Finance
The covenant carbon never had
There is 6.2 trillion dollars of labelled sustainable debt governed by covenants, ratchets and performance targets. Almost none of it links a single payment to the delivered, monitored performance of the carbon credits a company actually buys. That is the gap, and project finance already built the tools to close it.
Sustainable finance has spent a decade building elaborate machinery to tie money to environmental performance, and it has tied it to the wrong thing. Cumulative labelled sustainable debt reached roughly 6.2 trillion dollars by December 2024. Green bonds outstanding passed three trillion in the third quarter of 2025, with annual issuance around one trillion. Sustainability-linked loans ran near 418 billion dollars in 2025, sustainability-linked bonds around 35 billion. Every one of these instruments comes with covenants, reporting obligations, and in the linked variants a financial consequence (a coupon step-up, most commonly 25 basis points, or a margin ratchet) triggered by performance against a target. The infrastructure for binding capital to environmental outcomes is mature, standardised and large.
And it all points inward. Sustainability-linked instruments, governed by the ICMA Sustainability-Linked Bond Principles and the LMA loan principles, link the borrower's coupon or margin to the borrower's own emissions KPIs measured against its own Sustainability Performance Targets. That is a fine thing to do. But it means the entire 6.2 trillion dollar apparatus is silent on the carbon credits a company buys. Nothing in this market links a payment to the delivered, monitored performance of the underlying credit or project. There is no covenant on the tonne.
The shape of the gap
Consider what a corporate carbon purchase actually is, viewed as a financing professional would view any forward commitment. A buyer agrees today to pay for environmental performance that will be delivered over years, sometimes over a decade or more. The forward market makes this concrete: announced forward offtakes in 2025 ran to somewhere between roughly 12.3 and 13.7 billion dollars (Sylvera, WEF), structured to deliver around twelve million credits a year through 2035 at a weighted average near 180 dollars. Abatable counts around 5.8 billion in forward agreements and 15.8 billion in project financing. The forward book, not the spot market, is where the risk and the money concentrate.
Now ask the questions a structured-finance desk asks of any forward delivery. What happens if the tonnes are not delivered? What happens if they are delivered but degraded (a reversal, a permanence failure, a reclassification that strips the vintage of eligibility)? Is payment staged against verified delivery, or paid up front against a promise? Is there a holdback? An escrow? A buffer-release mechanism tied to independent monitoring? A delivery bond standing behind the developer's obligation? In project finance these questions have standard answers and standard instruments. In carbon, in the overwhelming majority of cases, they have none. The buyer wires money for a forward promise and carries the entire delivery and quality risk with no contractual machinery to manage it.
The asymmetry is worth dwelling on, because it is the kind of thing that reads as obvious once stated and is nonetheless tolerated at scale. A treasury team that would never accept an unsecured ten-year forward on any other commodity, with no staged payment, no performance bond and no independent measurement of delivery, accepts exactly that on carbon as a matter of routine. The reason is partly cultural (carbon purchasing grew out of sustainability functions rather than treasury) and partly that the enabling machinery simply was not on the shelf. Neither reason survives contact with the new accounting. Once a forward position has to be impairment-tested and presented gross against its obligation, the absence of delivery protection stops being a procurement detail and becomes a balance-sheet exposure the auditor will ask about.
The buyer wires money for a forward promise and carries the entire delivery and quality risk with no contractual machinery to manage it.
The instruments that should exist
The striking thing is how little invention is required. The tools to close this gap were built decades ago for project finance, infrastructure and trade finance. They simply have not been imported into carbon. Three are obvious and overdue.
01 · The covenant-linked forward offtake
A forward offtake is already a contract. What it lacks is enforceable performance covenants wired to independent monitoring. Import the discipline directly: stage payment against verified delivery rather than paying up front; tie a holdback to monitored permanence over a defined tail; release buffer reserves only on independent confirmation that the underlying tonnes remain stored. The structural features of a sustainability-linked bond (the step-up on a missed target, the ratchet on performance) already exist as a template. The only change is to point them at the delivered performance of the credit rather than the borrower's own inward KPI. The same legal architecture, aimed at the right object.
02 · The carbon delivery bond
Performance bonds are standard in construction and infrastructure: a third party stands behind a contractor's obligation to deliver, and pays out if delivery fails. A carbon delivery bond is the same instrument. It guarantees that contracted tonnes will be delivered to specification, with payout triggered by independent monitoring rather than by the developer's own attestation. This converts an unsecured forward promise into a secured one and gives the buyer's auditor something concrete to point to when testing the recoverable value of a forward position.
03 · Holdback and escrow tied to monitoring
The simplest and most immediately deployable mechanism. A portion of the purchase price is held back or placed in escrow and released in tranches only as independent monitoring confirms the underlying performance over time. This aligns cashflow with delivered reality, it gives the developer a continuing incentive to maintain the project rather than bank the cash and move on, and it directly addresses the central criticism of the existing sustainability-linked market: that penalties are too weak to change behaviour. The average sustainability-linked bond step-up is around 31 basis points, under twelve per cent of the coupon, widely judged inadequate to alter conduct. A monitoring-triggered holdback is not a small coupon adjustment. It is the principal itself, conditioned on performance.
Why monitoring is the enforceable trigger
Every one of these instruments has the same dependency, and it is the dependency the existing market has never solved. A covenant is only as good as the trigger that fires it, and a performance trigger is only as good as the independent measurement behind it. The criticism that has dogged sustainability-linked debt (weak and non-material KPIs, unambitious targets, low penalties) is at root a measurement and verification failure. If nobody can independently and continuously establish whether the performance condition has been met, the covenant is decorative.
This is exactly where carbon is better positioned than the existing labelled-debt market, not worse. The underlying performance of a carbon project (whether the tonnes are being delivered, whether they remain stored, whether a reversal has occurred) is continuously observable in a way a borrower's diffuse emissions trajectory often is not. The trigger can be real, continuous and evidence-based rather than annual and self-reported. The missing piece has never been the legal structure, which project finance long ago perfected. The missing piece has been an independent, continuous, finance-grade read on the underlying performance that a covenant can actually fire against. Build that read, and the covenants become enforceable. Without it, they remain decorative, exactly as too many sustainability-linked structures already are.
A covenant is only as good as the trigger that fires it, and a performance trigger is only as good as the independent measurement behind it.
Insurers are partners, not competitors
There is a parallel effort already underway, and it is important to read it correctly. The carbon insurance market is growing from roughly 1.8 billion dollars in 2025 toward roughly 6.2 billion by 2034, a compound rate near 14.7 per cent, with Kita, Oka, CarbonPool, Howden, Munich Re, Chubb and Aon all active. Kita writes non-payment and buffer insurance; CarbonPool holds a carbon balance sheet and offers in-kind cover. These players are approaching the same problem (delivery and quality risk on carbon purchases) from the risk-transfer side.
They are partners to a covenant-and-structuring layer, not competitors to it, and the division of labour is the natural one from any structured transaction. The insurer holds the capital and prices the cover. What the insurer needs in order to price it is an independent, continuous read on the underlying risk: the same read a covenant needs in order to fire. The structuring brain that designs the covenant and the holdback, and the monitoring layer that supplies the trigger, can be capital-light and independent precisely because the insurer carries the balance sheet. The relationship mirrors the one between an independent risk model and a reinsurer, or between a rating and a guarantor. One party reads and structures the risk. The other party holds it. They are stronger together, and neither displaces the other.
The prize, in finance terms
Put the pieces together and the conclusion is uncomfortable for how the market currently operates. There is 6.2 trillion dollars of labelled debt disciplined by covenants, and a forward carbon book in the tens of billions disciplined by almost nothing. The instruments to close that gap (covenant-linked offtakes, delivery bonds, monitoring-tied holdbacks) are not novel. They are the standard apparatus of project finance, performance bonds and escrow, waiting to be pointed at the right object. The one ingredient that has never existed is an independent, continuous, finance-grade read on the underlying credit performance to serve as the enforceable trigger and the pricing input. Supply that, and carbon purchases stop being unsecured forward promises and start being structured, covenanted, monitored financial commitments, the way every other forward of this size already is.
Kyroq supplies the independent, continuous risk read that makes a carbon covenant enforceable and a delivery instrument priceable, and structures the holdbacks, buffer releases and covenant terms that wire monitored performance into the forward offtake. Insurers and banks hold the capital; Kyroq is the independent risk and record layer, capital-light, never taking title, always paid by the buyer. The covenant carbon never had is finally buildable, because the trigger behind it can now be real.