Europe’s carbon border charge has yet to show up fully in steel prices because pre-emptive stockpiles are still being worked off, and distributors say the reckoning is coming
WARSAW, 20 September 2026 – The European carbon border levy has been law for long enough to have changed prices, and it has not changed them as much as anyone expected. Panellists at the EUROMETAL Regional Meeting Central Europe this week explained why, and the answer should worry anyone who assumed the transition had been absorbed: the market is still digesting inventory that was imported ahead of the rules.
The delay is temporary. When the stockpiles clear, the full price effect arrives, and it will land on a European distribution sector that is already operating at close to zero margin and on manufacturers who say they cannot pass increases on.
The panellists also delivered a message on policy that has become the consistent refrain of European metals distribution in 2026. The Carbon Border Adjustment Mechanism is directionally correct and substantively insufficient. Unless it reaches downstream to the manufactured goods that compete with European fabrication, it will raise European input costs without protecting European manufacturing.
Why the price signal has been muted
Energy and labour costs have been more dominant factors in European steel pricing during 2026 than carbon, according to the Warsaw panel. That is not because the carbon charge is small. It is because the market has not yet had to pay it on most of the material currently in circulation.
Marcin Matysiak, head of long products purchasing at Polish distributor Konsorcjum Stali, attributed the muted signal to the large inventories accumulated during the front-loading of imports ahead of both the carbon mechanism and the tariff-rate quota system. “Now it’s just a matter of inventories being digested by the market, and this will come,” he said, in remarks reported by Kallanish.
Front-loading is entirely rational behaviour. When importers know that a cost is coming on a defined date, they bring material in before that date. The result is an inventory bulge that suppresses prices in the short term and then unwinds, at which point the cost appears in full.
The quantitative picture supports this. Carbon pricing did boost European steel prices by roughly 5 per cent in the first quarter of 2026, according to Roland Fazekas, chief executive of Carboferr. That is a meaningful figure but considerably less than the mechanism implies at full application, and it reflects a market in which most transactions were still drawing on pre-regime stock.
The broader price movement over the year has been driven by other factors. European domestic hot-rolled coil was assessed by Platts at 730 euros per tonne ex-works Ruhr in Northern Europe and 725 euros per tonne ex-works Italy in Southern Europe in mid-September, both up 110 euros per tonne since the start of the year, while the MEPS Europe Average hot-rolled coil price has risen more than 16 per cent between January and September. Those increases track the tariff-rate quota cut and energy costs more closely than they track carbon.
The 1.5 per cent problem
The most vivid illustration of the downstream gap came from Fazekas, whose company invested a decade ago into machining and engineering capability to produce parts for sectors including automotive.
Describing his firm’s competitive position, he said: “We are fighting with the tier ones and the OEMs at the same time. Plus, we are having the mills on the other side.”
He then gave a specific comparison from a contract to supply parts for BMW production in Hungary, where the question arose of what to outsource to China and what to keep in domestic production. “The effective EU import duty on a Chinese precision tool is 1.5 per cent, roughly. If we just look at quarto plates directly imported from China, it is 70 per cent import duty. I think the Chinese are laughing all the way down on us, that how come we are so stupid?”
That contrast is the whole argument in one data point. A European manufacturer who imports the raw plate to make a precision tool faces a prohibitive charge. A European customer who imports the finished Chinese precision tool faces almost nothing. The policy therefore makes it more attractive to buy the finished foreign product than to make it in Europe, which is the opposite of the intended effect.
The 70 per cent figure reflects the combination of instruments now applying to Chinese plate: anti-dumping duties, the ordinary customs duty, and the 50 per cent out-of-quota rate under the steel import regulation that applied from 1 July 2026, which cut tariff-free quota volumes by 47 per cent to 18.3 million tonnes a year.
The distributors’ position
Wojciech Gruszka, chief executive east at ArcelorMittal Distribution Solutions, placed the distribution sector at the point where the policy failure becomes visible.
“We are at the first line between the upstream and our customers. Customers create demand,” he said. “If we allow to push this production of metals, processing of metal outside of Europe, we will die. We know that they are starting switching to importing the semi product. They are cutting costs.”
The behaviour he describes is the mechanism by which the policy unwinds itself. A European manufacturer facing higher steel costs does not necessarily close. It substitutes toward imported semi-finished material, or toward imported components, or it moves the fabrication step offshore and imports the result. Each of those choices reduces European steel demand by more than the original import restriction reduced European steel supply.
Gruszka’s verdict on the carbon mechanism was that it is “a good direction but this is not enough”.
Jan Moravec, chief executive of Czech distributor Ferona, added a set of process criticisms. Trade barriers should be limited in duration and have clear targets, he argued. European mills are suffering losses while distributors operate at close to zero profit margin. Environmental, social and governance reporting requirements mean companies “have to have departments, maybe scientific teams for analysing all the rules and the standards”.
His summary judgement was that Green Deal targets should be realistic and that market protection should move faster, as it does in the United States, rather than European authorities overanalysing.
Fazekas identified the reason for the caution. There may be resistance from authorities to implementing tariffs on downstream goods because of fears of retaliation against European manufacturing exports.
Adaptation: what distributors are actually doing
Alongside the policy argument, the Warsaw meeting produced a picture of how European distribution is restructuring its own operations, and the numbers are striking.
Zenon Jedrocha, a board member at Polish distributor Stalprofil, described a transformation in order profile. “Two to three years ago, we sold more or less about 200 items, and the average sale was two tonnes. Now we are selling almost 1,000, average 100 kilo. It means that it’s a complete change.”
That is a fivefold increase in the number of stock keeping units and a twentyfold reduction in average order size. It is a shift from bulk distribution to something closer to industrial retail, and it requires entirely different infrastructure. “We have to invest plenty of money for the two-high storage system, and it’s a fully automatic system that allows us to sell such more items,” Jedrocha said.
He also noted the financing environment. Plenty of distributors have invested at high interest rates in Poland, and “we are waiting since years for a better time”.
Konsorcjum Stali has completed a new service centre investment responding to changing customer logistics and quality requirements. Ferona is investing in process digitalisation.
The underlying driver is that European manufacturers are holding less inventory, ordering more frequently and in smaller quantities, and demanding more processing from their suppliers. That is a rational response to cost pressure and uncertainty, and it transfers working capital and complexity from the manufacturer to the distributor.
Green steel is not yet a market
A final theme from Warsaw concerned the product that the entire carbon architecture is meant to promote.
Demand for premium-priced green steel has been tempered, largely by the squeezed profit environment across the chain. Gruszka explained the customer position bluntly: “Every single change on our side, regarding prices, even during the contract time, one quarter, two quarters, for them is a problem that they cannot accept. They cannot take the burden of the increasing of the raw material from our side. They are not working on three-digit margins, right?”
Moravec was more direct still. Without government incentives for green steel use, “our customers do not care at the moment”. He added that the definition of green steel first needs to be agreed before a market can be created.
That last point is substantive rather than rhetorical. There is no single European standard defining what qualifies as green steel, which means a premium product cannot be reliably specified, procured or verified. Until it can, buyers will not pay a premium and producers cannot recover the cost of the investment the carbon architecture is designed to incentivise.
Outlook
On demand, the Warsaw panel was cautious. Moravec expects consumption to remain flat for the next two to three years. Gruszka expects distributors to grow in tandem with returning European steel capacity even if demand is flat. Jedrocha pointed to European Union funded projects as a source of demand in Poland. Matysiak said his company is prepared for expected improvement.
Fazekas offered the most interesting structural observation: Central and Eastern Europe should be a beneficiary of manufacturing relocation from western Europe. That is relocation within the single market rather than out of it, driven by energy and labour cost differentials, and it would leave the region’s distributors better placed than their western counterparts.
Implications for importers and exporters
Three practical conclusions follow for companies outside Europe.
The inventory overhang is temporary and the price effect is deferred, not avoided. Exporters modelling European demand on 2026 pricing should assume that carbon costs become more visible in 2027 as pre-regime stock clears. The 5 per cent first-quarter effect is a floor, not a ceiling.
The downstream gap is real and currently favourable to foreign suppliers of finished goods. A Chinese precision tool at a 1.5 per cent effective duty against Chinese plate at 70 per cent is an arbitrage that exists today and that European industry is actively lobbying to close. Exporters currently benefiting from it should treat the advantage as time-limited and should price accordingly.
The change in European buying behaviour, toward many more line items in much smaller quantities, changes what European customers want from foreign suppliers. Smaller, more frequent, more processed consignments carry higher logistics and compliance cost per tonne. Exporters who can serve that pattern will hold business that bulk-only suppliers will lose, and the compliance burden of quota timing and melt-and-pour documentation makes small frequent consignments harder, not easier.
The European market is becoming more expensive, more administratively demanding and more fragmented in its ordering patterns. None of those trends reverses when the inventory clears.
Anatomy of the front-loading
The inventory overhang that is masking the carbon price signal deserves a closer look, because understanding how it was built tells you how long it will take to clear.
Front-loading occurs when market participants can see a cost arriving on a known date. Two such dates converged in 2026. The carbon border mechanism moved into its definitive phase, ending the transitional reporting-only period and beginning actual certificate obligations. And the steel import regulation applied from 1 July 2026, cutting tariff-free quota volumes by 47 per cent to 18.3 million tonnes a year and doubling the out-of-quota duty to 50 per cent.
Importers responded rationally. Material was brought in ahead of both deadlines, financed and warehoused. The inventory that resulted is not evenly distributed. It is concentrated in the hands of the operators who could fund it, which means large distributors, integrated traders and service centres with strong balance sheets.
That concentration has two consequences. The first is that the price signal is being suppressed selectively rather than uniformly, because the participants holding cheap stock can undercut those who must buy at current prices. The second is that the unwinding will not be gradual. Inventory carried at high interest rates gets liquidated when the carrying cost exceeds the expected price gain, and liquidation tends to happen in a cluster rather than a trickle.
Reporting from MEPS this month captured the early stage of exactly that dynamic: some European distributors are selling material significantly below current mill offers as they seek to convert stock into cash before year end, against a backdrop of high inventories and low demand.
For a buyer, that creates a specific and time-limited opportunity in the fourth quarter. For a seller into Europe, it means the near-term market will look softer than the policy settings imply, and that the underlying tightening will become visible only in 2027.
The pass-through problem in numbers
The reason European manufacturers cannot absorb the increases is arithmetic rather than sentiment, and the figures from the Warsaw discussion and from market assessments allow it to be set out.
European domestic hot-rolled coil was assessed by Platts at 730 euros per tonne ex-works Ruhr and 725 euros per tonne ex-works Italy in mid-September, both up 110 euros per tonne since January. Imported coil was assessed at 585 euros per tonne CIF Antwerp and 580 euros per tonne CIF Southern Europe, up 85 euros per tonne over the same period.
A European fabricator buying domestically therefore pays roughly 145 euros per tonne more than the world price available to a competitor outside Europe. On a product where steel is 30 per cent of the cost of goods sold, that is a cost disadvantage of several percentage points on the finished item before any consideration of energy, labour or compliance.
Against that, the effective import duty differential that Fazekas described is the mirror image. Chinese quarto plate at an effective 70 per cent charge, against a Chinese precision tool made from that plate at roughly 1.5 per cent. The policy taxes the input at close to fifty times the rate it taxes the output.
No commercial operator can absorb a gap of that shape. The rational response is not to absorb it but to change the sourcing decision, which is precisely the behaviour Gruszka described when he said customers are starting to switch to importing semi-finished product.
What the SKU shift really means
The operational change Jedrocha described, from roughly 200 items at an average of two tonnes to nearly 1,000 items at an average of 100 kilogrammes, is one of the more revealing pieces of information to come out of the European market this year, and its implications extend well beyond Poland.
A twentyfold reduction in average order size means European manufacturers are no longer buying in production runs. They are buying to immediate requirement. That is a response to uncertainty: uncertainty about their own order books, about price direction, and about whether material ordered today will still be needed when it arrives.
It also transfers cost. Holding a thousand line items rather than two hundred requires more storage capacity, more sophisticated handling and far more capital tied up per euro of revenue. The automated high-bay systems Jedrocha described are the technical answer, and they are expensive, particularly when financed at Polish interest rates.
For foreign suppliers, this changes what European customers actually want. A European distributor now needs breadth of range, short lead times and reliable small-lot delivery far more than it needs the cheapest large-lot price. An exporter geared to shipping full vessels of a single specification is selling into a market that has moved away from that model.
It also intensifies the quota problem. Quota allocation is first-come, first-served on release into free circulation, which rewards large consignments cleared early in a quarterly period. Small, frequent consignments are administratively harder to time against quota windows and carry proportionally more documentation cost, including the melt and pour evidence required from 1 October 2026.
The European market is therefore asking for smaller shipments while the European regulatory system rewards larger ones. That contradiction sits squarely on the distributors, which is part of why their margins are where they are.
Green steel: the missing market
The green steel discussion at Warsaw identified a gap that has significant implications for whether the entire carbon architecture achieves its purpose.
The carbon border mechanism is designed to make low-carbon European production competitive by pricing the carbon in imports. That logic assumes a market exists in which low-carbon steel commands a premium. The Warsaw panel’s assessment is that it does not, for two reasons.
The first is margin. Customers operating on thin margins cannot accept price increases within a contract period, let alone a structural premium for an environmental attribute that does not change the physical performance of the product.
The second is definitional. There is no agreed European standard for what constitutes green steel. Without a definition, a buyer cannot specify it, a procurement department cannot verify it, and a producer cannot prove it. Moravec’s point that the definition must precede the market is a practical observation about how industrial procurement works rather than a philosophical one.
Until both are resolved, the carbon mechanism functions as a cost on imports rather than as an incentive for low-carbon production, which is half of what it was designed to do.
Watch list
Four developments will determine how the deferred price effect lands.
The clearance of front-loaded inventory, visible in distributor stock levels and in the spread between mill offers and distributor transaction prices. When that spread closes, the digestion is complete.
The fourth quarter quota period opening on 1 October, alongside the melt and pour evidence requirement taking effect the same day. The speed of quota exhaustion will indicate how binding the 47 per cent volume cut really is.
The trilogue between Parliament and member states on extending the carbon mechanism downstream, following the Parliament’s 15 September vote backing a scope potentially exceeding 400 product lines.
And European manufacturing output data. If the substitution toward imported semi-finished goods that Gruszka described is happening at scale, it will show up as falling European steel consumption alongside rising imports of fabricated products, which is the signature of the policy failing on its own terms.