The CBAM factor nobody could look up is now 100 %
For most of this year, one of the inputs to the CBAM calculation could not be looked up anywhere. Not because it was hard to find — because it did not exist yet. The cross-sectoral correction factor, the CSCF, sits in the formula that decides how much of your embedded emissions is treated as free of charge, and for the allocation period beginning 1 January 2026 it was simply unset.
It is set now. 100 %, for every year from 2026 to 2030 (Article 1 of Implementing Decision (EU) 2026/1862 of 23 July 2026, published in the Official Journal on 24 July 2026).
That is the short version, and for most declarants it is good news of the dullest kind: the number people assumed turned out to be the number that was adopted. The part worth six minutes is what it does — and does not do — to figures that were already calculated.
Where the factor sits
Under the definitive regime, the free-allocation adjustment for a good is:
SEFA = CBAM factor × CSCF × benchmark
Three terms, and only the middle one was missing. The CBAM factor is the share still allocated free in a given year and steps down on a schedule written into the law — 97.5 % in 2026, 95 % in 2027, 90 % in 2028, 77.5 % in 2029, 51.5 % in 2030 (Article 10a(1a) of Directive 2003/87/EC). The benchmark is the goods-specific figure published in the annex to Implementing Regulation (EU) 2025/2620. The CSCF is a uniform factor that exists to keep the total of all free allocation inside the emissions-trading cap: if the sum of what installations are entitled to exceeds the ceiling, everyone's allocation is scaled down by the same percentage (Article 10a(5) and (5a) of Directive 2003/87/EC).
That is the whole mechanism, and it explains why the factor is so easy to miss. It is not a CBAM instrument. It is adopted under Article 14(6) of Delegated Regulation (EU) 2019/331, in the emissions-trading file, on the schedule of the free-allocation cycle — and it lands in the CBAM formula from the side. A declarant watching the CBAM acts for the numbers they need would not have seen it coming, and would not see it change.
What was decided, and why
The decision is short. Article 1 sets the factor for the whole allocation period at once:
"For each year in the allocation period 2026 to 2030, the uniform cross-sectoral correction factor for the adjustment of free allocations of emission allowances in accordance with Article 10a(5) and (5a) of Directive 2003/87/EC shall be 100 %."
The reasoning is in the recitals, and it is arithmetic rather than policy. Of the maximum amount available for free allocation in 2021–2025, 125,586,948 allowances went unused, alongside the entire 3 % reserve for that period (recital 8). Against that headroom, the Commission puts the total draw needed across 2021–2030 at 289,226,630 allowances from the reserve (recital 9). The entitlement fits inside the ceiling, so nothing has to be scaled down — the factor is 100 % and the term is neutral.
Two things are worth noting beside it. The same factor was 100 % for 2021–2025 (Implementing Decision (EU) 2021/927), so the adopted value continues an unbroken line rather than breaking one. And it is not to be confused with Decision (EU) 2026/1312 of 15 June 2026, which approved the national implementation measures — the installation lists — five weeks earlier and sets no factor at all. If you went looking in June and came back empty-handed, that is probably what you found.
What 100 % means in the arithmetic
It means the middle term drops out. SEFA = CBAM factor × benchmark, and the figure you get is the
figure you would have got by assuming 1.0.
A worked line, using a cement consignment of the kind that shows up in every second CBAM file — grey clinker under CN 2523 10 00, benchmark 0.6660 tCO₂/t, imported in 2026:
SEFA = 0.975 × 1.000 × 0.6660 = 0.64935 tCO₂/t
Now the counterfactual, because it shows how much was riding on an unpublished number. Had the factor
come in at 90 %, the same line would read 0.975 × 0.900 × 0.6660 = 0.584415 — a free allocation
lower by 0.064935 tCO₂ per tonne. On a 512-tonne consignment that is 33 certificates, about
€2,500 at the Q2 2026 certificate price of €75.28. One consignment, one commodity, one
percentage point set in a decision most CBAM declarants never read.
The number is right. The status is not.
Here is the part that catches tidy operations.
If you computed a 2026 obligation before 24 July 2026, you used an assumption. The decision does not retroactively turn that into a calculation — it produces the same digits, which is a different thing. Every figure computed under the assumption is now confirmable, and until somebody confirms it, it is a figure whose basis is a guess that happened to be correct.
That distinction has a practical edge, because the failure mode is invisible. A recomputed figure and an assumed figure are identical to the last decimal. Nothing in the output flags which is which. If your process marks provisional figures — and it should — then those marks are the only thing standing between "we checked" and "we were lucky", and they do not clear themselves when the act is published. Somebody has to recompute against the adopted value and re-date the result.
In our own system that is deliberate: an assessment carries the provisional flag it was stored with, the declaration export stays locked while any figure on it is provisional, and publishing the decision does not silently unlock anything. The figures do not change. Their standing does, and only once they have been run again.
Anyone who assumed something other than 1.0 has the opposite problem and a bigger one. A factor whose entire purpose is to scale allocations down invites a cautious guess below 100 %, and a cautious guess is the wrong kind here: it understates the free allocation and therefore overstates the certificate obligation. That is the error that costs money rather than the one that costs a penalty, and nobody is going to write to you about it.
One more thing, and it is genuinely useful
The decision covers 2026 to 2030 in a single article. That is five reporting years in which this particular input does not move, does not need an annual lookup, and does not belong in the list of things to re-check each January. The next time it is open is the allocation period starting in 2031.
For a regime where almost everything else is annual — the CBAM factor steps every year, the benchmark annex is revised, the certificate price moves quarterly in 2026 and weekly from 2027 — one input that is fixed for five years is worth writing down as fixed.
What to check this week
- Find out which CSCF your figures used. Not which one your provider intended — which one is in the stored calculation. If the answer is "1.0 by assumption", you are fine numerically and owe yourself a recomputation for the record.
- Recompute anything calculated before 24 July 2026 and re-date it, rather than reasoning that the result would be unchanged. It would be. That is not the point of the exercise.
- If any figure came in with a factor below 100 %, redo it. It overstates your obligation, and nobody will send you a correction for paying too much.
- Take this one off your annual list until 2031, and put the source on it: the Official Journal entry, not a summary. One practical note for anyone automating that check — EUR-Lex answers scripted requests for this document with an empty response, so the text has to be read by hand.
This article explains the rules as we read them; it is not legal advice. The factor is set by Article 1 of Implementing Decision (EU) 2026/1862 of 23 July 2026 (OJ L, 24 July 2026), adopted under Article 14(6) of Delegated Regulation (EU) 2019/331; the free-allocation mechanism in Article 10a(5) and (5a) of Directive 2003/87/EC; the CBAM free-allocation adjustment and its benchmarks in Implementing Regulation (EU) 2025/2620; the preceding period's factor in Implementing Decision (EU) 2021/927. The recital figures are quoted from the decision itself. Verify your own position with your customs agent or the competent authority in your member state — in Germany, the DEHSt.