EU Tyre CVD Disclosure: Combined AD+CVD Hits 27.50%-57.09%, Ruling Due 5 Dec

The Analysis

How does this shipment actually move? That is the question to settle first, and it is the one no headline answered. On 19 October the European Commission issued its pre-final disclosure in the countervailing duty investigation into passenger car and light truck tyres from China. The subsidy rates run from 24.63 per cent to 29.63 per cent. Set the anti-dumping rates from the 7 July final determination alongside them, zero to 27.46 per cent, and the combined burden lands somewhere between 27.50 per cent and 57.09 per cent. The disclosure also settles the thing everyone was quietly braced for: no retroactive collection. Cargo already shipped and cargo already cleared is not going to be chased backwards. The countervailing ruling is due on 5 December. Written comments close on 29 October. A hearing has to be requested within three days of the disclosure coming out.

I read that range and I do not see a range. I see a lookup table, and I see one field on the import declaration that decides which row of the table applies to my container. Everything else in the document sits downstream of that field. A rate you cannot map to a name on a form is a number in a press release, and a number in a press release has never paid anybody's duty bill.

So this piece walks the five boxes the cargo passes through, classification through declared particulars to inspection, duty payment and release, and points at the one that moved. It lays the seven rate buckets side by side and shows which half of the bill is genuinely still open. It runs the arithmetic on an assumed container. It finishes with what to pull together, who to ring, and by what hour.

Start at the front of the flow. Classification, declared particulars, inspection, duty payment, release. Every trade measure eventually lands inside one of those five boxes, and if you cannot say which one, you cannot say what has to be fixed before the vessel arrives. That is the discipline. Not the rate, not the politics, not the sector commentary. The box.

This one does not touch classification. The scope is drawn by product description, passenger car and light truck tyres, and a given tyre either sits inside that description or it does not. Nobody is asking anyone to reopen the tariff line or to argue about where a size range falls. What moves is the second box, the declared particulars: which producer is named, which exporter is named, and the additional code hanging off the tariff line that tells the customs system which rate to pull. That is where the entire outcome is decided, and it is decided silently, by a database lookup, without anybody at the port asking you whether you meant to do it that way.

Which is why the first thing out of my mouth on a file like this never changes. Check the code, then check whether any relief or measure attaches. In that order. People ring me about the percentage before they have established which row of the table they are even standing in, and then they are surprised when the assessment comes back at a number they never modelled. The order matters more than the number.

Here are the seven buckets, written out in full, because the shape of the table tells you more than the headline range ever will. Hankook group lands at 27.50 per cent combined, being 27.50 countervailing plus zero anti-dumping. Linglong group lands at 31.00 per cent, being 24.63 plus 6.37. Yongsheng Rubber lands at 48.02 per cent, being 29.63 plus 18.39. The forty-eight other companies that cooperated in both investigations land at 28.76 per cent, being 27.50 plus 1.26. Companies that cooperated only in the anti-dumping investigation and not in the subsidy one land at 36.19 per cent, being 29.63 plus 6.56. Companies that cooperated only in the subsidy investigation and not in the anti-dumping one land at 49.66 per cent, being 27.50 plus 22.16. Every other company, uncooperative in both, lands at 57.09 per cent, being 29.63 plus 27.46.

Now look at what repeats, because this is the part that reframes the whole file. On the countervailing side there are only three values anywhere in that table: 24.63, 27.50 and 29.63. Two of the three are not calculations at all. They are switches. 27.50 is what you receive if you answered the subsidy questionnaire. 29.63 is what you receive if you did not. The distance between them is 2.13 percentage points, and no amount of argument about your cost structure moves you across that line, because the line is drawn by participation, not by arithmetic. On the anti-dumping side the values scatter from zero all the way to 27.46, a spread of 27.46 points, and those came out of company-specific calculations.

Put the two halves together and the picture is uncomfortable. Of the 29.59-point gap between the cheapest combined rate and the dearest, 27.46 points come from the anti-dumping column and 2.13 points come from whether you filled in a subsidy questionnaire. The anti-dumping column was finally determined on 7 July. It is closed. What remains open on 5 December is, in structural terms, a two-point question about cooperation status. That is not to say the two points do not matter, only that anybody spending their week relitigating the dumping margin is pushing against a door that was locked three months ago. The live question is which row you are in, and the row is selected by a name.

Why the table has this lopsided shape at all comes down to the fact that two investigations ran on parallel tracks with separate clocks. A company could engage with one case and miss the other, or engage with both, or engage with neither, and the Commission has built a bucket for each of those four states, then set individual rates inside the buckets for the groups it examined on their own merits. That is the mechanism producing a 36.19 bucket sitting below a 31.00 bucket, which looks wrong until you see that the first is a non-cooperation marker with a modest dumping rate and the second is a cooperating group with a higher one. The ordering is not intuitive. Reading the table as a league table of company performance is how people end up in the wrong row.

Put a container against all of it, using an explicit assumption. Take one 40-foot container of PCR tyres declared at 80,000 US dollars of customs value at the EU port. At 27.50 per cent the duty is 22,000 dollars. At 28.76 per cent it is 23,008 dollars. At 31.00 per cent, 24,800. At 36.19 per cent, 28,952. At 48.02 per cent, 38,416. At 49.66 per cent, 39,728. At 57.09 per cent, 45,672. Between the floor and the ceiling of this table you are looking at 23,672 dollars on a single container.

Now assume you ship forty containers a month into the EU on this lane. The distance between landing in the both-cooperated bucket at 28.76 and dropping into the all-others bucket at 57.09 is 28.33 points, which is 22,664 dollars a container and 906,560 dollars a month. The same two-point cooperation switch, on the same assumption, is 1,704 dollars a container, 68,160 dollars a month. One of those numbers is worth a week of somebody's time. The other is worth a phone call this afternoon.

Timing is where most of the avoidable damage happens, and the clock on this file is not the one people are watching. Comments on the disclosure close on 29 October, ten days after publication, and that is the only window in which an interested party puts written arguments on the disclosed calculations. A hearing has to be requested within three days of the disclosure issuing. That window is already gone for anybody who read the trade press a week late, and it is worth asking inside your own organisation whether there is any standing process for catching a three-day window at all, because they show up in every trade case and they get missed almost every time. Until 5 December you keep declaring at the anti-dumping rates from the 7 July final determination. After 5 December the anti-dumping order formally takes effect and the countervailing rates ride on top of it. Nothing that arrived and cleared before then gets reopened, because retroactive collection has been ruled out in the disclosure.

And here is the piece that no headline carried, the piece that will actually cost money. The China Rubber Industry Association asked the companies involved to read the disclosure document carefully and to check their English company name. That line reads like housekeeping. It is not housekeeping, it is the key to the entire table. Your rate is not written against your Chinese name, or your business licence, or the name your sales team prints at the top of a quotation. It is written against a string of Latin characters in an annex to a Commission document. If that string does not match what your importer declares, the customs system does not find you in the 28.76 row, and it does not send anybody a question, and it does not pause. It defaults. The default row is 57.09 per cent.

On the assumed 80,000-dollar container, that is the difference between paying 23,008 dollars and paying 45,672 dollars. Twenty-two thousand six hundred and sixty-four dollars, turned by a full stop. A Co., Ltd where the annex has Co. Ltd. A group parent named where the annex names the operating subsidiary. A subsidiary named where the annex names the parent. A factory that was reorganised in March and is still being declared under the old spelling. Every one of those is a live way to lose twenty-eight points, and every one of them is invisible until the assessment comes back, by which point the cargo is long gone and the money is already out of the account.

The boundary conditions matter at least as much as the rates, and that is where Dr. Ingrid Voss would put her knife down first: not on the percentage, but on what has to be true before the percentage applies at all. Two conditions have to hold. The product has to fall inside the scope as it is described in the measure, and the exporter has to be an entity the measure actually names. Break either one and the rate never attaches. Satisfy both in a configuration you did not plan for and it attaches anyway, whether you budgeted for it or not. That is precisely why the English-name check is not administrative trivia. It is the test of whether the second condition is met, run one character at a time.

Disclosure documents are beautiful. They are typeset, indexed, footnoted, with every legal entity spelled out in full in an annex and every calculation step shown. The terminal is not like that. The terminal is a declaration going in at eleven at night with whatever the forwarder had on file from the last booking, and a system that looks up a code and produces a number, and a duty bill arriving three weeks later that nobody wants to sign for. Papers that do not reconcile at the front of the process become problems all the way down it, and by the time they surface you are arguing about money instead of fixing a spelling. This is not cynicism, it is the ordinary shape of the work.

Three things could still move, and each of them changes what a sensible person does this week. A disclosure is not a determination. The 5 December ruling can land at different numbers, and while the architecture of the buckets rarely gets rebuilt at that stage, individual rates have been known to shift. On the same assumption, one percentage point either way on a company rate is 800 dollars a container and 32,000 dollars a month on forty containers, so the sensitivity is real but it is not the difference between viable and not viable. Build your landed-cost model with a band in it rather than a single figure, and agree with your counterparty in writing what happens if the final number sits outside the band.

The second variable is whether rushing cargo through before 5 December buys anything. The disclosure rules out retroactive collection, which means what has already shipped and cleared is safe. Whether an order that has not shipped yet can be landed and cleared ahead of the ruling and thereby sit outside the new measure depends on the entry provisions of the order itself and on how the port you use actually processes it, and that is a port-by-port question. I am not going to hand anybody a legal conclusion on it, and I would be wary of anybody who does. What I will say is that a plan built on beating a date by three days, on a lane where berth windows and terminal dwell are not yours to control, is not a plan. Treat any saving from it as zero until it is confirmed in writing by somebody who will put their name to it.

The third variable is the one every board asks about within a day: move the sourcing. Shifting production or buying from a plant outside China is a sourcing decision with an origin question attached, and the origin question is the one that bites. A changed invoice header and a new factory address do not by themselves change origin where a substantial transformation test applies, and the test is applied at the port, on the evidence in the box, not on the strength of a supplier assurance. Whether a given re-routing survives that test is a question for qualified counsel in the importing market and for your own origin documentation, not for me. What I can tell you is what it does to the paperwork: every change of manufacturing source resets your origin evidence, your supplier declarations, your product compliance files and your customer's approved-vendor records, and that reset has a cost and a lead time of its own.

So, to the front of the flow again, and to what goes in the box this week. Pull the annex and copy the English entity names out character by character, including punctuation and the legal suffix. Compare them against the producer and supplier master data your broker actually declares from, not against the names on your website or your letterhead, because those are not the same thing and they drift. Where a group has several legal entities selling into the EU, establish which entity is named in the measure and which entity is on the invoice, and if they differ, establish why before anything ships. Lock the master data once it is corrected, and put a date on it, because a name that was right in October and reverted in January is the same problem wearing different clothes.

On the importer side, the conversation nobody wants to have is the Incoterms one. On a delivered-duty-paid term, a jump of this size lands on the seller, and a seller who priced a DDP programme before 19 October has now written a cheque for a number that did not exist when the quotation went out. On a term where the buyer carries the import duty, the buyer needs the revised landed cost before the next order, not after the next arrival. Either way the contract needs a clause saying who bears a change in trade-remedy duty and what happens to orders already placed and prices already agreed.

Then the diary. Twenty-ninth of October for written comments on the disclosure, and if you intend to file, the drafting has to start now, because ten days is not ten days once lawyers and technical staff are involved. The three-day hearing window has closed for this disclosure, and the useful output from missing it is a standing rule inside your organisation that any trade-remedy publication triggers a same-day check of deadlines, with a named owner. Fifth of December for the ruling, and the operational date that matters to you is not the ruling date but the arrival date of the first vessel that clears after it. Work out which sailing that is on your lane and put it on the same sheet as your inventory cover, because the question you will be asked in the first week of December is how many weeks of stock you have at the new duty and whether the old-priced stock is enough to bridge a renegotiation.

If you are in the both-cooperated group, your job this week is small and specific: confirm the name, confirm the code, confirm your broker is declaring from the corrected master record, and ask for one test declaration to be run and the assessed rate checked before volume moves. If you are in any of the non-cooperation buckets, understand that the 5 December ruling is unlikely to be the moment your rate improves, and plan the commercial conversation with your EU customer on that basis rather than on hope. If you do not know which bucket you are in, that is the actual emergency, and it is the one thing on this list that cannot wait for a committee.

What to pull together, who to ring, and by when. Today, before close of business: the annex, the entity names, the broker's master data, and a written answer to which bucket you are in. This week: the China Rubber Industry Association and your own trade counsel, on the comment deadline and on whether anything in the disclosed calculations is factually wrong about your company. Before the end of the month: your EU customers, with a revised landed cost and a proposed contract clause. Before the first post-5 December arrival: one test declaration, checked line by line, so the first container that meets the new measure is not also the first container that teaches you how it works. The cargo can be saved. Not by a strategy document, by a name, a code, and this afternoon.

↑ Back to the key points

— By Derek Xu