Markets strand assets when demand changes, not when they expire

An asset does not strand because it stops working. It strands because the buyers who used to want it no longer do. The coal plant still burns; the demand for coal power moved. The same logic now applies to a large slice of the corporate carbon book. On 11 June 2026 the Science Based Targets initiative finalised version 2.0 of its Corporate Net-Zero Standard, and in one move it changed what counts. The credits did not expire. The rule that gave them value did.

More than 10,000 companies hold validated science-based targets, and over 12,600 carry targets or commitments, representing more than 40 per cent of global market capitalisation across 90-plus countries. Many of them bought avoidance and REDD+ credits in good faith to help close the gap to their targets. SBTi V2.0, effective 31 January 2027 with a transition running to January 2028, tells them those credits no longer count toward the target they were bought to meet. That is a demand shock, delivered by standard.

What V2.0 changes

01 · Credits are out of target accounting

Under V2.0, carbon credits cannot count toward Scope 1, 2 or 3 target implementation, and they cannot be netted from a company's emissions inventory. They are relegated to contribution claims: a company may say it has funded climate action beyond its value chain, but it may not say that funding reduced its own footprint against its target. The accounting role the credits were bought to play has been removed.

This is the crucial distinction for anyone holding a book. A credit that contributes is not worthless, but it is worth something quite different from a credit that counts. The market priced these instruments on the assumption they could be used against a target. Strip out that use case and the price has to find a new, lower level, set by the much softer demand for voluntary contribution.

02 · The obligation moves to durable removals

In place of credits-against-target, V2.0 builds future neutralisation on durable removals: the long-lived, high-permanence removal of carbon, scaling from a minimum of 1 per cent of residual emissions in 2035 to 100 per cent by a company's net-zero year. SBTi adds tiered recognition, Engaged, Advanced and Leadership, to reward companies that fund removals along the way. The direction is unambiguous. The cheap, abundant, near-dated credit is out. The expensive, scarce, long-dated removal is in.

The cheap, abundant, near-dated credit is out. The expensive, scarce, long-dated removal is in.

Durable removals are not a like-for-like replacement for the avoidance book. They are a different asset class with a different cost curve and a different risk profile: longer delivery horizons, permanence obligations measured in decades, and forward offtake structures rather than spot purchases. The forward market is already where the money concentrates. Announced offtakes in 2025 ran to roughly 12 to 14 billion US dollars, set to deliver around 12 million credits a year through 2035 at a weighted average near 180 US dollars, against a spot market whose weighted average price was about 6.10 US dollars. The two markets are an order of magnitude apart on price and on time.

Two exposures, not one

02 · The legacy book

The first exposure is the legacy book already on the balance sheet. These are the avoidance and REDD+ credits bought to count toward a target that, from January 2027, will no longer accept them. Their use case has been withdrawn. Under the new FASB standard, ASU 2026-02, these same voluntary credits are carried at cost less impairment and tested at each reporting date. A demand shock that lowers their realisable value is exactly the kind of event that drives an impairment, and that impairment is irreversible.

So the legacy book is not a dormant asset quietly waiting to be retired. It is a live impairment exposure, repriced by a standards change, that flows through earnings on the accounting clock. The company that bought heavily into avoidance to hit an interim target now holds a position that the standard-setter has demoted and the accounting framework forces it to mark down.

03 · The forward removal programme

The second exposure is the forward removal programme the company must now build. Meeting the durable-removal pathway means committing, often years ahead, to long-dated offtakes at prices many multiples of historic credit costs, for delivery of a tonne that has not yet been removed and whose permanence must hold for decades. This is a procurement and delivery-risk problem dressed as a sustainability commitment. The risk is not that the removal expires; it is that it under-delivers, reverses, or fails to materialise on the schedule the target assumes.

These two exposures sit at opposite ends of the carbon book and are typically managed by no one as a single position. The legacy avoidance credits are a finance and impairment problem. The forward removal offtakes are a procurement and delivery-risk problem. They are rarely held in one view, rarely priced against one another, and rarely monitored on a common basis. That is the unmanaged gap V2.0 has opened.

The organisational reality makes the gap worse. The legacy book usually sits with sustainability or procurement, booked years ago by people who may no longer be in post. The forward programme is being negotiated now, often by a head of carbon removal working alongside energy and finance. The impairment consequence lands on the controller and the audit committee, who inherited both without choosing either. Three constituencies, three time horizons, and no single owner with a complete picture of what the company holds, what it owes, and what it has promised to deliver.

The quality spread sharpens the point. Across 2025 high-rated afforestation, reforestation and revegetation credits roughly doubled, from around 14 US dollars to above 26 US dollars, while the weakest tier languished near 3 US dollars. Quality, not volume, drove value: spot end-user spend was about 1.04 billion US dollars on 168 million credits retired, with retirements actually down on the year. A book weighted toward the cheap, abundant avoidance vintages of three years ago is precisely the book most exposed to the V2.0 reset, and least able to lean on a recovering price for support.

Why "they don't expire" misses the point

04 · Vintage and use case, not shelf life

The reassurance a company tells itself is that its credits do not expire, so nothing has really changed. That is the coal-plant fallacy. The credits still represent the same underlying project; what changed is the demand for them in the only application that gave them their price. When the largest single source of corporate demand, use against a science-based target, is withdrawn by the body that defines that target, the value does not erode gradually. It resets.

The credits still work. The rule that gave them value does not.

The same dynamic is reinforced from other directions. The ICVCM Core Carbon Principles and CORSIA eligibility act as a sorting machine that reprices whole methodologies and vintages at a stroke. The EU's Empowering Consumers Directive, applying from 27 September 2026, bans offset-based product claims such as "carbon neutral" outright, removing another demand channel for avoidance credits. SBTi V2.0 is the largest of these resets, but it is not isolated. The avoidance book faces a coordinated withdrawal of the use cases that priced it.

History rhymes here. The Clean Development Mechanism under the Kyoto Protocol generated more than 8,000 projects across 100-plus countries between 2004 and 2012, and the vast majority of its claimed reductions were later judged unlikely to be additional. That was the original integrity failure the voluntary market inherited. The lesson is not that credits are worthless; it is that the demand underpinning a class of credits can be withdrawn, by a standard-setter, a regulator or a court, faster than the holders adjust. A book valued on yesterday's rules is always one announcement away from a reset, and the reset arrives as a price move, not a notice of expiry.

The enforcement backdrop removes any comfort that this is theoretical. In March 2024 the Amsterdam District Court found against KLM in the first successful airline greenwashing ruling, holding that 15 of 19 statements were misleading, including the claim that offsets neutralise emissions. A Delta class action built on a "carbon neutral" claim survived a motion to dismiss in December 2024, and 21 airlines agreed with the European Commission to drop offset-based neutrality claims. The same avoidance credits that V2.0 demotes are the ones that exposed companies to these actions. The legal risk and the accounting risk now point at the same shelf of the book.

What disciplined buyers do now

The companies that manage this well will stop treating the carbon book as a single sustainability line and start treating it as two distinct financial exposures. They will classify the legacy book against the new SBTi rules, what counts, what is contribution-only, what is effectively stranded, and carry a defensible, independent view of its realisable value into the impairment test. They will treat the forward removal programme as a structured procurement exposure, with delivery, permanence and performance monitored continuously rather than confirmed years late at a verification point.

Above all they will recognise that a standards change is a market event. SBTi V2.0 did not destroy any carbon; it moved the demand. The buyers who understand that markets strand assets when demand changes, not when they expire, will reprice and restructure now, on their own terms. The ones who wait for an auditor or a regulator to force the issue will do it later, at a worse price, in public.