EU Parliament votes to extend CBAM to 457 downstream product categories

The European Parliament made its move on 15 September, voting 464 for, 50 against and 159 abstentions to push CBAM beyond the primary metals and into downstream products. The number to lock in your head is 457: that is the count of product categories the Parliament wants covered, things like fasteners, wire, springs and household goods made from or containing steel and aluminium. Set that against the Commission's earlier 180 and the Council's 200, and you see the Parliament is pulling the scope wider than either of the other two institutions, and when all three disagree by that much, the only safe planning assumption is that the final line lands somewhere in the wide middle, not at the narrow end. CBAM itself, the carbon border levy on imports, entered its definitive phase on 1 January 2026, so this is not a future proposal, the mechanism is live, only the boundary of what falls inside it is still being hammered out, and that boundary is the thing that will land on your declaration.

Let me explain what CBAM actually is for anyone who files it loosely, because I still meet importers who think it is a tax someone else pays. CBAM is the EU's carbon border charge: instead of paying EU carbon price on goods made inside the bloc, importers of carbon-heavy goods pay a corresponding amount at the border so that foreign producers cannot undercut on carbon cost. Since 1 January 2026 the definitive phase means real declarations and real financial obligations, not the transitional reporting of the pilot years. Up to now the focus was cement, iron, steel, aluminium, fertilizers, electricity and hydrogen. The Parliament's 15 September vote says that is too narrow and wants the net cast over the finished and semi-finished things made from those metals, and that is the moment the problem stops being the steel mill's and starts being yours.

The reason this should worry an importer of finished goods is simple and easy to miss: a fastener, a length of wire, a spring or a household item that contains steel or aluminium may soon carry a carbon cost that today it does not. If your product sits in one of those 457 categories, you will need to know the embedded emissions of the metal inside it, not just the metal you buy as a raw billet. That is a much harder data job, because the carbon is now buried in a component or a finished article, and tracing it back to a mill's emission factor is exactly the paperwork the Parliament is voting to create, and the paperwork is where most teams actually fail, not the price.

Begin by mapping your EU-bound catalogue against the 457 categories while the list is still a stance and not yet law, because the window to prepare is the time before the law, not after. I would pull your harmonised tariff codes, the HS codes, for every steel- or aluminium-containing product you ship to the EU and check which ones fall in the fastener, wire, spring and household-goods neighbourhood. Do not wait for the trilogue to finish in 2027, because the data you need to collect takes longer to assemble than the law takes to pass, and a catalogue map you build this quarter is the foundation every later step sits on. Hand this mapping to your compliance lead by the end of October, with a flag on every code that even looks like it could be swept in, because the codes on the margin are the ones that will surprise you.

Next, start demanding embedded-emissions data from your suppliers now, not later, and write it into the commercial relationship rather than requesting it politely after the fact. Under CBAM you need the actual carbon content of the goods, and for downstream products that means your fastener maker or spring maker must tell you the emissions of the steel they used, traceable to a specific installation. I would write this into every new purchase order from this quarter: supply the CBAM-embedded-emissions figure or the contract gets a price-review clause. Suppliers who cannot produce it by 2027 will become a liability, and you want to know which ones those are this year, not when the border officer asks, because a supplier who cannot report is a supplier you may not be able to import from.

Now let me put a number on the exposure with the assumption stated plainly, so you can see the order of magnitude rather than guess. Assume you import 2 million euros a year of aluminium household goods into the EU, and assume the embedded carbon works out to 1 tonne of CO2 per 1,000 euros of product, so 2,000 tonnes of CO2 a year. Assume the EU carbon price, the ETS, sits around 75 euros per tonne for the sake of the example. That is 150,000 euros a year in CBAM cost that today you do not pay and under the Parliament's 457-category scope you would. None of those figures are your actuals; they are illustrative assumptions so you see the order of magnitude, and your real number depends on your product mix and your suppliers' emission factors, but if your volumes are larger than the example, scale the 150,000 euros up proportionally and see whether it moves your margin.

After that, build the declaration process before you are forced to, because the administrative load is the part that hurts small teams, not the carbon price itself. The definitive phase since 1 January 2026 already requires importers to handle CBAM declarations, and extending scope only multiplies the lines you file. I would stand up a simple spreadsheet or a small system that captures, per shipment, the HS code, the country of production, the installation, and the embedded emissions, owned by one named person. The teams that scramble in 2027 are the ones that treated each shipment as a one-off; the ones that sail through are the ones that already log this as routine, and routine is cheaper than heroics every single time.

And look at whether you can shift the carbon burden upstream or change the origin mix, because the cost is negotiable at the source even if it is fixed at the border. If one of your 457-category products is made from high-carbon steel in one country but a lower-carbon mill exists elsewhere, the CBAM cost difference between the two can be real money under a 75-euro carbon price. I would run a per-product comparison of landed cost including the illustrative CBAM charge for your current source versus a greener or differently-sourced alternative, and let the number decide, not habit. This is also where talking to your supplier about switching to green steel or aluminium can pay, because the carbon cost ultimately lands on your declaration, and a supplier who lowers your embedded emissions is lowering your bill.

Consider the option of holding inventory inside the EU to smooth the declaration flow, because a steady, documented stock position is easier to declare than a chaotic monthly arrival pattern. If your volumes are predictable, pre-positioning finished goods in an EU warehouse lets you declare against known, stable embedded-emissions data rather than scrambling for figures on every inbound shipment. The warehousing cost is real, but for high-volume 457-category products the predictability can be worth more than the rent, and it also buffers you against the trilogue timing risk, because your stock is already in and declared under the rules as they stand.

Talk to your customs broker and your EU subsidiary about this in the same breath, because the declaration is a customs act and the people who file it need to know your product mix as well as you do. I would schedule a call this month with the broker walking through your 457-category exposure line by line, and ask specifically how they intend to source embedded-emissions data for downstream products, because a broker who shrugs is a broker you will outgrow when the law bites. The earlier that conversation happens, the cheaper the later compliance.

Set a monitoring cadence on the trilogue, because the scope number will move and you need to know when. I would check the Parliament, Council and Commission positions monthly and note where the 457, the 200 and the 180 converge, because the convergence point is your real risk line. The 464-to-50 vote tells you the Parliament is serious, the 159 abstentions tell you it is not settled, and the teams that watch the drift are the ones that adjust their supplier asks in time rather than in panic.

The pitfall I see is teams assuming the 15 September vote is the final rule, and that assumption is the most expensive one in this whole story. It is not. It is the Parliament's negotiating position going into the trilogue, the three-way talks between Parliament, Council and Commission that run through 2027. The final scope could land closer to 200 or 457 or somewhere between, and the timing of application for downstream goods is still undecided. Betting your plan on the exact 457 is as dangerous as betting it never happens; the safe move is to prepare for the wider scope and be pleasantly surprised if it narrows.

Another trap is only watching the steel and aluminium headline and forgetting the declaration machinery, and that is how small teams get hurt. Even at today's scope, importers must prepare declarations under the definitive phase, and the administrative load is where small teams get hurt, not the carbon price itself. I would assign the declaration ownership clearly and rehearse it on your current metal imports so the wider scope, when it comes, is a data extension, not a fire drill, and a fire drill in customs is the kind you do not recover from quickly.

My honest worry is that finished-goods importers have treated CBAM as somebody else's problem, a raw-billet story for steel mills, and that comfort is about to end. The 464-to-50 vote says the Parliament disagrees strongly, and 159 abstentions means even the opponents did not kill it. If you ship fasteners, wire, springs or household goods with metal content to Europe, the carbon cost is coming to your declaration, and I would rather you spend this quarter building the data trail than spend 2027 explaining to customs why you have none, because an explanation is not a declaration and customs does not accept the former in place of the latter.

A practical point for your procurement team is that the embedded-emissions ask changes your supplier scorecard, not just your customs file. Once you need the carbon figure per shipment, the supplier who can produce it cleanly becomes cheaper to work with than the one who cannot, even if their unit price is a touch higher, because the cost of chasing missing data shows up as delay and as a declaration you cannot file. I would add a carbon-data readiness column to your supplier scorecard this quarter and rank your sources on it, because the 457-category scope will make that column the difference between a smooth import and a blocked one.

There is also a contract angle most importers miss. If your purchase agreement is silent on embedded emissions, you have no lever when the supplier fails to provide the number, and CBAM obligations land on you regardless. I would revise your standard purchase terms this year to include a CBAM data clause: supplier delivers the embedded-emissions figure with the commercial invoice, or a price adjustment applies. A clause like that turns a compliance risk into a shared responsibility, and shared responsibility is the only kind that survives a 457-category expansion.

Do not underestimate the internal ownership question, because CBAM sits between trade, finance and customs and falls through the cracks when nobody owns it. I would name one person as the CBAM owner across those three functions, with a mandate to collect supplier data, run the declaration, and report the cost, because a number that nobody owns is a number that arrives late and wrong. The companies that handled the definitive phase well from 1 January 2026 are the ones that assigned the owner before the deadline, not after.

The longer-game point is that carbon cost at the border is not going away even if this specific scope narrows. The Parliament voted 464 to 50 to widen it, and even a narrowed final rule still covers more than the original metals, so the direction of travel is clear. I would treat the embedded-emissions data you build now as a permanent asset, not a one-time compliance project, because the next extension will ask for the same data on more products, and the importer who already has it wins the next round without extra work.

My honest closing is that the abstentions matter as much as the yes votes. The 159 who abstained did not kill the widening, they left the door open, and in EU trade politics an open door on carbon usually closes in the direction of more coverage, not less. If you import metal-containing goods to Europe, build the data trail now while the law is still being written, because the price of waiting is a scramble in 2027 that you could have avoided with a quiet quarter of preparation this year.

A practical staffing note: the importer of record carries the CBAM obligation, so if you sell to the EU through a distributor or a subsidiary, the responsibility may sit with them, but the data still has to come from you. I would put the data-supply obligation in the commercial agreement regardless of who files, because a declaration filed by someone else still needs your embedded-emissions number, and the person filing cannot invent it. The 457 scope makes this non-negotiable, because downstream products hide the metal inside a finished article.

Consider using an authorised CBAM declarant rather than building the capability in-house if your volume is modest. The definitive phase since 1 January 2026 lets a named declarant file on your behalf, which can be cheaper than hiring a compliance analyst for a handful of declarations a year. I would price both the in-house and the declarant route this quarter, because for a small importer the declarant is often the rational choice until volume justifies a team.

Scenario planning is the discipline that turns a scary 457 into a managed risk. I would model three scopes, the Council 200, the Commission 180 and the Parliament 457, and for each compute the embedded-emissions data you would need and the cost you would carry. The widest scope is your planning case, the narrowest is your hope, and the convergence of the three institutions is your realistic line, and knowing all three means no outcome surprises you.

Do not forget to talk to your customers about the cost pass-through, because a CBAM charge on your declaration is a cost you will want to recover. I would model the per-product carbon cost and decide which SKUs absorb it and which pass it to the buyer, because pretending the cost does not exist is how margins quietly vanish. The 150,000 euro example from earlier is real money on a 2 million euro flow, and real money deserves a pricing decision, not a shrug.

My added point is about lead time for the law itself. The trilogue runs through 2027, but the data you need takes longer than the law to assemble, so the smart importer builds the data trail now and lets the lawyers argue the scope. I would treat the embedded-emissions register as a permanent asset, because every future extension will ask for the same numbers on more products, and the importer who already holds them wins the next round without scrambling.

The last thing I would tell a finished-goods importer is to stop waiting for certainty, because certainty in EU trade law is a luxury you pay for in lost preparation time. The Parliament has spoken at 464 to 50, the abstentions keep the door open, and the trilogue will move through 2027 regardless of your schedule. The embedded-emissions data you start collecting this quarter is useful under every scenario the institutions are debating, so the only losing move is to collect nothing and hope the scope stays narrow.

• Map your EU-bound steel- and aluminium-containing HS codes against the 457 categories by end of October, owned by compliance.
• Insert a CBAM embedded-emissions data clause into every new purchase order from this quarter, owned by procurement.
• Build a per-shipment CBAM tracker capturing HS code, origin, installation and embedded emissions, owned by one named person, this quarter.
• Run a per-product landed-cost comparison including the illustrative CBAM charge for current versus greener sourcing, review by end of November.
• Brief your customs broker on your 457-category exposure line by line this month and confirm their data-sourcing plan.
• Pre-position high-volume 457-category stock in an EU warehouse to smooth declarations against stable emissions data.