Verification Is the Arbitrage: How Credits Cross the Quality Divide

The voluntary carbon market has split in two. Credits carrying the Core Carbon Principles label trade at a clear premium, while generic avoidance offsets have fallen below $1 per tonne. The obvious lesson is to buy quality. The more valuable one is that the dividing line is verification, and verification is where the returns are.

Every carbon commentator will tell you the market is bifurcating in 2026. They are right, and it is the least interesting thing to say about it, because it is already priced in. This is written for allocators and corporate treasuries that hold a carbon position: the question is not whether quality wins, but where the value accrues when it does. The answer is not the premium credit. It is the act of proving a credit deserves the premium.

The split, in numbers

The gap is no longer subtle. Generic avoidance credits without a quality label have fallen below $1 per tonne. High-integrity credits carrying the CCP label trade at multiples of that, and the premium has more than doubled since 2023.

The clearest signal is that the two ends are moving in opposite directions at the same time. Investment-grade rated credits (BBB+ and above) averaged around $20 per tonne in early 2026, up on the year, while lower-rated credits averaged under $8 and fell. Quality is repricing up. Everything else is repricing down. And the filter is narrow: only a small fraction of independently rated projects currently earn the CCP label.

Removals sit in their own tier above all of it. Nature-based removals run roughly $7 to $24 per tonne, biochar trades around $125 to $200, and direct air capture averages well above $500, reaching four figures at the top. Nature-based avoidance below about 15 euros is already disappearing from buyer-grade portfolios as CCP filtering tightens.

Why the obvious move is the wrong one

The consensus takeaway is to buy the premium tier. It is crowded and it is fragile, for two reasons. First, supply is thin. Only a small share of projects carry the label, so buying the top tier means competing with the deepest-pocketed buyers in the market for a narrow pool of credits. You are a price-taker in the most contested corner of the market.

The concentration problem

Second, demand at the top is dangerously concentrated. Microsoft alone accounted for roughly 90% of carbon-removal purchases in 2025. In 2026 it paused new removal purchases while it reviewed its climate strategy, partly because its own footprint was growing as it built AI data centres. Existing deals were left in place, but the message to the sector was immediate: when one buyer is nine-tenths of your demand, that buyer's pause is your revenue cliff. A strategy of chasing the same premium credits as the giants leaves you exposed to the same concentration risk, without their bargaining power.

The real divide is not quality. It is verifiability.

Here is the reframe. A credit's tier is not a fixed property of the project. It is a function of what can be proven about it. A mid-quality credit with weak documentation sits in the discount bin. The same credit, independently verified against a recognised methodology, can move up into the financeable tier. That movement, from unproven to proven, is where value is created, and it is the part almost no one is pricing.

Most buyers treat the quality tier as given and shop within it. The more valuable position is to own or back the thing that moves credits between tiers: verification. When verification reclassifies a credit from discount to premium, the spread it crosses is the return.

The Verification Arbitrage

Put it on two axes. One axis is whether a credit is independently verifiable. The other is whether it currently sits in the premium tier. That gives four boxes. Top-tier and already verified credits are fully priced; there is no edge left there. Discount credits that cannot be verified are cheap for a reason; leave them. The interesting box is the one everyone ignores: credits with real underlying quality that are stuck in the discount tier only because their evidence is thin. Verification lifts them across the divide, and the infrastructure that performs that lift, audit-grade MRV, is the asset an allocator should want to own. You are not betting on which credits are good. You are backing the mechanism that decides.

Why this is structural, not a cycle

Bifurcations driven by sentiment reverse. This one is written into disclosure rules, which do not. Under ESRS E1-7 and ISSB-aligned reporting, a corporate buyer must now document the quality, additionality, and permanence of any credit it claims, or it becomes a disclosure liability on the balance sheet. That permanently removes the buyer-grade market's appetite for cheap, undocumented credits. The demand for verification is not a fashion. It is a compliance requirement with a date on it, and it only tightens.

That is why nature-based avoidance below roughly 15 euros is already leaving serious portfolios, and why the premium for provable quality keeps widening. The market is not rewarding expensive credits. It is rewarding provable ones.

What to do now

For an allocator, the move is to position in the verification and MRV layer rather than speculate on individual credits. The credit market is contested and concentrated. The layer that determines which credits qualify is neither, and every tightening disclosure rule feeds it. This is the infrastructure Tokere builds, and where tokenization sits as an optional layer on verified supply.

For a corporate treasury, build a portfolio rather than chase a tier: a mix of avoidance, nature-based removals, and durable removals, with independent verification as the non-negotiable due-diligence gate. Do not over-index on the exact supply the mega-buyers favour, because you will pay their prices and inherit their concentration risk. Weight toward credits whose quality you can prove yourself, not credits whose quality depends on a label everyone is fighting over.

Where the arbitrage breaks

One honest limit. Verification lifts credits that have real underlying quality but thin evidence. It cannot rescue a project that was never additional, or a tonne that was never real. Proving a bad credit only produces a well-documented bad credit, the same lesson the tokenization experiments taught the market a few years ago. The arbitrage works on the gap between quality and proof, not on the absence of quality.

That is why the discipline is verify-first, not verify-everything. The value sits in finding credits the market has underpriced because their evidence is incomplete, and closing that evidence gap. It is not in laundering weak supply into the premium tier, which is exactly the failure that discredited the voluntary market in the first place.

The next two years

Expect three things. CCP-eligibility hardens from a quality signal into a procurement gate, the credential you need to sell into corporate buyers at all. Removals scarcity bites, because net-zero targets increasingly require removals to cancel residual emissions, and the durable supply is small and largely spoken for. And verification becomes the margin, the difference between a stranded credit and a financeable one, and therefore the part of the value chain worth owning.

The buyers who win the next phase will not be the ones who bought the most premium credits. They will be the ones who understood that the premium itself is manufactured by verification, and positioned where that value is created rather than where it is merely paid.

Frequently asked questions

Why do high-integrity carbon credits cost more?

Because buyers now carry disclosure and reputational risk for the credits they use, and high-integrity credits (for example, those with an ICVCM CCP-approved methodology) reduce that risk. In 2026 the premium has more than doubled since 2023, while generic avoidance credits have fallen below $1 per tonne.

What is the CCP label and how rare is it?

The Core Carbon Principles label, from the Integrity Council for the Voluntary Carbon Market, marks credits from methodologies assessed as high-integrity. It is a genuine filter: only a small share of independently rated projects currently qualify, which is part of why labelled credits command a premium.

What is the difference between avoidance and removal credits?

Avoidance credits represent emissions prevented, for example protecting a forest. Removal credits represent carbon actively taken out of the atmosphere and stored, for example biochar or direct air capture. Removals are scarcer, more durable, and far more expensive, ranging from roughly $7-24 for nature-based up to $500 and above for direct air capture.

Is the carbon market bifurcation permanent?

It is structural rather than cyclical, because it is driven by disclosure rules like ESRS E1-7 and ISSB standards, not by sentiment. As long as buyers must document the quality of credits they claim, demand for cheap, undocumented credits stays suppressed and the premium for provable quality persists.

Where is the investment opportunity in carbon markets in 2026?

Less in picking premium credits, which is crowded and concentrated, and more in the verification and MRV layer that decides which credits qualify for the premium. As disclosure tightens, that layer captures value from every tier of the market rather than betting on one.