Field research
Independent notes on what a loss record would have caught, and why it now lands on a balance sheet rather than in a sustainability report.
The market’s largest insurance reserve is 24.8 per cent consumed, thirteen years into a hundred-year promise. The figures have been public throughout. Nobody had added them up.
The analytics layer is consolidating into underwriting at speed. Every pairing joins a view formed before the fact to the capital carrying the risk. None of them is an outcome record, because the asset they would pair with has never been built.
Roughly 43 million tonnes of engineered removals contracted, about 1.2 per cent verifiably delivered. The number is real, widely misused, and the way it is usually calculated is wrong. What an honest delivery record requires.
Exhibits
Each is drawn from source data and sits inside the piece that argues from it.
The gap
Why this asset class has no record of its own failures, and what that costs the people holding it.
Property catastrophe, credit, structured finance and commodities all built the same infrastructure in the same order, and the order is the lesson. Carbon is attempting it backwards.
0.2the weakest correlation between two vendors on the same assetsthirteen vendors, one hundred propertiesThirteen vendors assessed the same one hundred properties and their damage estimates correlated as weakly as 0.2. None publishes an error rate, because none has anything to measure against. Carbon has the same structure, one step earlier.
4 yearsof carbon insurance written without a loss tablethe market’s own flagship reportCarbon insurance is scaling on capacity and conviction, not loss history. What underwriting carbon actually requires, and where the independent read behind the quote has to come from.
19%of 52 REDD+ projects met their reported emissions targetScience, 2025 · twelve countriesIt was not a rogue project. It was the assurance chain working as designed. When everyone in the chain is paid to move the credit forward, the bias is structural, and methodology alone cannot fix it.
27 against 25consulting fees against audit fees, one client, one yearAndersen and Enron, before Sarbanes-OxleyEvery mature market learned this once, often the hard way. Nature finance still lets the originator grade its own pool. Why independence is the precondition for a price, not a nicety.
$60bnprivate capital committed to nature, cumulativeForest Trends and The Nature Conservancy, 2026Private investment in nature has passed sixty billion dollars. It is still priced, monitored and reported off the developer’s own paperwork. A reading of where the risk infrastructure has to go next.
The balance sheet
What happens when the position stops being a sustainability line and becomes an audited number.
Books full of credits carried at face value, a large share worth a fraction, graded by the people who sold them, and now an accounting rule forcing the reckoning. The analogy to issuer-paid ratings is not loose. It is precise.
FY2028first year the book is measured and tested every reporting dateFASB ASU 2026-02FASB ASU 2026-02 turns the voluntary carbon book into an audited, impairing balance-sheet item. The controller and the external auditor now own a market they have never had to value.
3 inputsa broker quote, a rating, a registry serial. None is a markwhat fair value actually requiresA finance-native walk through what fair value and impairment actually require for a carbon book under the new accounting, and why a broker quote, a registry serial and a project rating are each insufficient as the audit-grade input.
the sellerwhere the mark on most carbon books ultimately comes fromask, and that is the answerFunds, desks and treasuries carry carbon at numbers somebody chose. Ask where the mark came from, and the answer is usually the seller. The case for an independent valuation opinion, before someone else asks the question.
0underwriter models that can validate their own book to a reinsurerindependence is the whole of the pointCarbon underwriters are building and buying good analytics. None of it answers the question a reinsurer, a capital provider or an auditor actually asks, and the moment a book gets sold on is arriving now.
$6.2tnof labelled sustainable debt governed by covenantsalmost none of it linked to a monitored tonneThere is 6.2 trillion dollars of labelled sustainable debt governed by covenants, ratchets and performance targets. Almost none of it links a single payment to the delivered, monitored performance of the carbon credits a company actually buys. That is the gap, and project finance already built the tools to close it.
The regulation
Eligibility, standards and the dates on which a standing position reprices.
Three ways a carbon position loses value without anything happening to the tonne, six dated decisions in thirty months, and why holders find out at the same time as the public.
13%of legacy credits carry host country approvalUNFCCC · c.128 Mt of c.980 Mt requestingThe CDM to Article 6.4 transition is the largest eligibility event the carbon market has produced. Around 13 per cent have host country approval, the two largest historical suppliers are absent, and nobody is keeping the score in public.
10,000+companies holding validated science-based targetsSBTi · over 40 per cent of global market capitalisationSBTi V2.0 cuts avoidance and REDD+ credits out of target accounting and pushes the obligation into expensive, long-dated durable removals. The legacy book and the forward removal programme are now two separate, unmanaged exposures.
$1bnend-user spend, roughly flat while volume fellmarket reporting, 2025The Integrity Council's Core Carbon Principles and the CORSIA eligibility regime function as binary sorting machines. A methodology or a vintage is in, or it is out, and when the line moves it can reprice a whole cohort at once. There is no watchlist, no gradual downgrade and no advance notice. A book that holds certificates rather than a live read on risk finds out last.
1 rulingcan reprice an entire category at oncethe carbon market’s version of a ratings migrationReclassification is the carbon market’s version of a ratings migration. A methodology revision or a council ruling can move a whole category at once. The exposed capital should see it coming, not read about it after.
27 Sep 2026offset-based product claims banned across the EUEmpowering Consumers DirectiveFrom 27 September 2026 the European Union bans offset-based product claims outright. The Green Claims Directive that was meant to soften the regime has been withdrawn. The result is a hard prohibition with no safe harbour, fines reaching four per cent of turnover, and a litigation record that is already winning. The carbon book bought to support a claim has become a liability the general counsel now owns.
35 yearsthe minimum storage life recognised in productsEU Carbon Removal Certification FrameworkThe Carbon Removal Certification Framework pulls durable removals inside regulation. Regulated assets need an independent, continuous risk read. That infrastructure does not yet exist.
3xthe strongest credits against the weakesttraded ranges, late 2025The strongest credits already trade at roughly three times the weakest. The spread is widening, and it is invisible to any book that holds a certificate instead of a risk read.
The method
How the evidence is assembled, and the disciplines that decide what may be published.
The same eligibility loss has been published as 32 per cent, a third, and 30.6 per cent. All three are correct. The convention that fixes it is small, unglamorous, and the reason a figure can be cited rather than merely quoted.
3% to 5%the discount rate move that flips the sign of the answerflood barrier appraisal, New YorkEstimable, attributable, decision-linked and inside your mandate, all four at once. Fail one and the honest output is a direction, labelled as such. Why we apply the test to ourselves.
96%of mangrove regeneration was spontaneous, not plantedForest Ecology and Management, 2020Verra’s ARR methodology moved baselines from assumption to observation, and proved a forest can be read from orbit. It also drew, more clearly than any document before it, the line where methodology ends and risk begins.
2 to 4%the wildfire contribution to California’s buffer poolCARB forest protocol 2015 · total 8.7 to 19.2%Forty years of monitoring, pooled reserves, and a reversal definition still being standardised. As fire-return intervals shorten, the question is whether the buffer is sized to the risk it actually carries.
168,000credits issued for carbon the trees never capturedsatellite reassessment · of close to 500,000 issuedThe world’s largest mangrove restoration sold close to half a million credits. A peer-reviewed satellite reassessment found roughly 168,000 of them were for carbon that never existed. Every audit had passed.
Send us a slice of what you hold, or one submission you are about to price, and we will return an independent read from public data.
The register stays free whether you engage or not.
The first edition publishes in October. Leave your details and we will write to you the day it does.
Thank you. You are on the list, and we will write when the register publishes.
Independent reads open when the register launches in October. Join the waitlist and we will tell you the day it does.