International credits, shut out of the EU’s carbon market since 2020, are back. Under Article 6 of the Paris Agreement, they allow a country to count emissions cuts made abroad towards its own target, while channelling finance into decarbonisation elsewhere. The EU’s amended Climate Law allows up to 5 % of the 2040 target to be met through credits, with at least 85 % of the reduction still having to be domestic.
The European Commission has given itself six years to build an international credit market that barely exists. Its July proposal on revising the European Emissions Trading System set aside EU allowances to buy international credits from 2036 – but only if a review in January 2033 finds enough high-quality, high-integrity, cost-effective credits on the market to justify it.
What the proposal doesn’t set out is who supplies those credits, their quality and how to ensure there’s enough of them in time. Building such a market takes five to ten years, meaning it likely won’t happen before 2033. The Commission shouldn’t wait to start deciding what happens next.
The EU would buy credits collectively, rather than hand operators the right to purchase them directly, loosening the EU’s domestic target – if the market can support it. Relief would come through the cap. A central facility will buy up to 260 million tonnes of international credits between 2036 and 2040, worth roughly 2 % of 1990 EU net emissions. This would avoid a race towards the cheapest, less scrutinised supply.
A pilot phase without a plan?
A pilot phase, running from 2031 to 2035, is meant to prepare the ground. The specifics are still to come and the Q4 2026 proposal revising national climate targets and flexibilities is the moment when it should arrive. International credits purchased during the pilot phase won’t count towards the 2040 target, something the EU itself pushed for at COP26.
International credits currently trade well below EU allowances, not because the market is oversupplied but because almost nobody’s buying yet: The entire Paris Agreement Crediting Mechanism (PACM) has issued 1.11 million tonnes since February 2026, against an EU compliance pool that will stay empty until 2036. That’s when the EU’s own facility would start buying.
But it only solves half the problem. Businesses need a reliable cap trajectory to plan investment now – not in 2033. By delaying costlier domestic investment for the 2033 review, they risk a negative verdict that would only leave three years to 2036, which isn’t enough time to find an alternative plan.
In short, the cap gets met through idled output and closures instead of the decarbonisation investment the mechanism was meant to turbocharge.
Supply is still uncertain
The EU hasn’t decided whether to build the supply of international credits through bilateral government-to-government deals or a centralised, UN-run mechanism. As the facility already centralises purchases to avoid the integrity risk of scattered buying, the better design would be a single EU body which vets candidate credits, drawing on already-operating infrastructure without inheriting the need for a bilateral agreement per host country. Whether this is what the Commission intends, upcoming legislation needs to settle this – and settle it fast.
The urgency is arithmetic. Only 28 parties have authorisation arrangements in place and the gap between what exists today and what the mechanism requires is essentially most of the market that still needs to be built. Closing it will mean widening what counts, not only building more of what already exists: transition credits linked to retiring coal plants earlier than planned are one option to scale supply, helping to finance a just transition in host countries as well as decarbonising the power sector.
How that supply gets built isn’t neutral. A technology-led approach, backing specific decarbonisation methods across many host countries at once, scales the moment a credit type is approved. Country-based frameworks such as the Clean Trade and Investment Partnerships move slower: the EU’s only one so far, with South Africa, has spent 10 months in dialogue without a delivered offtake agreement.
A credible EU purchase commitment could give that architecture the demand signal required, but that depends on relationships which take years to build. For the volume the pilot phase needs by the early 2030s, the Commission should build the near-term pipeline through technology-led credit types and let country partnerships deepen the market once they mature. Who buys that early volume – and for what – is a separate question the proposal still leaves open given it can’t count towards the 2040 target.
The EU won’t be bidding alone. Switzerland and Japan have already traded emission reductions bilaterally, while Norway, Singapore, Sweden and South Korea have either signed agreements or have already pledged funding without one. China and the US remain wildcards, and if just one becomes a serious buyer, this would split an already thin pipeline further.
The Commission shouldn’t wait for 2033. A stated volume without a stated channel gives host countries and developers nothing to build against, and with lead times this long, there’s no room for legislation that arrives after the fact. Deciding early is itself the signal, like Norway’s USD 740 million commitment. Alas, speed isn’t the only answer: a fully stocked pipeline tells the EU it has enough credits, but not whether they’re good enough.
The standard that nobody’s defined
There’s currently no legal definition of ‘high-quality’ or ‘high integrity’ credits. The PACM is expected to serve as a reference point that may be complimented and the 2033 report is required to assess environmental integrity, accounting robustness and verification requirements. The problem is there’s nothing to set the criteria.
That’s why the Commission should quickly publish its own standard, treating PACM’s methodology as the floor and possibly building on the Oxford Principles for Responsible Engagement with Article 6 to define what the EU requires on top, so buyers and project developers know the quality bar they’re expected to meet well in advance of the 2033 review.
The first PACM credits show why such an absence matters. Rating agencies scored them differently, the Supervisory Body cut the credited volume (before approving the projects anyway) and NGOs argue that the reduced volume overstates the savings.
As one of the largest buyers after 2036, the EU has real leverage to set a standard the market converges around. That’s worth nothing if 2033 arrives before the EU has decided what it’s buying.
About the author:
Isabel Scheckenbach is Head of the Climate Programme within the Energy, Resources and Climate Change (ERCC) Unit.