Nepal and the carbon market
How a country that earns almost nothing from emissions became one of Asia’s most credible sellers of forest carbon — and what still stands in the way.
Opening
13 November 2025
9.4US$m
Million US dollars, received. The first results-based forest carbon payment Nepal has ever had, for 1.88 Mt verified in July 2025.
8 December 2025
2082
The Carbon Trading Regulation, gazetted. Nepal’s first legal framework for selling credits, at home and abroad.
23 January 2026
55US$m
Million US dollars, at most. The LEAF agreement: up to 4 Mt at a US$10 floor, contingent on verification. First in Asia.
Before these five months, Nepal spent twenty years in carbon markets and earned almost nothing.
Opening
Seven words carry the rest of the briefing. Only the third has to stay in your head — it is the one everything later turns on.
The corresponding adjustment is Article 6.2 and 6.4 of the Paris Agreement.
Opening
The useful distinction is not who buys. It is what forces the purchase.
| Market | What forces the purchase | What the buyer therefore wants |
|---|---|---|
| Compliance | A law, a cap or a tax | The cheapest compliant tonne |
| Voluntary | Nothing at all | Credibility it can defend in public |
Nepal is now in both: correspondingly adjusted credits for compliance buyers through LEAF, and — since the Regulation of December 2025 — a domestic voluntary market inside Nepal. A law forces the compliance buyer, so it shops on price. Nothing forces the voluntary buyer, so it shops on credibility — which is why an unglamorous, government-run, independently baselined programme is worth more to it than a clever one.
01
01 · How carbon became sellable
Fifty-five years from a property-rights argument to a global framework with a trading article in it.
1960
Coase
the property idea
1982
Lead credits
the first trades
1990
Acid Rain
SO₂ allowances
1992
UNFCCC
the framework
1997
Kyoto
the protocol
2005
EU ETS
the big market
2015
Paris
Article 6
The Acid Rain Program is the proof: allowances traded at about a tenth of the price the EPA had projected, and the programme cost a fraction of what industry forecast. Not the same as saying it cut emissions at a tenth of projected cost — that conflates allowance price with programme cost, and the literature criticises it.
01 · How carbon became sellable
This is the single most important idea in the briefing. Under Kyoto only the buyer had a target, so one ledger did the counting. Under Paris everyone has an NDC — so unless the seller subtracts what the buyer adds, the same tonne is counted twice.
Kyoto, 1997 — only the buyer had a target
Paris, 2015 — every country has a target
That subtraction is the corresponding adjustment. Nepal’s Regulation provides for it in Rule 16 — and LEAF is the first place anywhere it is offered to private buyers.
01 · How carbon became sellable
The mechanism changed, and the identical project now earns less.
Credits for one Myanmar cookstove programme, 2021–22 vintagetCO₂e
February 2026, the first Article 6.4 issuance — the first of about 165 transitioning projects. The CDM bar is implied, not published: it is what the reported 40% gap works back to. Nepal’s own biogas programme left the CDM on 5 July 2026 in the same transition.
01 · How carbon became sellable
The point is the spread, not any one bar — roughly 220-fold, top to bottom.
Mid-2026US$ per tCO₂e · log scale
Each gridline is ten times the last. Nepal sits in the bottom third — and moved from $5 to $10 in fourteen months. EU ETS converted from €82.85/t at ~1.16 USD/EUR, the same basis as the price table in the calculator.
01 · How carbon became sellable
What lost credibility
Project-scale credits with developer-set baselines — the seller choosing the counterfactual it would be judged against.
What buyers moved to
Jurisdictional, government-led programmes on independent baselines. Nepal was already in that format.
2023 investigation, one registry’s REDD+ portfolio: most of the projects it could analyse showed no measurable benefit — and the underlying study drew a formal published rebuttal. This is a claim about what buyers now want, not proof of performance. Jurisdictional programmes carry their own baseline, leakage and reversal risks.
02
Ten years. Everything done right. Almost nothing earned.
02 · The CDM era, 2005–2015
Registered 27 December 2005 as projects 136 and 139 — and then squeezed.
| Biogas credits, BSP-Nepal Activity-1 | Volume | Status |
|---|---|---|
| Expected over the 21-year life | 1,971,543 | Expected, never issued |
| Actually issued, three periods | 223,389 | Issued |
| Requested and refused | 95,468 | Rejected after assessment |
US$7m committed by the World Bank’s Community Development Carbon Fund; run by AEPC across 55 and 57 of Nepal’s 75 districts. An earlier application covering 19,396 plants was refused after the fund had already paid. The PoA registered 31 January 2013 with an initial 20,000 digesters; the route eventually covered more than 140,000.
03
September 2025 to January 2026.
Badikhel Community Forest, Lalitpur · Shadow Ayush / Wikimedia Commons · CC BY-SA 4.0 · responsive crop
03 · Five months that changed the position
Note the order. The Benefit Sharing Plan had to be approved in October before any money could move in November.
Feb 2021
ERPA
signed, 9 Mt at $5
May 2025
NDC 3.0
−17.12% by 2030
Jul 2025
Verified
1.88 Mt validated
Oct 2025
Plan
benefit sharing
Nov 2025
Paid
US$9.4m received
Dec 2025
Law
Regulation 2082
Jan 2026
LEAF
up to US$55m
In February 2026 the programme was extended by twelve months, to 31 December 2026 — the first extension. The money had arrived too late in the year to spend: the Benefit Sharing Plan was approved on 15 October 2025 and the payment landed on 13 November, leaving under a month before the original 31 December 2025 closing date.
03 · Five months that changed the position
Three numbers, and they must stay distinct.
Against the 2021 ERPAMillion tonnes CO₂e
Disbursed
20.9%
US$9.41m of US$45m. US$35.59m remains.
World Bank rating
Moderately Unsatisfactory
For the achievement of the objective — primarily delays in distributing payments. Not the programme as a whole.
03 · Five months that changed the position
Two failures. Only one of them is documented.
Documented — the disbursement failure
The Bank rates the achievement of the objective Moderately Unsatisfactory, primarily for “delays in distributing ER payments” — and, second, for delays in implementing the programme.
Not documented — why the tonnes were fewer
The volume shortfall it attributes partly to over-estimation in its own programme document — the second reason it gives. Nothing about measurement capacity.
Do not claim the shortfall was caused by weak measurement — the Bank does not say that. The stronger point is that creditable performance and forest performance are different quantities. The first reference level counted only fuelwood extraction as degradation; fire, grazing and felling came after the UNFCCC team’s 2018 recommendation.
03 · Five months that changed the position
“There is no national-level emission reduction assessment executed for the forestry sector to evaluate the mitigation outcome progress against the FRL.”— National REDD+ Strategy 2025–2034, p.44. Verbatim.
Nobody has measured whether Nepal’s forests beat their reference level. This is the purest evidence for the argument the briefing is making.
03 · Five months that changed the position
Both figures are correct. They answer different questions — and only the second one can be sold.
Third National Communication
17.1Mt
What the forest does — net annual sink. All pools, whole country, no counterfactual.
National Forest Reference Level
1.186Mt
What the accounting counts — 929,325 t deforestation plus 408,500 t degradation, less 151,077 t enhancement.
The credit is what you can prove the forest did differently from a counterfactual, in two carbon pools, against a reference level.
03 · Five months that changed the position
The terms
Up to US$55m for up to 4 Mt, at a US$10 floor — double FCPF’s 2021 price. Gandaki, Bagmati and Lumbini, on ART-TREES, through Emergent.
The buyers
The UK and Norway guarantee an initial 1 Mt; the rest goes to corporate buyers. First anywhere to offer them correspondingly adjusted credits.
The risk
Its baseline is 2017–2021, with crediting from 2022. After thirty years of community forestry a recent baseline is a much harder bar than FCPF’s.
Nepal delivered 1.88 Mt against 9 Mt contracted on the easier baseline. LEAF asks for up to 4 Mt against a harder one — which is what “up to US$55m” is really saying. FCPF’s baseline rests on a ten-year period ending around 2010–11.
03 · Five months that changed the position
Community forest user groups
23,601
Within 35,280 community-based groups of every modality.
Hectares they manage
2.51m
Of 3.35m hectares under community-based management overall.
Affiliated households
3.17m
Over three million households, and about 45% forest cover.
Nepal did not build community forestry to sell carbon. It built it for firewood, fodder and floods — and tens of thousands of women have held elected forest-committee positions. The carbon market arrived thirty years late to something that already worked.
04
04 · What still stands in the way
Enacted 8 December 2025 under the Environment Protection Act 2076. What the government takes, or requires:
| Rule | What the government takes, or requires |
|---|---|
| Rule 14 | 5% of verified and issued credits count toward Nepal’s own NDC |
| Rule 15 | NPR 100 per tonne sold, to the government |
| Rule 18 | 10% of private developers’ trading profit; government entities keep theirs |
| Rule 16 | A corresponding adjustment, to prevent double counting |
MoFE is the national authority; unauthorised carbon projects are prohibited (r.34). A Technical and a Steering Committee assess proposals (rr.29–30), and until the National Carbon Registry exists MoFE may use an international registry. Eligible proponents include community forest user groups — which sets up the next problem.
04 · What still stands in the way
An open question, not settled law.
Regulation 2082
Carbon trading runs through MoFE as the national authority. Community forest user groups are eligible proponents in their own right.
Forest Regulation 2079, r.107
Forest carbon needs the user groups’ prior informed consent, price by agreement, and at least 80% of income into the Fund — with r.115 splitting what is in it.
Why both exist
Carbon rights are retained federally; every other environmental service sits with provinces and forest user groups.
As far as I can find, nothing published reconciles the two regimes. If anyone here knows otherwise, please tell me. Do not blur the two 80 percents: r.107(8) sends 80% of income into the Fund; r.115 distributes 80% of what is in the Fund to the user group. Nobody has read the gazetted Nepali text.
04 · What still stands in the way
| Article 6 readiness | Status |
|---|---|
| Designated national authority | MoFE, designated |
| Authorization arrangements | In place |
| Tracking arrangements | In progress |
| Reporting | None |
Nepal is in the 86 countries with arrangements under way — not the 14 with both in place, not the 20 that have reported. The bilateral-agreements row is an inference: the page has no such row, and Nepal is absent from every counterparty list. Singapore has eleven full Implementation Agreements, about 28 arrangements overall.
04 · What still stands in the way
Illustrative, not an official allocation: the 80% is a target still being negotiated, a further ~10% may go in operational costs, and provincial capacity is uneven. Communities are pushing for the full 80%. What is left is plumbing — and plumbing is solvable.
05
Ten minutes on our own footprint.
05 · Now: what about us?
Nepal’s power is about fifteen times cleaner than India’s. Fuel and flights are where the footprint is.
Default grid emission factorsgCO₂ per kWh
43 is a floor value, assigned identically to eleven other countries — Albania, Andorra, Bhutan, DR Congo, Iceland, Lesotho, Malawi, Monaco, Paraguay, San Marino and Eswatini. It is a default for hydro-dominated grids, not a measurement of Nepal’s, and the dataset is from July 2019. 673 ÷ 43 = 15.7.
The last thing left unresolved is what a Nepali kilowatt-hour actually costs in carbon. The calculator below is that argument applied to this office: every factor Nepali where a Nepali factor exists, every one editable, every one linked to the document it came from.
Step one
Your month: electricity units, cooking fuel, commute, any flight this year.
Step two
IPCC for fuels, DEFRA for flights, Nepali survey data for everything on the road. The grid factor is the argument.
Step three
Compare the spread in this room against Nepal’s average of about 1.06 t a year — a 2017 figure on a 2012–2017 trend, printed in 2020.
The fuel factors are IPCC 100%-mineral values, not blended headline figures — Nepal does not blend, and the blended ones would understate the answer by about a tenth.
The calculator
A GHG Protocol inventory for a Nepali law firm. Every factor is Nepali where a Nepali factor exists, every one is editable, and every one links to the document it came from — including the grid factor, which is the argument.
Fuel the firm burns and gases it leaks, on its own premises and in its own vehicles. Small for a law firm — until you count the generator and the air conditioning.
Same tonne, chemically identical. The price depends entirely on who verified it and under what rules. Cheapest to dearest: 222×.
Nepal and the carbon market
1.88 million tonnes is what thirty years of community forestry bought. It became worth US$9.4 million because somebody measured it — in two carbon pools, against a reference level. The law arrived in December. The registry has not.
Open question
Two forest-carbon regimes, one forest. Has anyone reconciled them?
Tap me whenever you want a short prompt for the section you are reading.