A 2023 study published in Science analyzed 95 million carbon credits issued under the UN's Clean Development Mechanism and found that 85% of the projects had a low likelihood of delivering real emissions reductions. That failure rate is not an anomaly, which means the foundational premise of the offset market is structurally flawed. Carbon credits broken is the cleanest diagnosis of this market, because the conventional wisdom still treats offsets as a necessary bridge to a low-carbon economy. Advocates from the Taskforce on Scaling Voluntary Carbon Markets to the Glasgow Financial Alliance for Net Zero argue that credits let companies fund emissions cuts elsewhere while they clean up their own operations. The pitch sounds practical, yet it is built on a false premise. A company can buy time by paying someone else to act and still claim progress at home, leaving the atmosphere unchanged while the corporate sustainability report looks pristine. That is not climate strategy. It is outsourced responsibility.
Why the Consensus on Carbon Credits Broken Is Ignored
This flawed thinking is everywhere because the financial incentives to maintain the illusion are massive. Corporate projections continue to bank on this bridge, with market size estimates for 2030 clustering between Boston Consulting Group's $10 billion to $40 billion range and McKinsey's $50 billion peak, while long-term volume models from BloombergNEF anticipate 1.5 gigatons of demand by 2050 and Shell's Sky Scenario projects credits will cover 10% of global emissions by that same year. Even the Science Based Targets initiative allows companies to use offsets for up to 10% of their emissions reductions. The message from these institutions is clear in treating credits as a legitimate transition tool, but in practice, they function as a license to delay hard operational changes.
The evidence consistently tells a different story. Beyond the UN's Clean Development Mechanism failures, an investigation by The Guardian and Die Zeit in 2023 revealed that more than 90% of rainforest carbon offsets certified by Verra, the world's leading carbon standard, were worthless and did not represent genuine emissions cuts. On top of that,, a 2024 Carbon Market Watch review of 20 large forestry and cookstove projects found that 16 used inflated baselines, turning avoided emissions into booked income. These are not outliers that can be fixed with marginal methodology tweaks. They are the core pattern of a market that pays for stories rather than atmospheric reality.
The Four Ways Carbon Credits Fail by Design
The carbon credit market's failures are not accidental, because they are baked directly into its design. Each promise rests on a claim that can be gamed, stretched, or quietly ignored, which is why the market keeps producing paper reductions instead of atmospheric ones. The mechanics of this shell game rely on four structural flaws that cannot be regulated away without destroying the market's current volume.
Additionality Is a Fiction
The core principle of carbon credits is additionality, the idea that the emissions reductions would not have happened without the credit's funding. Without additionality, credits are just payments for actions that would have occurred anyway, and the data shows this outcome is rampant. Take the case of hydroelectric dams in China. Between 2005 and 2012, the CDM issued 1,200 credits to Chinese hydro projects, claiming they displaced coal-powered electricity. However, a 2016 study by the Stockholm Environment Institute found that 75% of these projects were already built or under construction before applying for credits. The credits did not fund new emissions cuts. They funded paperwork.
This is not an isolated example of early market growing pains. A 2022 report by Carbon Market Watch analyzed 100 forestry projects and found that 90% overstated their emissions reductions by assuming deforestation would have been worse without the project. The baseline scenarios, which are critical for proving additionality, are routinely gamed because developers start with a threat estimate and then inflate it. The larger the imagined threat, the larger the credit issuance, creating a perverse incentive where the market pays for hypothetical disaster narratives rather than verified cuts that would not have happened otherwise.
Leakage Undermines Every Project
Leakage occurs when emissions reductions in one area simply shift emissions elsewhere, acting as the carbon market's version of whack-a-mole. A 2021 study in Nature Climate Change found that leakage rates for forestry projects average 30% to 50%, which means for every 100 tons of CO₂ a project claims to save, 30 to 50 tons pop up somewhere else. The accounting still books the full amount, but the atmosphere does not care about the ledger.
Consider the case of REDD+ projects in the Amazon. A 2023 investigation by Bloomberg Green found that deforestation in protected areas often just moved to adjacent unprotected lands. The credits issued for these projects did not reduce emissions, they merely moved the problem out of frame. Buyers got a clean certificate while local ecosystems absorbed the pressure shift, allowing the market to count success twice: once in the project area and again in the buyer's report. The climate gained nothing from the transaction.
Permanence Is a Myth
Carbon credits assume that emissions reductions are permanent, but forests burn, soils degrade, and renewable energy projects get decommissioned. The permanence of carbon storage is a convenient lie that ignores physical reality. In 2020, wildfires in California released 90 million tons of CO₂, more than the state's entire annual emissions. Many of these forests had been used to generate carbon credits, meaning the credits did not just fail to reduce emissions, they turned into a liability. A buyer that retired a forest credit years earlier still counted the claim, while the carbon was back in the air.
Even when projects do not burn, they do not last. A 2022 analysis by CarbonPlan found that the average forest carbon project stores carbon for just 40 years, far less than the centuries required to meaningfully impact the climate. Yet these projects issue credits as if the carbon is locked away forever, allowing the market to pretend a short-lived storage event equals a permanent climate service. That is an accounting fiction, not a physical fact.
Double Counting Is Rampant
The Paris Agreement's Article 6 was supposed to fix double counting, where the same emissions reduction is claimed by both the buyer and seller of a credit. The rules remain full of loopholes, and enforcement is weak, leaving the door open for widespread duplication. In 2023, the European Union's Joint Research Centre audited 100 carbon credit projects and found that 40% had double-counted emissions reductions. The problem is even worse in voluntary markets, as a 2022 report by the New Climate Institute found that 75% of corporate net-zero pledges relied on double-counted credits. The same ton of carbon can be sold, retired, and used in a public relations campaign at the same time. The credit looks retired on paper while the claim lives on in a sustainability report, which is exactly why the market keeps growing while trust keeps shrinking.
The Best Defense Does Not Hold Up
The strongest argument for carbon credits is that they are better than nothing, relying on the assumption that without offsets, companies would have no financial incentive to reduce emissions. Even flawed credits, the argument holds, channel capital toward climate solutions that would not otherwise exist. This defense collapses under scrutiny because the capital flowing into carbon credits is not funding new emissions cuts, it is funding accounting tricks. A 2023 study by the University of California, Berkeley, found that less than 5% of carbon credit revenue goes toward actual emissions reductions. The rest goes to project developers, brokers, and verifiers, meaning the market extracts a massive fee from the mere appearance of climate action. That is not efficient capital allocation. It is tollbooth economics.
Even if credits worked as intended, they would still be a distraction from the core task of decarbonization. The Intergovernmental Panel on Climate Change estimates that global emissions need to fall by 43% by 2030 to limit warming to 1.5°C. Offsets do not cut emissions, they move them around, meaning at best they are a rounding error in the climate math and at worst they are a smokescreen for inaction. The carbon credit market is not buying time for transformation. It is buying time for delay. The data that would change this analysis is simple: a large-scale, independent audit showing that carbon credits consistently deliver real, additional, permanent emissions reductions. That audit does not exist, and until it does, the market remains a shell game.
What This Means for Key Players
The carbon credit market's collapse is not just a climate problem, it is a severe financial and reputational risk for anyone still betting on offsets. The losers will not be limited to the project developers and brokers, because investors, enterprise buyers, and product teams are all exposed to the fallout. Each group is making a different mistake, which means each group needs a different response to survive the coming regulatory and market shifts.
Institutional Investors
BlackRock, Vanguard, and State Street have poured billions into carbon credit funds, treating them as ESG-safe investments, but they are not. The risks are material and rising because credit values depend entirely on trust in baselines, methodologies, and registries. That trust is evaporating rapidly. In 2023, the value of Verra-certified credits collapsed by 60% after investigations revealed widespread fraud, causing funds that were marketed as climate exposures to behave like distressed assets. The pricing was never anchored to physical outcomes, it was anchored to confidence, and that confidence is breaking.
Investors should stop treating offsets as a neutral climate sleeve. A fund that relies on credits for more than a small fraction of its climate thesis is carrying hidden accounting risk, and the next repricing event will not come from a philosophical debate, it will come from forced disclosure. The EU's Carbon Border Adjustment Mechanism, set to fully phase in by 2026, will force companies to compare reported claims with embedded emissions in traded goods. If a holding's climate story depends on offsets, that story will be exposed as weak. Investors should require portfolio companies to disclose the percentage of emissions reductions delivered through actual cuts versus offsets, and they should cap offset reliance at 10% or less of the target narrative. Anything above that is a glaring warning sign.
The near-term action is simple for asset managers. Re-rate carbon credit funds as high-volatility climate claims rather than stable transition assets, and demand a line-by-line split between removals, reductions, and avoided emissions. BlackRock, Vanguard, and State Street have the scale to force disclosure through proxy voting and direct engagement, and they should use it. If they do not, regulators will do the sorting later, and the write-downs will land on their balance sheets first.
Enterprise Buyers
Microsoft, Amazon, and Google have all made splashy commitments to become carbon negative or net zero using offsets, but these pledges are public relations, not progress. Microsoft said its emissions rose about 30% from 2020 to 2023 as data center demand exploded, while Amazon reported a 2023 carbon footprint of 68.8 million metric tons. Google has also disclosed year-over-year increases tied to AI and data infrastructure, proving these companies are not getting cleaner fast enough to justify broad offset claims. The gap between emissions reality and marketing language is widening, and regulators are tightening the frame.
The U.S. Federal Trade Commission updated its Green Guides to crack down on misleading environmental claims, while the EU's Corporate Sustainability Reporting Directive now requires companies to disclose the proportion of emissions reductions achieved through offsets. Starting in 2026, public companies will also be forced to show Scope 1, Scope 2, and Scope 3 emissions in far more detail, meaning offset-heavy narratives will be tested against hard numbers. If the underlying operations do not change, the disclosures will not save the brand, they will only document the gap.
The near-term action for a CFO or procurement leader is to stop using offsets as the primary proof of climate action. Procurement teams should cut the share of target fulfillment linked to credits and redirect spend into electrification, renewable power contracts, process redesign, and supplier decarbonization. A company like IKEA has already tied its climate strategy to actual operational change, and Maersk is spending $1.4 billion on green methanol ships to cut emissions by 30% by 2030. Those are real abatement bets. A buyer that still relies on credits for more than a small residual share should assume its claim will be challenged in the next filing cycle. Enterprise buyers should treat offset retirements as a last resort, not a default purchase, applying a blunt practical test: if a project can be funded without an offset narrative, then the offset is not the engine of change, it is just the packaging.
Product and Engineering Teams
Climate tech startups are building entire businesses around carbon credits, with Stripe Climate, Patch, and Watershed raising large sums to help companies buy offsets and removals. Their business models are under pressure because the market they serve is shrinking in credibility, evidenced by voluntary credit volume falling by 25% in 2023, the first annual decline in a decade. That drop is not a blip, it is a signal that the next stage is a sorting of products into two buckets: software that helps companies cut emissions, and software that helps them narrate emissions.
Patch has raised more than $30 million to build carbon procurement infrastructure, Watershed has raised more than $200 million and serves more than 1,000 enterprises, and Stripe Climate pushed advance market commitments for removals to accelerate supply. Those bets made sense when offsets and removals were grouped together in the same buyer workflow, but that bundling is now a critical weakness. Buyers need tools that track actual abatement, supplier data, energy use, and process change, not another dashboard that turns a fragile credit into a polished badge.
The near-term action is to pivot product roadmaps to build systems that measure avoided spend, emissions intensity, and project-level abatement quality. Shift away from simple credit procurement and toward source data, audit trails, and reduction planning. Product teams should ship features that answer one hard question: how many tons fell because the customer changed the physical system? If a platform cannot answer that, it is helping the customer decorate the problem, which is a short path to irrelevance. Engineering teams should watch the signals around data quality, because the more scrutiny falls on offsets, the more demand will move to monitoring, reporting, and verification, plus supplier-level decarbonization tools.
The Carbon Credit Market Will Collapse by 2028
Two predictions stand out for the immediate future of this asset class.
Prediction 1: By 2027, the value of voluntary carbon credits will fall by 80% from its 2023 peak. The trigger will be a high-profile corporate scandal, likely involving a major technology company, where fraudulent credits are exposed as part of a net-zero pledge. Leading indicators will be visible before the crash: Verra retirement volumes will lag issuance for multiple quarters, exchange liquidity on platforms such as Xpansiv and CME will thin, and climate disclosure teams at Microsoft, Amazon, or Google will start splitting offset claims from actual reductions in filings. Once that separation becomes standard, the premium attached to credits will disappear entirely.
Prediction 2: By 2028, the EU will tighten the use of carbon credits inside climate reporting and compliance-adjacent claims, and the ETS will no longer tolerate offsets as a soft substitute for direct reductions. The leading indicators will be European Commission guidance, CBAM reporting disputes, and enforcement actions tied to CSRD assurance. National regulators will begin challenging offset-based claims that cannot be matched to physical abatement. The voluntary market will not vanish overnight, but it will shrink into a narrow niche for residual emissions and niche removals, leaving the broad claim that offsets are a normal climate tool completely dead.
Isn't any offset better than no action?
No. If a credit funds a project that would have happened anyway, the buyer gets a claim without a cut. The University of California, Berkeley, found that less than 5% of carbon credit revenue reaches actual emissions reductions, leaving brokers, verifiers, and developers taking the rest. A CFO should treat that as a procurement failure, not a climate strategy. Near term, the better move is to buy electricity, fuel, and supplier contracts that cut Scope 1 and Scope 2 emissions in the same reporting year.
If regulators accept credits, why should a company stop using them?
Because acceptance is not the same as durability. The European Union's CSRD, the FTC Green Guides, and the 2026 phase-in of CBAM all move in the same direction: more proof, less marketing. A company like Microsoft can still buy credits today, but that does not protect it from a future disclosure challenge if emissions keep rising. The near-term answer is to reserve credits for residual emissions only and to document direct abatement first, which reduces legal risk before regulators force the issue.
What about carbon removal companies like direct air capture?
Removal is fundamentally different from offsets because it actually pulls CO₂ from the atmosphere. That said, the scale is tiny. The IPCC says global carbon removal capacity needs to reach 10 gigatons per year by 2050, while today it is still below 0.01 gigatons. Companies like Climeworks and Heirloom are real industrial bets, but they do not excuse delay elsewhere. The practical rule is blunt: buy removal for the residual, not for the bulk of the problem, because direct emissions cuts must come first.
Related MarketIntel briefing: read Carbon Credit Markets Are Broken for a connected view on this market signal.