every dollar the sovereign borrows — which is where the fiscal case for resilience actually lives.

Research for UN Environment by Buhr and colleagues at Imperial College London and SOAS found that climate vulnerability was already embedded in sovereign borrowing costs: for every ten dollars of interest a sample of developing countries paid, roughly one dollar was attributable to climate vulnerability. Over the decade studied that amounted to about US$40 billion in additional interest payments on government debt alone, and the authors projected the next decade's additional cost at US$146 billion to US$168 billion.

Three things follow. First, the premium is charged on the whole debt stock, not on the exposed project, so the return to reducing it is proportional to how much a country borrows rather than to the size of any individual investment. Second, it compounds with the other capital-cost penalties that vulnerable economies already face, which an earlier report documented on the corporate side. Third — and this is the part that appraisal systems miss — a resilience programme large enough to move a rating or a risk assessment generates a saving that appears in the interest line of the budget, not in the project's own cash flows.

The same logic runs through the insurance market, where the price of risk is quoted daily. Of US$318 billion in economic losses from natural catastrophes in 2024, 57 per cent were uninsured, leaving a protection gap of US$181 billion (Swiss Re Institute). Where cover is absent, the loss is not avoided by anyone; it is retained by households and governments, and it is financed after the event at emergency prices.

3. Five Instruments That Price an Avoided Loss

Each of these converts a counterfactual into something with a date and an amount. All five have live transactions behind them.

3.1 Spread compression: resilience as a balance-sheet investment

If vulnerability is priced into the sovereign spread, then reducing vulnerability is a debt-management operation. The saving is real, appears in the interest line, and accrues to the finance ministry — which is also the entity that can borrow against it. The practical requirement is measurement: rating agencies and investors need a credible account of what a resilience programme changes, which means published hazard exposure, asset registries and completion data rather than a plan.

3.2 Climate-resilient debt clauses: turning a disaster into liquidity

A climate-resilient debt clause allows a borrower to defer principal — and in some designs interest — when a defined disaster occurs, converting the shock into breathing room rather than a default. The clause costs the creditor timing, not principal, which is why it has spread quickly. Grenada, Barbados and the Bahamas have carried such clauses in bonds; the World Bank introduced pilot clauses for small states allowing deferral for up to two years, the IDB offers principal deferral to twelve eligible countries, and the EIB adopted clauses in 2024, with other development banks considering them. At the 2023 Paris Summit, 73 countries called for creditors to introduce them (CPI primer for Finance in Common).

3.3 Debt conversions: buying the fiscal space and earmarking it

A debt conversion refinances expensive debt behind a credit enhancement and commits the savings to a specified purpose. Two transactions define the current state of the art.

Barbados, November 2024. The government refinanced US$293.3 million of public debt behind US$300 million of guarantees — US$150 million each from the IDB and the EIB, the first joint guarantee by the two institutions in the Caribbean — generating about US$125 million in fiscal savings contractually directed to water and sewerage infrastructure. It is the first conversion structured explicitly for climate resilience rather than conservation.

Ecuador, May 2023. The Galápagos operation repurchased US$1.6 billion of bonds at roughly 40 cents on the dollar, saving an estimated US$1.1 billion in debt service over seventeen years and committing US$450 million to conservation. The replacement bond carried an US$85 million IDB guarantee and US$656 million of political risk insurance from the US Development Finance Corporation, which is what allowed it to price far above Ecuador's own credit.

Both structures share a mechanism worth naming: a guarantee converts an institution's credit rating into fiscal space for the borrower, and the conversion agreement then earmarks that space. The protective investment is paid for out of a spread differential rather than a revenue stream — the closest thing in current practice to financing an avoided loss directly.

3.4 Insurance and parametric cover: the market's own price for the counterfactual

Insurance is the one instrument that puts an explicit price on an avoided loss every year. Its second function is more useful than its first for this argument: a premium that falls after a resilience investment is a market valuation of that investment, produced by an entity with money at stake. Parametric structures, which pay on a measured trigger rather than an assessed loss, make that valuation faster and cheaper to run in places with thin claims infrastructure. The obstacle is that the protection gap is widest precisely where the premium is least affordable, which is why sovereign risk pools and premium subsidy are part of this instrument rather than an alternative to it.

3.5 Cross-subsidy from the mitigation half

The last instrument uses the bankable half to pay for the unbankable one. The Paris Agreement crediting mechanism does this by rule: a five per cent share of proceeds from issued credits goes to the Adaptation Fund, more than double the two per cent levied under the Kyoto-era mechanism. The same principle can be applied at project level — a wind farm's revenues funding the substation flood defence that keeps it connected — and it has the advantage of being invisible to the credit committee, because the combined asset is financed on the revenue it produces.

Table 1 — Instruments by the cash-flow shape they address

The mapping is the author's, applied to the transactions and rules cited in this section. Nothing here is investment advice.

4. Finance It Once

The single most consequential structuring decision is whether the joint asset is presented to the market as one financing or two.

The rule this report proposes is simple to state and unusual in practice: finance the combined asset once, against the revenue-bearing half, and write the resilience specification into the loan documentation as a covenant rather than pursuing it as a separate project.

Three things follow from doing it that way.

The protective half inherits the timetable of the commercial half. A covenant is a condition of disbursement; a parallel grant application is a hope. This alone addresses the most common failure mode, in which the revenue asset is commissioned on schedule and the protection is deferred to a phase that is never funded.

The lender's own interest is engaged. A financier with exposure to a wind farm's cash flows has a direct stake in the substation staying dry, because an asset that stops producing stops servicing debt. Resilience specified as a covenant is not a concession extracted from the lender; it is protection of the lender's collateral, and it should be priced as such.

The blended element goes where it is needed. Concessional funds are scarce and are most often used to reduce the cost of the whole package. Used instead to pay for the protective component inside a commercially financed asset, the same money buys the thing the market will not buy, and leaves the market to fund what it already funds happily.

5. What the New Goal Changes

The headline number at COP29 was a tripling. The structural question it left open is the one this report is about.

The new collective quantified goal agreed at Baku sets a target for developed countries to mobilise at least US$300 billion a year for developing countries by 2035, replacing the US$100 billion goal, with a broader call on all actors to work towards US$1.3 trillion a year by the same date and a Baku-to-Belém roadmap to identify how.

The gap between the two numbers is the whole of the problem this report describes. The US$300 billion is largely public and mobilised; the additional trillion is expected to come from private capital. But private capital finances cash flows, and the majority of what the additional trillion is supposed to buy in vulnerable countries is loss-avoiding. Unless the instruments in Section 3 are scaled deliberately — spreads that respond to resilience, clauses that are standard rather than pilot, conversions that are repeatable rather than bespoke, insurance that is affordable at the point of greatest gap — the larger figure will be met, if it is met at all, overwhelmingly on the mitigation side.

That has a direct convergence consequence, which is why this report sits in a series about it. A country that finances its mitigation and defers its adaptation is buying an asset base that produces revenue until the year it is damaged. Sustained convergence requires both halves, and the financing architecture currently supplies one.

6. Recommendations

Ordered by who can act without waiting for anyone else.

6.1 Finance ministries and debt managers

• Treat resilience as debt management. Quantify the vulnerability component of the sovereign spread, publish the hazard and asset data that would let it be reassessed, and present major resilience programmes to investors and rating agencies as interest-cost reduction.
• Insist on climate-resilient debt clauses in every new instrument, official and commercial. They cost creditors timing rather than principal, which is why the ask succeeds.
• Build a standing conversion capability rather than treating each swap as a one-off; the arrangement cost is what makes these transactions rare.

6.2 Multilateral development banks and guarantors

• Price guarantees as the core product, not an accessory. Both landmark conversions above ran on guarantees and political risk insurance; the guarantee is the mechanism that turns an institution's balance sheet into a borrower's fiscal space.
• Standardise the conversion structure so that a transaction which took a year of bespoke legal work becomes a template with a term sheet.
• Direct concessional money at the unbankable component inside otherwise commercial projects, rather than at softening the whole package.

6.3 Commercial lenders and investors

• Underwrite the resilience of your own collateral. Physical protection of a revenue asset is credit enhancement; it belongs in covenants and in pricing, not in a sustainability annex.
• Accept one financing rather than two where a joint asset exists, and set disbursement conditions on the protective works.

6.4 Insurers and risk pools

• Publish what resilience is worth. A premium reduction quantified for a specific measure is the most credible price signal available for adaptation, and it is currently invisible to the public sector.
• Extend parametric cover where claims infrastructure is thin, and pair it with premium support in the jurisdictions where the protection gap is widest.

7. Conclusions

The integration argument has been won. The structuring argument has not been had.

Adaptation and mitigation belong in the same investment, and most institutions now say so. What follows that agreement is a financing committee, and there the two halves of the asset stop being complements and become different products: one with a receivable and one with a counterfactual. Private capital supplies about nine per cent of tracked adaptation finance, against a mitigation total that is overwhelmingly commercial, and that ratio is a fact about cash-flow shape rather than about conviction.

Compounding risk raises the stakes because it prices itself into everything a vulnerable sovereign borrows — roughly one interest dollar in ten on the sample studied, tens of billions over a decade and a projected US$146 to US$168 billion over the next. That is simultaneously the cost of doing nothing and the return available from doing something, and it accrues to the treasury rather than to any project.

Five instruments already convert an avoided loss into a cash flow: spread compression, resilient debt clauses, debt conversions, insurance, and cross-subsidy from the revenue half. Four of them require a public balance sheet, which is the honest explanation for why adaptation finance remains public. The fifth — financing the joint asset once, on the revenue it earns, with protection written in as a covenant — is the one available to a project team on a Tuesday, and it is the one this report would put first.

The new goal makes the question urgent rather than academic. Three hundred billion dollars a year of largely public money by 2035, and a call for a trillion more from capital that finances receivables: unless the instruments that price avoided losses are scaled deliberately, the trillion will arrive on the side of the ledger that was never short of financiers.

References

Every quantitative claim above is attributed inline. The principal sources are collected here.

• Climate Policy InitiativeGlobal Landscape of Climate Finance 2025 — adaptation finance of about US$65 billion in 2023 on the headline cut (US$63 billion on the adaptation-only series), of which roughly US$5.7 billion private. See also Tracking and Mobilizing Private Sector Climate Adaptation Finance on why the private figure is both small and under-observed.
• Buhr, Volz, Donovan, Kling, Lo, Murinde & PullinClimate Change and the Cost of Capital in Developing Countries (Imperial College London, SOAS and UN Environment, 2018) — one interest dollar in ten attributable to climate vulnerability; about US$40 billion of additional interest over the decade studied and a projected US$146–168 billion over the next. Summary at SOAS.
• Swiss Re Institutesigma 1/2025: Natural catastrophes — US$318 billion of economic losses in 2024, 57 per cent uninsured, a protection gap of US$181 billion.
• Climate Policy InitiativeClimate-Resilient Debt Clauses: A Primer for FiCS Members (2025) — adoption by the World Bank, IDB and EIB, use in Grenada, Barbados and the Bahamas, and the 73-country call from the 2023 Paris Summit.
• Inter-American Development BankBarbados launched the world's first debt-for-climate-resilience operation, and the European Investment Bank release — US$293.3 million refinanced, US$300 million of joint guarantees, about US$125 million released for water and sewerage.
• Ecuador Galápagos conversionReported structure and terms — US$1.6 billion of bonds repurchased at about 40 cents, US$1.1 billion of debt service saved over seventeen years, US$450 million for conservation, with an US$85 million IDB guarantee and US$656 million of DFC political risk insurance. See Galápagos Life Fund and Climate Home News; a critical assessment of such structures is in the Yale Journal of International Affairs.
• UNFCCCAdaptation Fund and the Article 6.4 arrangements — a five per cent share of proceeds from issued credits, against two per cent under the Kyoto-era mechanism.
• UNFCCCCOP29 agrees to triple finance to developing countries — US$300 billion a year by 2035 and the call to work towards US$1.3 trillion. Explanation of the two figures at WRI.
• H HeuristicsRelated reports in this series: integrating adaptation and mitigation, whose physical argument this report assumes; climate-resilient industrialization pathways, on the corporate cost of capital; and the economics of prevention, on why protective spending loses budget arguments.