The Registry Reform Fantasy Has Run Out of Road
In 2023, a California federal court received a class action lawsuit alleging that Delta Airlines' claim to be carbon neutral since 2020 constituted deceptive advertising. Delta had followed the rules of the world's largest carbon registries, yet those rules proved insufficient to make the claim true. This legal action exposes a foundational design failure within carbon credit markets: the product being sold cannot be verified, the accounting cannot be trusted, and the incentive structure rewards model manipulation over physical integrity. The voluntary carbon market, which reached a valuation of roughly $2 billion in 2021 and was once projected by BloombergNEF to hit $50 billion by 2030, is not suffering from a governance gap that better administrative procedures can close. No registry revision will fix a market whose core asset, a tonne of carbon not emitted, cannot be observed, only modeled.
The integrity problem is structural, not procedural, and the evidence has been accumulating long enough that continuing to treat it as a growing pain is its own form of dishonesty. The conventional response to every scandal has been to call for tighter registry rules, independent audits, and more disclosure, but that response mistakes a symptom for the disease. By May 2026, the evidence will have accumulated to a point where the honest conclusion is that voluntary carbon markets cannot be reformed into reliability. They must be replaced by a system that measures physical carbon reductions rather than modeling hypothetical scenarios. The financial and reputational risks of pretending otherwise have become too high for institutional participants to ignore.
The dominant narrative in institutional circles holds that carbon markets are a good idea badly administered, suggesting that offset credits from forestry and methane projects allow companies to fund real-world emissions reductions at a lower cost than internal abatement. Proponents argue the market simply needs better standards, more rigorous verification, and independent ratings to function. Verra, the world's largest voluntary carbon registry, has repeatedly updated its Verified Carbon Standard in response to criticism, while Gold Standard positions itself as the premium option. A generation of ratings firms, BeZero Carbon and Sylvera among them, has raised capital on the promise of sorting good credits from bad ones.
This reform argument deserves to be taken seriously, as markets for complex, hard-to-observe assets do sometimes improve through better information infrastructure. Bond ratings, whatever their failures in 2008, eventually produced more accurate pricing for most of the credit universe. The analogy has intuitive appeal for financial institutions looking to build trading desks around environmental commodities, operating under the logic that if the underlying asset can be standardized, the market will eventually price the risk correctly. Yet the analogy breaks down precisely where it matters most. A bond's cash flows are contractual, whereas a forest's carbon storage is contingent on rainfall, fire, disease, political stability, and enforcement capacity across decades.
When South Pole, one of the largest carbon project developers in the world, issued credits from its flagship Kariba REDD+ project in Zimbabwe, it was not making a contractual promise; it was making a model-dependent projection. The entire premise of the REDD+ framework relies on predicting what would have happened to a forest if the project had not existed and then issuing credits against that hypothetical baseline. Researchers at the Berkeley Carbon Trading Project found the project had been issuing credits at a rate that assumed a deforestation baseline far higher than observed reality. The result was hundreds of millions of tonnes of credits sold to corporate buyers, including major consumer brands, that almost certainly did not represent real emissions reductions. Verra suspended issuance from Kariba in 2023, and South Pole parted ways with the project shortly after, but the credits had already been retired by corporate buyers, permanently locking the fictitious emissions reductions into corporate sustainability reports.
BeZero Carbon and Sylvera, the ratings firms meant to prevent exactly this outcome, had rated Kariba credits in the mid-range of their scales before the scandal broke. This is not a rounding error; it is a ratings failure comparable to AAA-rated mortgage securities going to zero. It happened inside a market that the reform consensus was already describing as maturing. The failure of these rating agencies to identify the baseline manipulation at Kariba suggests that the problem is not a lack of data, but the fundamental impossibility of verifying a counterfactual scenario. You cannot audit a tree that was never going to be cut down in the first place.
Four Numbers That Condemn Carbon Credit Markets Outright
The first number is 90. A 2023 joint investigation by The Guardian, Die Zeit, and SourceMaterial found that more than 90% of Verra's rainforest offset credits, the REDD+ category that constitutes the backbone of the voluntary market, were likely phantom credits with no real climate benefit. While Verra disputed the methodology, the investigation relied on peer-reviewed satellite analysis published by academics at multiple universities. This shows that the integrity problem is not concentrated in a few bad actors; it is a systemic failure of the underlying asset class. When nine out of ten units of a financial product are deemed worthless by independent academic review, the market is not experiencing a quality control issue; it is operating a structural fiction.
The second number is 235 million, representing the approximate number of carbon credits issued by the Kariba REDD+ project over its operational life, according to Verra's own registry data. The Berkeley Carbon Trading Project's analysis suggested that the project's deforestation baseline was overstated by a factor that rendered the vast majority of those credits fictitious. Major corporations that publicly committed to net-zero targets anchored in offset purchases retired these credits as proof of climate action, while their actual physical emissions remained unchanged. The market was not mispricing risk; it was pricing in a fiction and calling it due diligence. The sheer scale of the Kariba issuance demonstrates how easily model-dependent accounting can be weaponized to generate massive volumes of compliance-grade assets out of thin air.
The third number is $50 billion, BloombergNEF's projection for the voluntary carbon market by 2030, published when institutional appetite for credits was at its peak. By 2024, actual market activity had collapsed well below 2021 levels. Trading volumes fell sharply, and credit prices in many nature-based categories dropped by 70% or more from their highs. The collapse was not driven by reform; it was driven by the market finally pricing in what the academic research had been saying for two years: most of the product is worthless. The gap between BloombergNEF's $50 billion projection and the reality of a shrinking, illiquid market highlights the danger of extrapolating growth curves based on corporate public relations commitments rather than physical fundamentals.
The fourth data point is Delta Airlines. In 2023, a class action lawsuit was filed against Delta in a California federal court, alleging that the airline's claim to be carbon neutral since 2020, a claim built entirely on voluntary offset purchases, constituted deceptive advertising. Delta followed Verra's rules, but those rules were not sufficient to make the claim true. That distinction matters enormously, because it means compliance with existing standards is itself a liability, not a defense. If a corporation can follow every rule established by the world's largest carbon registry and still face a deceptive advertising lawsuit, the registry's standards possess zero protective value for the buyer. The legal burden has shifted from the registry issuing the credit to the corporation retiring it.
The Article 6 Objection Deserves a Direct Answer
The strongest counter-argument is that Article 6 of the Paris Agreement represents a fundamentally different architecture from the discredited voluntary market. Under Article 6.4, a new UN-supervised crediting mechanism would require host countries to make corresponding adjustments, meaning a credit sold to a foreign buyer would not also count toward the selling country's national climate target. This eliminates the double-counting problem that undermines much of the existing voluntary market. Several serious analysts, including researchers at the Carbon Market Institute, have argued that Article 6 creates conditions for a credible international carbon market that bears little resemblance to Verra's REDD+ credits.
This argument deserves respect because corresponding adjustments are a real improvement in accounting architecture. But Article 6 does not resolve the permanence, additionality, or measurement problems. A forest that earns credits under Article 6.4 can still burn, and the deforestation baseline can still be manipulated by host countries seeking foreign capital. The monitoring, reporting, and verification infrastructure in most host countries remains inadequate for the precision that carbon accounting requires. Moving the oversight from a private registry to a UN-supervised body changes the governance structure, but it does not change the physical reality of the asset being measured.
COP28 in Dubai and subsequent negotiations have repeatedly failed to finalize the Article 6.4 rules precisely because the technical problems are not administrative; they are physical. The data that would make this analysis wrong would be a peer-reviewed study showing that a cohort of at least 50 Article 6.4 credited projects, verified over a minimum 10-year period with independent satellite monitoring, delivered carbon reductions within 15% of their projected baselines. That study does not exist. Until it does, Article 6 is a promise, not a market. The reliance on UN supervision does not alter the laws of physics or the limitations of satellite forestry monitoring.
Who Gets Hurt When the Credits Are Fiction
The failure of carbon credit markets carries consequences that extend well beyond the companies that bought bad offsets. Three stakeholder groups face distinct near-term exposure, and each requires a different strategic response.
Institutional Investors
Asset managers who built carbon market exposure through funds linked to Xpansiv or through listed positions in carbon project developers are sitting on assets whose fundamental value proposition has been called into question. The more immediate risk is regulatory. The SEC's climate disclosure rules and their European equivalents under CSRD are pushing companies to distinguish between actual emissions reductions and offset purchases. As that distinction becomes mandatory in financial reporting, the accounting treatment of retired credits will come under audit scrutiny that the voluntary market was never designed to survive.
The near-term trigger is the first wave of CSRD-compliant filings from large European companies, expected through 2025 and 2026. Any company that has claimed carbon neutrality on the basis of voluntary credits faces potential restatement risk. Investors in those companies should be pricing that risk now, not after the first adverse judgment. The transition from voluntary sustainability reports to audited financial disclosures will brutally expose the gap between modeled carbon credits and physical reality. When financial auditors begin demanding the same level of proof for carbon offsets as they do for capital expenditures, the valuation models for these assets will collapse entirely.
Enterprise Buyers
Microsoft has publicly committed to being carbon negative by 2030 and has been one of the most active corporate buyers of carbon credits, including direct air capture credits from Climeworks and CarbonCapture. The distinction in Microsoft's portfolio, between high-quality engineered removals and lower-quality nature-based offsets, is the only distinction that will matter inside a courtroom. Companies that shifted early to credits with physical permanence guarantees are insulated, but companies still holding large positions in REDD+ credits are not.
The concrete near-term trigger for enterprise buyers is the first major greenwashing verdict against a Fortune 500 company in a U.S. or EU court. Legal observers tracking the Delta case and several European aviation cases involving carriers like Lufthansa expect a consequential ruling within the next 18 months. When that verdict arrives, the reputational cost of holding low-quality offset positions will force immediate portfolio reviews across every industry that made carbon neutrality claims between 2019 and 2023. The legal precedent will establish that ignorance of a registry's methodological flaws is not a valid defense against consumer fraud allegations.
Product and Engineering Teams
Companies building monitoring, reporting, and verification technology face a bifurcated future. Firms like Pachama, which uses satellite data and machine learning to verify forest carbon, must handle a market that is rapidly losing faith in the underlying asset. If the market consolidates around engineered removals with physical verification, demand for forest monitoring infrastructure collapses. If it pivots toward Article 6 compliance instruments, verification requirements become more stringent, and the competitive advantage shifts to firms with government contracting relationships rather than corporate sales teams.
The strategic pivot needs to happen before the market forces it, and the window is closing. Engineering teams that spent the last five years building software to optimize the issuance of voluntary credits must now retool their platforms to support compliance-grade physical verification. The capital markets will no longer fund technology that merely makes it easier to trade a flawed asset. The future belongs to infrastructure that can definitively prove a tonne of carbon has been permanently removed from the atmosphere.
A Timeline for Market Consolidation
First, at least one major consumer brand currently holding a carbon-neutral or net-zero product certification based primarily on Verra-certified nature-based credits will face a successful regulatory action or court judgment in either the United States or the European Union before the end of 2026. The most exposed names are in aviation and consumer goods. Delta, Lufthansa, and several fast-moving consumer goods companies that made carbon neutrality claims at the peak of the market in 2021 and 2022 are the names to watch. A single adverse judgment will trigger a cascade of restatements across the sector.
Second, by the end of 2027, traded volume in nature-based voluntary offsets will be less than 20% of its 2021 peak. The residual market will be concentrated in engineered removal credits priced above $200 per tonne. BeZero Carbon and Sylvera will have either pivoted to rating compliance instruments under Article 6 or will have been acquired by larger data providers. The $50 billion market BloombergNEF projected will not arrive. What arrives instead will be smaller, more expensive, and finally real. The market got the easy version of carbon accounting for two decades. The reckoning is not coming; it is already here.
Carbon Credit Markets
If voluntary credits are worthless, why do compliance markets like the EU ETS still function?
Compliance markets are fundamentally different instruments. The EU ETS sets a hard cap on emissions from covered sectors and reduces it over time, and a company that exceeds its allowance pays a real penalty. The EU ETS price traded above 50 euros per tonne through much of 2023 and 2024 precisely because the scarcity is enforced by state power. Voluntary REDD+ credits traded below $5 per tonne for the same reason: no cap, no enforcement, and no penalty for buying worthless product. The price gap is not a market inefficiency; it is the market correctly valuing two entirely different things. Investors looking for carbon exposure must recognize that compliance markets trade on regulatory scarcity, while voluntary markets trade on corporate public relations budgets. The failure to distinguish between these two markets has been one of the most costly analytical errors made by institutional investors in the climate space.
Didn't Verra's 2023 methodology revisions substantially fix the baseline problem?
Verra did revise its REDD+ methodology in 2023, introducing jurisdiction-level baselines to replace project-level ones. It is a genuine improvement in accounting logic, but the revision also invalidated much of the existing credit stock and caused further collapse in credit prices. The market's response to better standards was to exit, not to upgrade. That tells you something important about buyer motivation. The buyers who valued the old credits valued them precisely because they were cheap and easy to retire. Tighter standards produce fewer credits at higher cost, and corporate buyers are already questioning whether the reputational benefit justifies the expense. Reform is driving buyers out of the market, not restoring their confidence. When the primary appeal of a product is its low cost, improving the quality at the expense of the price destroys the core value proposition for the buyer.
What about companies actually delivering legitimate carbon removal?
They exist and deserve to be distinguished from the discredited offset market. Climeworks' direct air capture facility in Iceland permanently mineralizes captured CO2 in basalt rock. The cost is currently above $1,000 per tonne, and Microsoft and Stripe are among its buyers. This is real carbon removal with physical permanence, but it is not a carbon credit market in any meaningful sense. It is a procurement contract for a verified physical service. That distinction is the entire argument. The path forward is procurement, not offsets. The sooner corporate climate strategies are built around that distinction, the less legal and reputational exposure they carry into the next decade. Buying a physical service guarantees that the carbon is removed, whereas buying a credit only guarantees that a model has been satisfied.
Related MarketIntel briefing: read Carbon Credit Markets Are a Mirage, Not a Fix for a connected view on this market signal.
Source context: readers can compare this market signal with broader data from Gartner.
Related MarketIntel briefing: read Carbon Credit Markets Are Broken for a connected view on this market signal.