One label, two markets
The phrase carbon credit flattens two economically unrelated products. The first is avoidance - paying so that emissions which might have occurred do not, typically through forest protection or fuel switching. The second is removal - physically extracting carbon dioxide already in the atmosphere and storing it, whether in biomass, soils, mineralised rock or engineered reservoirs. They trade under the same nominal unit of one tonne of CO2, and that shared unit disguises a growing chasm in price and quality.
Avoidance credits have behaved like a low-grade commodity: abundant, hard to verify, prone to over-crediting scandals, and priced accordingly at a few dollars per tonne. Durable removal is a different asset. It is scarce, physically measurable, and increasingly demanded by corporate buyers who need removals rather than avoidance to make credible net-zero claims. The market has not commoditised these together because they are not the same thing.
The pricing tiers that are emerging
Price in this market tracks two attributes above all: durability and measurability. Durability is how long the carbon stays stored - a distinction between decades for some nature-based approaches and a horizon measured in centuries or millennia for mineralisation and geological storage. Measurability is how precisely the removed tonne can be quantified and verified rather than modelled. Credits that score high on both clear at a large multiple of the undifferentiated voluntary market.
Illustratively, the tiers run from low-single-digit dollars for cheap avoidance, to a middle band for shorter-duration nature-based removal, to durable engineered removal that has cleared at figures on the order of hundreds of dollars per tonne. That is not a temporary dislocation to be arbitraged away; it reflects genuine differences in cost of delivery and in the integrity a serious buyer requires.
Why operator-backed projects capture the premium
The integrity premium does not accrue to whoever holds a credit - it accrues to whoever can demonstrably produce a real, durable, measured tonne. That favours the operator over the trader. An operator-backed project controls the physical process, owns the measurement chain, and can stand behind delivery and reversal risk in a way a secondary-market intermediary cannot. Buyers paying a premium for integrity are paying for a counterparty they can trust to have actually done the removal.
This is why the investable position sits at the project and operator level rather than in trading undifferentiated credits. A single-asset structure that owns a durable removal process captures the spread between the cost of production and the integrity-tier clearing price. As disclosure standards tighten, that spread should widen rather than compress, because demand is migrating toward exactly the scarce, high-integrity supply these projects produce.
What it means for an allocator
The allocator's error is to treat carbon as a single volatile commodity and either avoid it entirely or trade the liquid, low-quality end. The more defensible exposure is to underwrite durable removal at the operator level, where the return is the production spread and the tailwind is a demand shift toward verifiable permanence. Diligence should concentrate on the physical durability claim, the independence and rigour of measurement, and the reversal-risk provisions.
The honest risks are policy dependence and demand timing. Much of the removal-price premium rests on voluntary corporate commitment and evolving standards, both of which can move. But the direction of travel - toward stricter definitions of what counts as a credible tonne - works in favour of high-integrity, operator-backed supply. For patient capital, the fact that removal credits are not yet a commodity is not a reason to wait; it is the entire source of the margin.