The buffer pool is a shared overdraft no one is reconciling.

Forest carbon offsets make two promises. The first, that a tonne was really avoided or stored, is the one auditors argue about. The second is quieter and harder: that the tonne will stay stored. A certificate can attest to a measurement made on a given day. It cannot promise that the forest will still be standing in fifty years. Permanence is the part the paper cannot guarantee, and the market's answer to that gap is a shared reserve, a buffer pool, that everyone draws on and almost no one is reconciling against the hazard it actually faces.

The promise a certificate cannot make

A forest carbon credit is a claim about the future. Under California's compliance programme, run by the California Air Resources Board, permanent has been interpreted to mean carbon kept out of the atmosphere for a minimum of 100 years. No certificate can underwrite a century. Trees burn, get sick, and are logged, and atmospheric CO2 lasts far longer than any individual stand. The market squares that circle not by promising each project will survive, but by self-insuring the portfolio.

The instrument is the buffer pool. Every project contributes a share of its issued credits to a communal reserve. When a project suffers an unintentional reversal, a fire or disease event outside the landowner's control, credits are retired from the pool to cover the loss, one buffer credit per tonne released. As long as the pool stays solvent, the permanence claim attached to every credit in circulation holds. It is, in effect, a shared overdraft facility. The question is whether anyone is checking that the balance can cover the withdrawals.

How the overdraft is meant to work

The arithmetic is specific. Under the 2015 forest protocol, projects contribute between 8.7 and 19.2 per cent of credited carbon to the buffer pool, split across risk categories. Wildfire risk draws a contribution of just 2 to 4 per cent of credits, lower for projects claiming active fire management. Disease and insects take a fixed 3 per cent. Other catastrophic natural risks, wind, ice and flood, take another 3 per cent. Financial and management risks, including bankruptcy, take 1 to 9 per cent.

As of early January 2022, roughly 31 million credits had been contributed in total, according to the peer-reviewed analysis by Badgley, Chay, Cullenward and colleagues at CarbonPlan, published in Frontiers in Forests and Global Change. About 19 per cent of the pool was earmarked for wildfire. The mechanism only works on two assumptions: that contributions match the real hazard, and that reversals are charged to the pool promptly and in full. Both are failing.

The reserve was drained in under a decade

The wildfire seasons of 2020 and 2021 tested the assumption directly. CarbonPlan combined two reversals already verified by the regulator, which retired over 1.1 million credits, with US Forest Service satellite-derived mortality maps for four further projects hit in 2020 and 2021. Their estimate: wildfire has caused the reversal of between 5.7 and 6.8 million credits. That is between 95 and 114 per cent of the entire wildfire allocation set aside to cover a century of fire risk across the whole programme.

Put plainly, at least 95 per cent of the buffer's hundred-year wildfire cushion was consumed in under ten years. The single largest event was the 2021 Bootleg Fire in southern Oregon, which burned through the Klamath East project and, as CarbonPlan documented, triggered automatic termination and the retirement of at least 1.14 million credits, the second-largest buffer withdrawal in the programme's history. A year earlier, the 2020 Lionshead Fire had torn through the Warm Springs tribal offset project in Oregon; the Confederated Tribes ultimately ended the roughly 25 million dollar project after the burn, with the regulator having issued more than 2.6 million credits against that forest.

A buffer sized for a century of fire, spent in a single decade, is not a reserve. It is an overdraft no one reconciled until the statement arrived.

Disease is the next unpriced liability

Wildfire is the visible drawdown. Disease is the one waiting on the books. CarbonPlan's scenario analysis of sudden oak death, a pathogen mediated by California bay laurel, found that mortality across the programme's tanoak-bearing projects could on its own fully encumber the credits set aside for all disease and insect risk. The disease component is a flat 3 per cent contribution, with, as the authors note, no published actuarial basis. Journalists who interviewed the buffer pool's designers reported the original risk ratings looked like educated guesswork.

This sits alongside a separate finding from the same research group. Their 2021 study in Global Change Biology concluded that around 29 per cent of the credits they analysed were over-credited relative to real climate benefit, roughly 30 million tonnes worth on the order of 410 million dollars. Over-crediting inflates the assets the buffer is meant to protect, while under-pricing of hazard shrinks the reserve protecting them. The two errors compound.

Why this is not California's problem alone

The same structure underpins the voluntary market. Verra's AFOLU Non-Permanence Risk Tool assigns each project a risk rating and a corresponding contribution to a shared, cross-fungible buffer account, cancelling buffer credits when a reversal is verified. The design logic is identical, and so is the failure mode: if contributions are set by rule of thumb rather than observed regional hazard, and if reversals are only charged after slow ground surveys, the pool can be insolvent long before anyone reconciles it. Verra has begun piloting insurance and fund-based alternatives, a tacit acknowledgement that the pooled model alone is under strain.

Fire-return intervals are shortening across the western United States and beyond. A buffer calibrated to twentieth-century hazard is structurally under-capitalised against twenty-first-century fire and disease. The problem is not that self-insurance is wrong in principle. It is that nobody is continuously matching contributions to the hazard that is actually arriving, or charging reversals at the speed they actually happen.