Climate tech, sustainability & carbon markets

Climate tech, sustainability & carbon markets

Industry ID: 20 | Slug: climate-sustainability | Researched: 2026-09-15 | Analyst: agent

Method note, read first. The shared WebSearch budget was exhausted before this sector began: one search was attempted and returned a budget message. All research below was conducted by direct WebFetch against primary sources — EUR-Lex, the European Parliament Legislative Observatory, the European Commission, EFRAG, the SEC, EPA, EIA, IEA, UNEP, UNFCCC, the World Bank, ICVCM, Gold Standard, Verra, SBTi, CDP and company reports. Tier-A verification is consequently strong; discovery is narrow, and several things a search would have surfaced are recorded as gaps in §13 rather than guessed at. This sector should be re-run at full search budget. Per the macro brief's standing instruction, the reduced discovery capability is recorded explicitly rather than hidden.

1. Definition and boundaries

In scope. Carbon dioxide removal and carbon markets, both voluntary and compliance; corporate sustainability reporting, assurance and the current retrenchment of both; climate risk analytics; adaptation and resilience; circularity and waste; industrial decarbonisation; water; and climate finance including climate-labelled venture capital.

Explicitly out of scope. Power generation and the grid, which belong to sector 05. That split is deliberate and this dossier exists to test it. Also out of scope: electric vehicles and the regulatory-credit pool (sector 10), insurance pricing and catastrophe-loss underwriting (sector 07), and the physical construction of resilience infrastructure (sector 18).

The boundary disputes, named.

1. CCUS sits on the line and the line moves. Point-source capture on a gas plant is arguably sector 05; direct air capture selling removal credits is unambiguously ours. The IEA reports them together at just over USD 5 billion in 2025. This dossier treats the credit and the removal service as in scope and the generating asset as 05's.
2. "Climate tech" venture funding is now mostly not climate. Sightline Climate's H1 2026 total is carried by data-centre deals classified under Built Environment. Those deals are sector 05's and sector 18's by substance. We report the figure and decompose it rather than inheriting the label.
3. Insurance. Sector 07 owns the pricing cycle and explicitly notes that "climate modelling itself belongs to a climate sector; the pricing consequence is ours." We therefore cover the availability and data side — who models the risk, who owns the data — and defer premium and loss-ratio analysis to 07.
4. Water is in scope but is the thinnest-evidenced part of this dossier; see §13.

2. Subcategories

3. Market structure

Concentration: extreme, and in an unusual place. This is not an oligopoly of sellers. It is a monopsony of buyers in its most-discussed segment. Microsoft accounts for 37,174,479 of the 49,468,711 tonnes of durable carbon removal ever contracted — 75.1% of the entire market (CDR.fyi, 2026-09-15). The next four buyers combined — Frontier (1,841,384 t), Altitude (1,630,030 t), Google (1,145,199 t) and the City of Stockholm (750,043 t) — total about 11%. For comparison, the macro brief's canonical concentration example is OpenAI and Anthropic taking 43% of H1 2026 venture funding. Durable CDR is nearly twice as concentrated as that, on the demand side.

Where margin actually sits. Not with the technology. The delivered-tonnes leaderboard is entirely biochar, biomass and mineralisation — Exomad Green (416,703 t), Varaha (185,841 t), Carboneers (180,763 t), Aperam BioEnergia (110,751 t), O.C.O. Technology (75,378 t). Meanwhile the contracted leaderboard is led by Vaulted Deep at 5,173,772 t. Margin today sits with low-capex biological routes in low-cost geographies, and with the registries and ratings agencies that arbitrate quality. It does not sit with engineered capture, which has contracts but not revenue.

Barriers to entry. Low for project developers, high for credibility. The ICVCM has approved 44 methodologies as CCP-eligible (2026-08-12) and rejected others outright. A methodology approval is now the effective licence to sell to institutional buyers, and it is issued by a private body.

Who has pricing power. In carbon removal: the buyer, overwhelmingly. In compliance carbon: the regulator. In reporting software: nobody, because the addressable market just contracted by an order of magnitude. In climate risk analytics: increasingly the incumbent financial-data vendors, as the free public baseline is withdrawn.

Market-size figures, with the modeller named.

The ratio that matters: CCUS is 0.23% of clean energy investment. The thing this sector is named after is a rounding error inside the thing it was split away from.

4. Who matters

Leading companies. Microsoft (https://www.microsoft.com/en-us/corporate-responsibility/sustainability) — the anchor buyer for the entire durable removal supply chain. Stockholm Exergi (https://www.stockholmexergi.se/) — BECCS, counterparty to Microsoft's 5.08 Mt permanent removal agreement. Aperam BioEnergia (https://www.aperam.com/) — an incumbent steelmaker's biomass arm outdelivering most venture-backed CDR startups. MSCI (https://www.msci.com/) — consolidating climate risk and carbon market data. Workiva (https://www.workiva.com/) — the listed company most exposed to CSRD scope reduction.

Notable startups. Vaulted Deep (https://www.vaulteddeep.com/) — largest contracted book. Exomad Green (https://exomadgreen.com/) — largest by tonnes delivered, Bolivia. Varaha (https://varaha.earth/) — India, highest delivery ratio among large suppliers. Carboneers (https://carboneers.earth/) — Netherlands. Charm Industrial (https://charmindustrial.com/) — bio-oil injection. Climeworks (https://climeworks.com/) — DAC, the most-covered and least-delivering route. Isometric (https://isometric.com/) — the fastest-moving removal registry. Sylvera (https://www.sylvera.com/) — credit ratings. First Street (https://firststreet.org/) — property-level physical risk, now MSCI-gated.

Active investors. Frontier (https://frontierclimate.com/) — the Stripe/Alphabet/Meta/ Shopify/McKinsey advance market commitment, second-largest buyer. TPG Rise Climate, Breakthrough Energy Ventures, Lowercarbon Capital and Galvanize Climate Solutions are the named category specialists; specific 2026 fund or round figures were not verifiable within this research budget and are not asserted.

Platforms and standards bodies. ICVCM (https://icvcm.org/) — Core Carbon Principles, 44 approved methodologies. Verra (https://verra.org/) — VCS, 2,579+ projects. Gold Standard (https://www.goldstandard.org/) — CORSIA and Article 6 infrastructure. EFRAG (https://www.efrag.org/) — ESRS and the Voluntary Standard. SBTi (https://sciencebasedtargets.org/) — 11,948 validated targets. CDP (https://www.cdp.net/) — 22,100+ disclosing companies. IFRS Foundation / ISSB (https://www.ifrs.org/groups/international-sustainability-standards-board/).

Regulators. European Commission DG FISMA (CSRD/CSDDD), DG TAXUD (CBAM), DG Environment (PPWR). US SEC (https://www.sec.gov/) — withdrew from its own rule. US EPA (https://www.epa.gov/ghgreporting) — proposing to narrow the GHGRP. California Air Resources Board (https://ww2.arb.ca.gov/) — could not be fetched; robots-disallowed.

Research institutions. Berkeley Carbon Trading Project (https://gspp.berkeley.edu/research-and-impact/centers/cepp/projects/berkeley-carbon-trading-project) — the Voluntary Registry Offsets Database, release v2026-06, six registries, free and bulk-downloadable; the best free dataset in the sector. Global Center on Adaptation (https://gca.org/). UNEP (https://www.unep.org/). IEA (https://www.iea.org/).

Trade organisations and civil society. VCMI, the Integrity Council, WWF and WRI (both SBTi partners), and the UN Global Compact. Named individuals verifiable from primary sources within this budget were limited to three — see §13.

5. Products, business models, technologies, customers

Major products. Carbon removal credits (biochar, BiCRS, bio-oil injection, BECCS, DAC, mineralisation, enhanced weathering); avoidance and nature-based credits; CBAM certificates and ETS allowances; ESRS and ISSB reporting software; limited assurance engagements; physical and transition climate risk scores; EPR and packaging compliance services; adaptation engineering and parametric risk transfer; water replenishment and efficiency projects.

How money is actually made today — and how that is changing.

• Carbon removal sells forward tonnes under long-dated offtakes. Revenue recognition follows delivery, and delivery is 3.4% of contracts. The business model is therefore currently project finance against one customer's balance sheet, not a product sale.
• Reporting software sold seats and modules against a mandatory compliance deadline. Directive (EU) 2026/470 removed most of those deadlines. The model is shifting to voluntary-standard tooling at a much lower price point.
• Assurance expected a ratchet from limited to reasonable assurance by October 2028. The directive deleted Article 26a(3), so that fee event is cancelled.
• Climate risk analytics competed on model quality against a free public baseline. As that baseline is withdrawn, the model shifts to data ownership and licensing — the index-and-ratings business model.
• Compliance carbon is the only segment where a customer is legally obliged to pay.

Technologies that matter. Measurement, reporting and verification (MRV) is the real technology of this sector — it is what makes a tonne saleable. Biochar pyrolysis and biomass burial are the routes that actually deliver. Geological storage capacity and injection permitting are the constraint on engineered routes. On the analytics side, asset-level hazard modelling and the emissions-factor databases underneath Scope 3 estimation are the durable assets.

Major customer segments and what they buy on. Hyperscalers buy removal on permanence, durability and reputational defensibility, and one of them is the market. Airlines will buy on CORSIA eligibility, which is a binary. EU importers buy CBAM compliance on embedded-emissions verifiability. Large EU corporates buy reporting tooling on what is still mandatory, which is now much less. Banks, insurers and asset managers buy climate risk analytics on supervisory acceptability, not accuracy.

6. Geography

Regulation concentrates in the EU, overwhelmingly. Every binding rule that moved in this sector in the last eighteen months was European: Directive (EU) 2025/794 (2025-04-17), Directive (EU) 2026/470 (2026-02-26), the revised ESRS (2026-07-03), CBAM's definitive regime (2026-01-01), and PPWR Regulation (EU) 2025/40 becoming applicable (2026-08-12). The US moved in the opposite direction on all three of its levers.

Capital concentrates in the US. Sightline Climate's H1 2026 total is carried by US-centric data-centre transactions, and the largest CDR buyer is American.

Delivery concentrates in the Global South. This is the least-remarked geographic fact in the sector. The top five durable CDR suppliers by tonnes delivered are led by Exomad Green in Bolivia (416,703 t) and Varaha in India (185,841 t), with Aperam BioEnergia in Brazil (110,751 t) fourth. The tonnes are produced in Bolivia, India and Brazil; the money and the standards are in Redmond, London and Brussels.

Non-US market covered with a regional source: the European Union. The regional primary sources used here are EUR-Lex (Directive (EU) 2026/470 and Directive (EU) 2025/794), the European Parliament's Legislative Observatory (procedure 2025/0045(COD)), EFRAG (Brussels) and the European Commission's own DG FISMA, DG TAXUD and DG Environment pages. EFRAG's State of Play 2026 assessed 900 assured 2025 sustainability statements prepared under the ESRS — the only hard measure anywhere of how many companies actually completed a full sustainability report under a mandatory regime.

Demand for credits is becoming multipolar via CORSIA and Article 6. Gold Standard (Geneva) recorded the first large-scale CORSIA retirement by a commercial airline on 2026-03-24 and launched a joint Article 6.2 reporting tool with Verra on 2026-08-20. Cross-border transfer requires host-country corresponding adjustments, which makes sovereign decisions in supplier countries the binding constraint on the whole market.

7. Historical trend patterns

Twenty-five years of this sector is a record of mechanisms that were announced, oversold, discredited and replaced, with the replacement repeating the cycle. Specific false positives from this sector, not from adjacent ones:

• The Clean Development Mechanism (2005–2012). The first global offset market. Collapsed when HFC-23 and adipic-acid projects were shown to be generating credits for destroying gases that were being manufactured to be destroyed, and when the EU restricted their use. False positive: "a UN-supervised methodology is sufficient proof of additionality." That belief is being reconstructed right now under Article 6.
• The 2008 cleantech venture wave. Roughly half the capital deployed by US venture firms into cleantech in 2006–2011 was lost. Solyndra, A123, Fisker and KiOR are the named casualties. False positive: "venture timelines fit capital-intensive physical assets." The current sector's Series C concentration — 40% of H1 2026 funding, up from 16% — is the same bet being placed again, later in the stack.
• Carbon neutrality claims (2019–2022). Airlines, energy majors and consumer brands declared neutrality on the back of cheap avoidance credits. Unravelled in 2023 when REDD+ baselines were shown to be systematically overstated. The consequence is visible today: no REDD+-specific methodology appears among the ICVCM's 44 approved.
• ESG fund labelling (2020–2023). Assets flooded into ESG-labelled products, then reversed under SFDR reclassification and greenwashing enforcement. False positive: "a label is a strategy."
• Hydrogen for industrial decarbonisation (2020–2024). Documented as a false positive in sector 05's own dossier: "announcement volume far exceeding docketed projects."
• The 2021–2023 carbon accounting software wave. Funded on the assumption that CSRD would compel tens of thousands of companies to buy. The compulsion has been withdrawn.

The pattern that repeats, stated precisely: this sector reliably builds market infrastructure — registries, standards, software, ratings — ahead of binding demand, on the assumption that regulation will arrive to create the demand. In every prior cycle it eventually did. In 2026, for the first time, it went the other way. That is what makes this year genuinely different from the four prior disappointments, and it is why the ICVCM label, the Article 6 plumbing and the CDR registries — all of which are good infrastructure — should not be assumed to attract the demand that would justify them.

8. What is changing now (as of 2026-09-15)

The macro frame. Per the shared brief, the FOMC held at 3.50–3.75% in July 2026 on a 9–3 vote with three dissents in favour of a hike, and named the AI buildout as an inflation driver. Capital cost is a live constraint, and this sector's core products — forward tonnes, compliance software, adaptation infrastructure — are all long-duration cash flows against no compliance-mandated revenue. That is the worst possible combination in a sticky-inflation, hawkish-lean environment. The 16 September 2026 FOMC decision falls one day after this research date and is unresolved.

The six operative facts of the sector right now.

1. The EU repealed the only binding corporate climate transition plan obligation in the world. Directive (EU) 2026/470, published 2026-02-26, strikes the CSDDD transition plan requirement as "disproportionate", deletes the EU-wide civil liability regime, and raises CSDDD thresholds from €450m/1,000 employees to €1.5bn turnover / 5,000 employees. Procedure 2025/0045(COD) ran from a Commission proposal on 2025-02-26 to Council adoption on 2026-02-24 — twelve months to dismantle what took three years to build.
2. CSRD scope collapsed and the supply chain got a legal shield. Reporting now applies only above €450,000,000 net turnover and 1,000 employees. "Protected undertakings" below 1,000 employees may refuse value-chain requests exceeding the voluntary standard, and contrary contractual provisions "shall not be binding." Wave two moved from FY2025 to FY2027 and wave three from FY2026 to FY2028 under Directive (EU) 2025/794. Sector-specific standards and the reasonable-assurance ratchet were both deleted. The Commission adopted the revised ESRS and a Voluntary Standard on 2026-07-03.
3. The US is dismantling the measurement layer, not just the rules. The SEC ended its defence of the climate disclosure rule on 2025-03-27. EPA proposed on 2025-09-12 to permanently remove GHGRP obligations for 46 source categories from a programme covering ~8,000 facilities — still proposed, not final, as of the page's 2026-07-30 update. EPA's proposed repeal of power-sector GHG standards (2025-06-11) was also still pending at 2026-02-27. NOAA's Billion-Dollar Disasters shows no release after 2025-06-12.
4. Carbon pricing went the other way. CBAM's definitive regime took effect 2026-01-01 across cement, iron and steel, aluminium, fertilisers, electricity and hydrogen, with real certificate purchase and surrender. The World Bank counts 47 carbon taxes, 40 ETSs and 34 crediting mechanisms in 2026. Disclosure regulation is retreating; pricing regulation is advancing. Conflating them is the commonest error.
5. Carbon removal has contracts, not deliveries, and one customer. 49,468,711 tonnes sold, 1.68 Mt delivered, 3.4%; USD 12.5bn committed; 858 suppliers, 1,240 purchasers; Microsoft at 75.1% of all tonnes. Everything delivered is biological.
6. Emissions stopped falling. EIA's 2026-09-09 STEO forecasts US energy-related CO2 at 4,821 MMmt in 2026 and 4,816 MMmt in 2027 — a change of about −0.1%. UNEP puts current policies at 2.8 °C and full NDC implementation at 2.3–2.5 °C, noting that methodological updates account for 0.1 °C of the apparent improvement and the US Paris withdrawal cancels another 0.1 °C. Net movement from the entire 2025 pledge round: approximately zero.

Testing the sector-split thesis from the climate side

Sector 05 tested whether climate capital and energy capital are diverging and returned: supported on drivers and policy, contradicted on deployment volume, real divergence arriving ~2029-30. Its synthesis was that the selection criterion has shifted from carbon to firmness, speed and interconnectability, and that anyone reading divergence into 2026 installation data is misreading a lagging indicator.

Verdict from the climate side: the thesis is SUPPORTED, and more strongly than from the energy side — but sector 05's timing is wrong by three to four years, because the two dossiers are reading different clocks. Sector 05 read installed volume, which lags. Climate-side indicators — policy, capital allocation and market classification — lead, and they have already separated. The two findings are consistent, not contradictory.

Evidence for divergence, from the climate side (strong).

• The specifically-climate spending categories are rounding errors and one is shrinking. IEA 2026: total energy investment USD 3.4tn, clean USD 2.2tn, grids USD 550bn and up ~20% y/y, storage >USD 100bn — against CCUS at just over USD 5bn and low-emissions fuels at USD 30bn. CCUS is 0.23% of clean energy investment. And the IEA states plainly that "the overall amounts going to renewable power projects have fallen year-on-year since 2024." Inside "clean energy", money has rotated from generation (climate-coded) to grids, storage and nuclear (reliability-coded).
• Climate venture capital is energy-infrastructure capital wearing a climate label. Sightline Climate, 2026-07-13: H1 2026 at USD 26.1bn, +55% — with deal count down 25% to a five-year low. Built Environment rose more than 800% on data centres and overtook Energy as the largest vertical; data centres were ~34% of all investment, with DayOne (USD 4.5bn) and NScale (USD 2bn) carrying much of it. Top ten deals took 42%. Series C nearly quadrupled to USD 10.5bn and took 40% of funding, up from 16%. And in the same dataset, "carbon equity funding collapsed." Strip the data centres out and the total is roughly USD 17bn on 25% fewer deals — flat to down. Within one dataset, the energy/compute leg is up 800% and the carbon leg has collapsed. That is the divergence, measured, in 2026.
• Policy has separated in direction, not just degree. In the same twelve months, the EU repealed the transition-plan mandate, cut CSRD scope tenfold, deleted sector standards and deleted the assurance ratchet; the US abandoned its disclosure rule and proposed to strip 46 categories from mandatory emissions reporting. Over the identical period, sector 05 documents FERC Section 206 orders to all six RTOs, state large-load tariffs, Texas SB 6, the PUCO 85% minimum-take tariff and a USD 1.4tn utility capex plan. Energy regulation is being written. Climate regulation is being unwritten. These are not two speeds of the same process; they are opposite signs.
• The remaining climate demand is voluntary, discretionary and is one company. Microsoft at 75.1% of all durable CDR tonnes is a more extreme concentration than the 43% that OpenAI and Anthropic took of H1 2026 venture funding. Energy demand is contracted load with regulated cost recovery. Climate demand is a line item in one firm's budget.
• And it does not deliver. 3.4% of contracted tonnes. USD 12.5bn committed, 1.68 Mt delivered. Energy's problem is turbines sold out to 2030 and 144-week transformer lead times — a delivery slot problem against real orders. Climate's problem is that the orders do not convert into anything.

Evidence against divergence, or qualifying it (also real, and it matters).

• Carbon pricing is advancing. CBAM's definitive regime from 2026-01-01 is a new climate regulation with genuine financial bite, imposed in the same year the EU gutted CSRD; 47 carbon taxes and 40 ETSs are in operation. So "climate policy is retreating" is too simple. Disclosure policy is retreating; pricing policy is advancing. This is the sharpest correction this dossier can offer to the thesis as usually stated.
• "Climate capital" is not one pool and the three pools are moving differently. Mitigation capital is being absorbed into energy capital. Carbon market capital has collapsed. Adaptation capital is stagnant and, in the one measured series, falling in absolute terms — USD 26bn in 2023 against USD 28bn in 2022, against a need of USD 310–365bn by 2035. Adaptation has genuinely diverged from energy, but downward, and for a different reason: it produces avoided losses rather than a sellable unit.
• The reporting infrastructure was built and has not been demolished. 900 assured ESRS statements exist for FY2025; EFRAG is extending a narrowed regime to non-EU undertakings via ESRS-40a. The regime was narrowed, not abolished.
• Two of the three US rollbacks are still only proposed. The GHGRP narrowing and the power-plant standards repeal were both unfinalised at their pages' latest updates. A court or an election reverses them. Sector 05's own warning about regulatory whiplash applies symmetrically here.

Synthesis. Energy and climate are now different businesses with different buyers, different regulators and different risk — sector 05 was right about that. But the reason is sharper from this side. Sector 05 framed the new axis as "dispatchable and deliverable vs neither." From the climate side the axis is "has an obliged buyer vs does not." Electricity has obliged buyers: load must be served, and cost recovery is regulated. Compliance carbon has obliged buyers: CBAM importers and ETS participants must surrender certificates. Everything else in this sector — durable removal, voluntary credits, disclosure software, transition planning, adaptation — has no obliged buyer at all, and in 2026 the legal scaffolding that was going to create one was removed. That is why climate capital is separating from energy capital now rather than in 2029-30: it is not waiting on a safe-harboured pipeline to run out, it is responding to the disappearance of a future mandatory customer. Policy and capital allocation are leading indicators; installed megawatts are a lagging one. Sector 05 read the lagging indicator and correctly said "not yet." We read the leading ones and they say "already."

One thing that would falsify this reading. If CBAM scope widens and the EU Carbon Removals Certification Framework creates a compliance-grade removal unit, durable CDR acquires an obliged buyer and the divergence closes fast from the climate side. Watch for a CRCF delegated act and any CBAM scope extension before end-2027.

9. The five lists

Five most important current trends

1. EU repeal of the mandatory corporate climate transition plan (T-20-01)
2. CSRD scope collapse and the value-chain information shield (T-20-02)
3. Dismantling of US public emissions and climate-disclosure data infrastructure (T-20-03)
4. The durable carbon removal delivery gap — 3.4% of purchased tonnes (T-20-04)
5. Carbon pricing advancing while carbon disclosure retreats (T-20-06)

Five fastest-growing signals

1. CCP labelling splitting the carbon market into approved and unapproved tiers (T-20-09)
2. CORSIA becoming the first real compliance floor under the voluntary market (T-20-10)
3. Article 6 infrastructure moving from negotiation to reporting plumbing (T-20-11)
4. Climate risk analytics consolidating into financial-data incumbents (T-20-12)
5. EU circular-economy rules becoming operative — PPWR applicable 2026-08-12 (T-20-15)

Five trends most likely to affect businesses

1. CSRD scope collapse and the value-chain shield (T-20-02) — overlaps list 1; suppliers below 1,000 employees can now lawfully refuse questionnaires
2. Carbon pricing advancing — CBAM definitive regime (T-20-06) — overlaps list 1; this is a cash obligation on every importer of six goods into the EU
3. The assurance escalation cancelled (T-20-16) — a withdrawn Big-Four revenue pool
4. Climate tech funding reclassified as data-centre infrastructure (T-20-07) — affects every fundraising benchmark in the sector
5. Dismantling of US emissions data infrastructure (T-20-03) — overlaps list 1; degrades every commercial climate analytics product silently

Five trends most likely to affect consumers

1. EU circular-economy rules becoming operative (T-20-15) — overlaps list 2; packaging formats, deposit schemes and product design are the consumer-visible surface
2. Adaptation finance falling while need compounds (T-20-08) — reaches consumers as uninsurable homes and unprotected infrastructure, with US coastal premiums already up 25%+ (GAO, 2026-02-27)
3. Corporate net-zero target counts as a proxy for decarbonisation (T-20-20) — consumers are the audience for claims that measure intent, not outcome
4. Climate risk analytics consolidating (T-20-12) — overlaps list 2; determines whether a household can get a mortgage or cover in a high-hazard area
5. Dismantling of US emissions and disaster data (T-20-03) — overlaps lists 1 and 3; removes the free information households and local governments use to assess their own risk

Acknowledged overlaps. T-20-03 appears on three lists; T-20-02, T-20-06, T-20-12 and T-20-15 each appear on two. The recurrence of T-20-03 is a finding, not padding: the withdrawal of public measurement is simultaneously a regulatory event, a commercial opportunity for data vendors, and a consumer-information loss. The sector's most consequential 2026 changes are all subtractions, which is why the same subtractions surface in several places.

10. Overhyped / overlooked / cooling / reversing

Most overhyped

Direct air capture at climate-relevant scale. The evidence is precise. No DAC company appears among the top five durable CDR suppliers by tonnes delivered — all five are biochar, biomass or mineralisation. Total durable CDR delivered is 1.68 Mt against 49.47 Mt contracted. Global CCUS investment — which includes point-source capture, not just DAC — was just over USD 5bn in 2025, against USD 2.2tn of clean energy investment. For scale: US energy-related CO2 alone is forecast at 4,821 million tonnes in 2026. The entire durable removal market has delivered roughly 0.035% of one year of US energy emissions. The tell is the same one sector 05 identified for SMRs: capital and attention are present, delivered revenue is not. DAC may matter in the 2040s. It does not matter this decade, and capital priced on a nearer timetable is mispriced.

Corporate net-zero target counts as a measure of progress. SBTi records 14,174 companies with targets or commitments and 11,948 validated; CDP records 22,100+ companies disclosing in 2025. Over the same period UNEP finds current policies still point to 2.8 °C, with methodological updates accounting for 0.1 °C of the apparent improvement and the US Paris withdrawal cancelling another 0.1 °C — net movement approximately zero — while EIA forecasts US emissions essentially flat at 4,821 then 4,816 MMmt. Only 899 of roughly 20,000 CDP-evaluated companies — about 5% — reached the A List. An input metric has been promoted to an outcome metric, and the promotion is now being cited as a reason the mandate was unnecessary.

Most overlooked

The delivered-tonnes leaderboard is entirely Global South biological routes, and nobody discusses it. Exomad Green (Bolivia, 416,703 t), Varaha (India, 185,841 t), Carboneers (Netherlands, 180,763 t), Aperam BioEnergia (Brazil, 110,751 t), O.C.O. Technology (UK, 75,378 t). The most successful carbon removal company in the world by the only metric that counts is a Bolivian biochar producer. Attention has missed it because biochar has no technological glamour, no US venture narrative and no marginal-cost-curve story — and because the sector's media gravity follows contracted volume, where Vaulted Deep's 5.17 Mt sits at the top of a different leaderboard entirely.

The value-chain shield in Directive (EU) 2026/470 is a bigger commercial fact than the scope cut, and almost nobody has priced it. The scope cut removes customers. The shield does something structurally different: it makes a contract term unenforceable. Protected undertakings below 1,000 employees may refuse any value-chain request exceeding the voluntary standard, and "any such provision shall not be binding." Every Scope 3 data strategy that assumed procurement leverage would compel supplier disclosure is now legally unenforceable in the EU. The commentary has focused almost entirely on the employee threshold.

Adaptation is the only climate category where demand is verified, rising and unmet — and it is falling in absolute funding. UNEP: USD 310–365bn needed annually by 2035 against USD 26bn delivered in 2023, down from USD 28bn in 2022, a gap of 12–14x, with the Glasgow doubling goal explicitly stated as unachievable. No venture category, no credit market and no disclosure regime addresses it, because adaptation produces avoided losses rather than a sellable unit. It is overlooked for a structural reason, which is why it will stay overlooked until insurers and municipalities — not climate policy — start buying it.

The EPA GHGRP proposal is an under-covered infrastructure risk. If 46 source categories stop reporting, the calibration set under most US climate analytics degrades permanently and the products silently become models rather than measurements. Buyers of those products are not pricing this.

Cooling

The assurance escalation. Indicator that turned: Directive (EU) 2026/470 deleted Article 26a(3) and its October 2028 reasonable-assurance deadline. Limited assurance is now permanent. Assurance requirements essentially never go backwards once legislated; this one did.

Carbon-market equity funding. Indicator that turned: Sightline Climate records carbon equity funding as having collapsed in H1 2026 while total climate tech funding rose 55%. Corroborated by CCUS at 0.23% of clean energy investment and 3.4% CDR delivery. Honest caveat: "collapsed" is the tracker's characterisation and no numeric figure was retrievable — the direction is well-supported, the magnitude is not.

Target-setting rigour. Indicator that turned: SBTi's Corporate Net-Zero Standard v2.0 (2026-06-11) offers "a suite of options to reflect business reality", and SBTi Services launched a paid Readiness Assessment Service five weeks later. Participation is not cooling — 11,948 validated targets is a real and growing base. Rigour is cooling.

Reversing

Corporate climate disclosure obligation. Mechanism, already executed: Commission simplification proposal → co-legislator agreement → amending directive. Twelve months from proposal (2025-02-26) to Council adoption (2026-02-24). Note the symmetry with sector 05's finding: political risk now moves faster than project risk, in both sectors and in both directions.

Trends that may reverse, and the mechanism by which they would.

• The disclosure rollback reverses via litigation or an election. Two of the three US actions are still only proposed; Iowa v. SEC is live; California SB 253/261 operates independently of federal policy. A single adverse ruling or a change of administration restores a mandatory federal floor.
• The carbon funding collapse reverses via an obliged buyer appearing. The mechanism is an EU Carbon Removals Certification Framework delegated act creating a compliance-grade removal unit, or CORSIA eligible-unit scarcity bidding prices up. Early indicator: a second large-scale airline CORSIA retirement, or any CBAM scope extension.
• The CDR delivery gap reverses via construction, not contracting. Early indicator: the first BECCS or DAC entrant into CDR.fyi's top five by tonnes delivered. Stockholm Exergi is the most likely candidate on Microsoft's 5.08 Mt agreement.
• The value-chain shield reverses via commercial rather than legal pressure — banks and large customers demanding more than the voluntary standard as a condition of business, which the directive does not prevent.
• Adaptation underfunding reverses via insurance withdrawal forcing public action, not via climate policy. Sector 07 records 305–331 day rate-approval delays in Colorado and California against 25%+ coastal premium increases — a configuration that produces political intervention rather than market clearing.

11. Risks and major uncertainties

Sector-specific risks.

1. Single-buyer withdrawal. Microsoft is 75.1% of durable CDR demand. Its AI capex cycle is now the dominant exogenous variable for carbon removal prices. Almost no supplier business plan models this, and the FOMC has explicitly flagged financial-stability risk from leveraged AI infrastructure financing.
2. Supplier failure cascading into the forward curve. 96.6% of contracted tonnes are undelivered. A high-profile insolvency would reprice every outstanding offtake and would be read as a verdict on the category rather than on one company.
3. Silent model degradation. If the EPA GHGRP narrowing is finalised, US climate analytics products keep producing numbers while their empirical basis erodes. The failure mode is invisible to buyers, which makes it worse than an outage.
4. Regulatory whiplash in both directions. The EU dismantled a three-year build in twelve months; the US has two major rollbacks still only proposed. Firms that stood down on the EU rollback and firms that stood firm on the US proposals can both be wrong.
5. Integrity relapse. The CCP label is a private certification with no measured price premium that this research could verify. If a CCP-approved methodology is later discredited, the sector loses its second credibility mechanism in three years.
6. Data concentration. Climate risk analytics consolidating into two or three financial data vendors means one methodology change could move asset valuations across a market, with no independent check.
7. Adaptation's missing unit of account. Without a sellable unit, no market can form, and the 12–14x gap persists regardless of how obvious the need becomes.

Genuine unknowns — and the distinction matters.

Things we do not know but could find out: the share of voluntary-market retirements that are CCP-labelled and whether a price premium exists (MSCI, Sylvera and Abatable have this and sell it); actual VCM issuance and retirement volumes for 2025-26 (in the Berkeley database's Excel files, requiring a bulk download this research could not perform); the count of SBTi commitments removed for failure to validate within 24 months (derivable from SBTi's own weekly export); CDP's 2024 disclosure figure, needed to establish whether participation is rising or falling; the post-OBBBA treatment of Section 45Q; California SB 253/261 deadlines and covered-company counts; ISSB jurisdictional adoption counts; PPWR's numerical recycled-content and reuse targets. Every one of these is a retrieval problem, not an epistemic one.

Things nobody can know: whether Microsoft sustains CDR purchasing through an AI capex downturn; whether host countries will authorise Article 6 corresponding adjustments at volume when doing so reduces their own NDC headroom; whether the EU's simplification agenda stops at the Environmental Omnibus or continues into CBAM and the PPWR; the November 2026 and 2028 US election outcomes that determine whether the EPA and SEC rollbacks are finalised or reversed; whether a physical climate event large enough to reset the political economy of adaptation occurs within the forecast period.

12. Scenarios to 2030

Base — "Voluntary equilibrium." The EU regime settles at roughly 1,000-employee scope with limited assurance and a voluntary standard as the supply-chain ceiling. CBAM operates and widens slowly. Durable CDR delivery improves from 3.4% toward 15-20% as biochar scales and the first BECCS plants commission, but Microsoft remains over half of demand. Carbon equity funding does not recover; consolidation takes out several ratings and marketplace firms. Adaptation finance stays flat in real terms. Falsifiable early indicator: CDR.fyi's delivered total passing 4 Mt before end-2027 without a new buyer above 3 Mt cumulative.

Upside — "An obliged buyer appears." The EU Carbon Removals Certification Framework creates a compliance-grade removal unit and CBAM scope extends downstream. CORSIA eligible units tighten and airlines bid. Durable removal acquires regulated demand, prices firm, offtakes become bankable and project finance flows. Carbon funding reverses within four quarters of the first compliance purchase. Indicator: a CRCF delegated act adopted before end-2027 and at least two further large-scale airline CORSIA retirements. Two is a market; one is a press release.

Downside — "Demand evaporates with the AI cycle." Hyperscaler capex retrenches, and voluntary climate budgets are cut first because they are discretionary and invisible to customers. Microsoft's purchasing slows; with no second buyer at scale, forward prices collapse and multiple suppliers fail before delivering contracted tonnes. Reporting software consolidates hard on a tenth of its projected market. Indicator: two consecutive quarters in which CDR.fyi records aggregate new purchases below the prior-year quarter, combined with any publicly disclosed CDR offtake cancellation above 100,000 tonnes.

Disruption — "Adaptation becomes the sector." Insurance withdrawal from high-hazard geographies forces municipal and national action, and adaptation acquires a financing mechanism — resilience bonds with measurable avoided-loss triggers, or parametric structures at municipal scale. The sector's centre of gravity moves from tonnes to avoided losses, and from voluntary corporate budgets to public infrastructure procurement. Indicator: the first sovereign or large-municipal resilience bond above USD 1bn with an avoided-loss-linked coupon, or international public adaptation finance rising above USD 30bn in any single year.

Regulatory — "Simplification continues into pricing." The Environmental Omnibus precedent extends: CBAM's de minimis widens, PPWR targets soften, and the EU's climate architecture converges on pricing-only with minimal disclosure. The sector becomes a compliance-services business attached to a customs regime. Indicator: any legislative proposal narrowing CBAM scope or raising its de minimis threshold before end-2027.

Failure — "The second integrity collapse." A CCP-approved methodology — most plausibly in biochar permanence or in an Article 6 corresponding-adjustment failure producing double-counting — is discredited. Having already survived the 2023 REDD+ collapse, the market does not survive a second one with its own certifier implicated. Corporate buyers exit rather than re-diligence, and carbon markets cease to be a viable climate-finance mechanism for the remainder of the decade. Indicator: any ICVCM withdrawal of a previously granted CCP approval, or a documented double-counting failure under Article 6.2.

13. Data gaps and limitations

Search capability was degraded and this is the primary limitation. The shared WebSearch budget (200/200) was exhausted before this sector began. One search was attempted and returned a budget message; all subsequent research was direct WebFetch on primary-source URLs. Tier-A verification of what was retrieved is strong — EUR-Lex directive text, the EP Legislative Observatory, EFRAG, SEC, EPA, EIA, IEA, UNEP, World Bank, ICVCM, SBTi, CDP. But discovery was narrow: this dossier found what it knew to look for. Recommend re-run at full search budget, consistent with the treatment of sectors 13–18.

What could not be verified — specifically.

1. Voluntary carbon market volume and price. The single largest gap, and it was a stated priority. Ecosystem Marketplace's SOVCM landing page carries no figures; the Berkeley database publishes statistics only inside downloadable Excel files; registry.verra.org is a JavaScript application that returns nothing to a fetch; MSCI's price index is paywalled. No VCM transaction volume, market value or average price for 2025 or 2026 appears in this dossier, and none should be inferred from it. Phase 2 must ingest the Berkeley workbook.
2. Carbon vertical venture funding magnitude. Sightline Climate says carbon equity funding "collapsed" without a published figure, and no second tracker (PitchBook, Net Zero Insights, Cleantech Group) could be reached. T-20-07 and T-20-17 are marked single_source with evidence_quality capped accordingly.
3. California SB 253 and SB 261. ww2.arb.ca.gov is robots-disallowed and could not be fetched. Deadlines, revenue thresholds, covered-company counts, regulation status and litigation are entirely absent. This is now the most significant mandatory US corporate climate disclosure requirement and this dossier says nothing about it. Highest-priority access gap; recorded in the source registry as S-20-25.
4. Section 45Q after OBBBA. Whether the 2025 tax law preserved, enhanced or changed the carbon capture credit — including any storage/utilisation parity and the commence-construction deadline — could not be verified. IRS's 45Q page 404'd, Treasury's press index showed nothing relevant, CRS reports and Norton Rose are robots-disallowed. Given that OBBBA terminated the wind and solar credits on 2026-07-04 (verified in sector 05), whether 45Q survived is materially important to the divergence thesis and is stated nowhere in this dossier as fact.
5. ESRS datapoint reduction. The revised ESRS adopted 2026-07-03 reduced datapoints substantially, but the before/after counts and percentage could not be sourced — the Commission press release ID guessed at was a different release, and EFRAG's deep news URLs 404. The qualitative changes (sector standards deleted, reasonable assurance deleted) are directly verified from the directive text; the quantitative reduction is not.
6. PPWR numerical targets. Recycled-content and reuse percentages are in the regulation's annexes, not on the Commission's topic page. Not retrieved, not guessed.
7. ISSB jurisdictional adoption. Neither the IFRS Navigator nor the use-by-jurisdiction page returned counts, market-cap share or emissions share. The IFRS entity record is marked confidence: low for this reason.
8. CDP year-on-year. 22,100+ companies disclosed in 2025, but the 2024 comparator could not be obtained (cdp.net/en/insights returned an empty landing page; the historical press release 404'd). Whether CDP participation is growing or shrinking is unknown, and this dossier deliberately does not assert a direction.
9. NOAA Billion-Dollar Disasters status. The citation page shows a 2025 dataset published 2025-06-12 with nothing since; an older NCEI article still describes the product as operational with 2021 data. Whether it was formally retired is not confirmed — treated as a signal, not a fact, in T-20-03 and T-20-12.
10. First Street / MSCI relationship. Inferred from an access gateway presenting MSCI terms of use. No transaction announcement was retrieved. Recorded as signal, entity confidence low, and flagged inside the trend's own contradictions array.
11. COP30 decision content. unfccc.int/cop30 is a directory page. The Belém Political Package exists and the conference closed 2025-11-22; the substantive decisions on the Global Goal on Adaptation, the Baku-to-Belém USD 1.3tn roadmap and Article 6 were not retrieved. Sizeable gap for a sector where multilateral outcomes matter.
12. Water. The thinnest-covered in-scope subcategory. unwater.org returned HTTP 403. The only verified water datapoint in this dossier is Microsoft's 14.2 million m³ FY25 replenishment. No trend record is dedicated to water, which is an honest reflection of the evidence rather than a judgement that water does not matter.
13. Named individuals. Only three people were verifiable from primary sources within this budget (Mark T. Uyeda from SEC Press Release 2025-58; Ede Jorge Ijjasz-Vasquez and Jamal Saghir from the GCA reports page). This meets the contract's 3–6 minimum but is at the floor, and it is a direct consequence of losing search. Uyeda's current role is unverified and the record says so.
14. World Bank State and Trends of Carbon Pricing 2026 returned HTTP 403 through the research proxy; only the dashboard instrument counts (47/40/34) were obtained. Global carbon pricing revenue and emissions-coverage share are missing.
15. Investor figures. No 2026 fund sizes or round amounts for Breakthrough Energy Ventures, Lowercarbon Capital, TPG Rise Climate or Galvanize could be verified. They are named as active participants with unverified funding rather than given plausible numbers.

Conflicting figures recorded rather than resolved. EPA's own pages contradict each other on the status of the power-sector GHG repeal (programme page updated 2026-02-27 shows a pending final action; the climate-regulatory-actions index, last updated 2025-09-25, lists no 2025-26 actions at all). The macro brief's core CPI 2.4% versus core PCE ~3.3% discrepancy is inherited and unresolved. Both are recorded in trend contradictions arrays.

Blocked or unusable sources, for the Phase 2 access budget. ww2.arb.ca.gov (robots-disallowed); eur-lex.europa.eu/search.html (robots-disallowed — direct legal-content URLs work); crsreports.congress.gov and nortonrosefulbright.com (robots-disallowed); unwater.org and openknowledge.worldbank.org (HTTP 403); registry.verra.org and finance.ec.europa.eu/publications/* (JavaScript-rendered or 404); MSCI carbon markets and First Street research (commercially gated). Direct curl through the research proxy failed entirely (exit 56) while WebFetch succeeded, so Phase 2 tooling should not assume shell-level HTTP access.

14. Ranking scorecard

15. Sources

1. "Directive (EU) 2026/470 amending Directives 2006/43/EC, 2013/34/EU, (EU) 2022/2464 and (EU) 2024/1760 (Omnibus I)" — Official Journal of the European Union (EUR-Lex) — https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=OJ:L_202600470 — 2026-02-26 — Tier A
2. "Directive (EU) 2025/794 of 14 April 2025 amending Directives (EU) 2022/2464 and (EU) 2024/1760 ('stop-the-clock')" — Official Journal of the European Union (EUR-Lex) — https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=OJ:L_202500794 — 2025-04-17 — Tier A
3. "Procedure file 2025/0045(COD): Certain corporate sustainability reporting and due diligence requirements (Omnibus I)" — European Parliament Legislative Observatory — https://oeil.europarl.europa.eu/oeil/en/procedure-file?reference=2025%2F0045%28COD%29 — 2026-02-26 — Tier A
4. "ESRS Knowledge Hub, news and State of Play 2026 (2026 Revised ESRS and Voluntary Standard as adopted by the European Commission on 3 July 2026)" — EFRAG — https://www.efrag.org — 2026-09-11 — Tier A
5. "Corporate sustainability reporting" — European Commission, DG FISMA — https://finance.ec.europa.eu/capital-markets-union-and-financial-markets/company-reporting-and-auditing/company-reporting/corporate-sustainability-reporting_en — 2025-12-09 — Tier A
6. "Carbon Border Adjustment Mechanism" — European Commission, DG TAXUD — https://taxation-customs.ec.europa.eu/carbon-border-adjustment-mechanism_en — 2026-01-01 — Tier A
7. "Packaging and packaging waste — Regulation (EU) 2025/40" — European Commission, DG Environment — https://environment.ec.europa.eu/topics/waste-and-recycling/packaging-waste_en — 2026-08-12 — Tier A
8. "SEC Votes to End Defense of Climate Disclosure Rules" (Press Release 2025-58) — US Securities and Exchange Commission — https://www.sec.gov/newsroom/press-releases/2025-58 — 2025-03-27 — Tier A
9. "Greenhouse Gas Reporting Program — proposed reconsideration" — US Environmental Protection Agency — https://www.epa.gov/ghgreporting — 2026-07-30 — Tier A
10. "Greenhouse Gas Standards and Guidelines for Fossil Fuel-Fired Power Plants" — US Environmental Protection Agency — https://www.epa.gov/stationary-sources-air-pollution/greenhouse-gas-standards-and-guidelines-fossil-fuel-fired-power — 2026-02-27 — Tier A
11. "Short-Term Energy Outlook, September 2026" — US Energy Information Administration — https://www.eia.gov/outlooks/steo/ — 2026-09-09 — Tier A
12. "World Energy Investment 2026" — International Energy Agency — https://www.iea.org/reports/world-energy-investment-2026 (PDF: https://iea.blob.core.windows.net/assets/64594543-cf6e-4fd9-8238-d3cae35daf48/WorldEnergyInvestment2026.pdf) — 2026-05-28 — Tier A
13. "Durable CDR market dashboard and leaderboards" — CDR.fyi — https://www.cdr.fyi/ — 2026-09-15 — Tier B (specialist tracker with a published methodology; used for the sold-versus-delivered distinction that no Tier A source publishes, and flagged for Tier A cross-check in Phase 2)
14. "2026 Environmental Sustainability Report (FY2025 data)" — Microsoft — https://www.microsoft.com/en-us/corporate-responsibility/sustainability/report — 2026-08-13 — Tier A
15. "CCP assessment status — approved and rejected methodologies" — Integrity Council for the Voluntary Carbon Market — https://icvcm.org/assessment-status/ — 2026-08-12 — Tier A
16. "Newsroom" — Gold Standard — https://www.goldstandard.org/news — 2026-08-27 — Tier A
17. "Verified Carbon Standard programme" — Verra — https://verra.org/programs/verified-carbon-standard/ — 2026-09-15 — Tier A (page is undated; undated: true, used only for project counts)
18. "Voluntary Registry Offsets Database, release v2026-06" — Berkeley Carbon Trading Project, UC Berkeley Goldman School of Public Policy — https://gspp.berkeley.edu/research-and-impact/centers/cepp/projects/berkeley-carbon-trading-project/offsets-database — 2026-06-01 — Tier A
19. "Carbon Pricing Dashboard — Carbon Pricing Around the World, 2026" — World Bank — https://carbonpricingdashboard.worldbank.org/ — 2026-09-15 — Tier A
20. "Adaptation Gap Report 2025" — UN Environment Programme — https://www.unep.org/resources/adaptation-gap-report-2025 — 2025-10-29 — Tier A
21. "Emissions Gap Report 2025" — UN Environment Programme — https://www.unep.org/resources/emissions-gap-report-2025 — 2025-11-04 — Tier A
22. "H1'26 climate tech funding up 55% to $26bn, thanks to data centers" — CTVC / Sightline Climate — https://www.ctvc.co/h126-climate-tech-funding-up-55-to-26bn-thanks-to-data-centers/ — 2026-07-13 — Tier B
23. "SBTi releases Corporate Net-Zero Standard Version 2, and 2026 news" — Science Based Targets initiative — https://sciencebasedtargets.org/news — 2026-06-11 — Tier A
24. "Companies taking action — target dashboard" — Science Based Targets initiative — https://sciencebasedtargets.org/companies-taking-action — 2026-09-15 — Tier A
25. "Companies scores — 2025 disclosure cycle" — CDP — https://www.cdp.net/en/companies/companies-scores — 2026-09-15 — Tier A
26. "COP30 Belém — Belém Political Package and conference decisions" — UNFCCC — https://unfccc.int/cop30 — 2025-11-22 — Tier A
27. "Reports — rapid climate risk assessments and investment briefs (DRC, Zambia, Kenya)" — Global Center on Adaptation — https://gca.org/reports/ — 2026-09-07 — Tier A
28. "Research library (access gated behind MSCI terms of use)" — First Street / MSCI — https://firststreet.org/research-library — 2026-09-15 — Tier B (used as a signal of access change only)
29. "Homeowners Insurance: Premiums Generally Tracked Inflation but Rose More in Disaster-Prone Areas" (GAO-26-107867) — US Government Accountability Office — https://www.gao.gov/products/gao-26-107867 — 2026-02-27 — Tier A (via sector 07 dossier)
30. "Sector 05 dossier: Energy and power" — Phase 1 trend-intelligence programme (internal) — file:///home/claude/research/dossiers/05-energy-power.md — 2026-09-15 — Tier A
31. "Macro Context Brief and Addenda 1–2" — Phase 1 trend-intelligence programme (internal) — file:///home/claude/research/frameworks/00-macro-context.md — 2026-09-15 — Tier A
32. "Billion-Dollar Weather and Climate Disasters (citation page; no release after June 2025)" — NOAA National Centers for Environmental Information — https://www.ncei.noaa.gov/access/billions/ — 2025-06-12 — Tier A
33. "Carbon Markets research and price index (paywalled)" — MSCI — https://www.msci.com/carbon-markets — 2026-09-15 — Tier B (free pages stale; Net-Zero Tracker page still dated 2024-08-31 — a live date-laundering example)
34. "Carbon capture, utilisation and storage technology deep dive" — International Energy Agency — https://www.iea.org/energy-system/carbon-capture-utilisation-and-storage — 2026-06-01 — Tier A
35. "Climate Corporate Data Accountability Act programme page" — California Air Resources Board — https://ww2.arb.ca.gov/our-work/programs/climate-corporate-data-accountability-act — not retrievable; robots-disallowed — Tier A source, zero coverage (listed so the gap is auditable)

More research in this sector · Explore trend records