USTR Port Fee Pause on China-Linked Ships Ends Nov 9; Talks Point to a Jan 10 Delay

The Analysis

The rule on the books says the pause ends on November 9, 2026. As of October 22, USTR has published nothing extending it. That is the story, and it is the part most coverage walks straight past, because the interesting conversation is about what people expect rather than about what the paper says. Platts reported on October 19 and 20 that Washington is expected to push the pause out by roughly two months, to January 10, 2027, and that US coal exporters are worried about freight getting dearer. Expectation is not publication. Absent a notice, the schedule reverts to the one that was drafted: USD 50 per net ton for vessels owned or operated by Chinese entities, and, for a non-Chinese operator running a China-built ship, USD 18 per net ton or USD 120 per container, whichever is higher.

Two tiers, four numbers, and one unit of measure that decides everything else: the charge is built on the ship, not on your cargo. Take that seriously and the same box can land anywhere between zero and well past six hundred dollars. Ignore it and you will budget one hundred and twenty, because that is the number that made the headline.

Hold one thing before the arithmetic starts. The date that binds you is not November 9. It is the day your cargo lands, and for anything booked out of Asia from today onward that day already sits past November 9, whether or not the extension arrives.

Start with the facts, and only with the facts. The USTR maritime action under Section 301 put a port fee on China-linked vessels and then paused it. The pause carries an end date on its face: November 9, 2026. On October 19 and 20, Platts reported that the pause is expected to be extended by roughly two months, to January 10, 2027, and that US coal exporters are worried about freight getting more expensive. As of October 22, no extension notice has been issued. Three facts, three dates, and only one of them is a legal fact, which is the end date. The other two are reporting about intentions, and intentions do not move a fee schedule.

Two intervals carry the whole decision. Eighteen days separate today from November 9. Sixty-two days separate November 9 from January 10. They behave in opposite ways. Eighteen days is shorter than any booking cycle you run, which means you cannot wait out the answer and still book normally. Sixty-two days is long enough to cover an entire peak-season arrival window, which means the extension, if it comes, is not a technicality. Your money sits in the space between those two numbers.

Why now, and who is doing the pushing. Follow the loudest American voice on this and you do not land on a container importer, you land on coal. Platts has US coal exporters worried about freight getting more expensive, and that worry is about capacity rather than about a fee line. A per-net-ton charge on China-linked tonnage makes US calls uneconomic for that tonnage, and tonnage that stops calling is tonnage that leaves the market. For a bulk shipper loading at a US terminal the cost does not arrive as a USTR invoice. It arrives as a fixture that is harder to cover and dearer once you find it.

Now take the number apart, because the number is where this whole thing gets mispriced. The figure circulating everywhere is USD 120 per container, and it is the figure that will end up in your budget if you let it. It is a floor, not an estimate. The rule reads USD 18 per net ton or USD 120 per container, whichever is higher. Read that as a one-way ratchet: the payer computes both branches and pays the larger one. The per-container branch only binds when the container count is high enough that 120 dollars a box outweighs 18 dollars a net ton. Below that count the net-ton branch binds, and every container taken off the discharge list makes the containers that remain more expensive.

Put a vessel on it. Assume a China-built box ship of 40,000 net tons, run by a non-Chinese operator, discharging boxes at one US port on one call. The two branches are 18 times 40,000, which is USD 720,000, and 120 times the box count. For the per-container branch to win, the box count has to reach at least 6,000, because 6,000 boxes at 120 dollars is USD 720,000, exactly level. That break-even is not a coincidence. It is 18 divided by 120, which is 0.15, times the net tonnage. Discharge 6,000 boxes or more and the per-box charge is 120 dollars. Discharge 3,000 and the fee is still USD 720,000, which spread over 3,000 boxes is 240 dollars per box. Discharge 1,200 and it is 600 dollars per box. The lighter the call, the heavier the box.

The formula underneath all of this is short enough to keep on a sticky note: the per-box charge equals the larger of 18 times net tonnage and 120 times boxes, divided by boxes. Because the net-ton branch binds below the break-even, the working version is 18 times net tonnage divided by boxes. It scales straight up with the size of the hull and straight down with the number of boxes. Assume a 55,000 net ton ship instead of 40,000: the fee becomes USD 990,000, the break-even box count rises to 8,250, and at 3,000 boxes discharged the per-box charge is 330 dollars. Assume a 25,000 net ton feeder instead: the fee is USD 450,000, break-even falls to 3,750 boxes, and at 3,000 boxes the per-box charge is 150 dollars. Same rule, same day, three hulls, anywhere from 150 to 330 dollars. Anyone quoting you a single per-box figure without telling you the tonnage and the discharge count is quoting you a guess.

Note what that does to ordinary commercial behaviour. A carrier trimming a US port out of a rotation, consolidating two calls into one, or cutting a low-volume discharge is doing something that looks like efficiency and reads on your invoice as a price increase. Under the 40,000 net ton assumption, going from 3,000 boxes to 1,200 boxes on the same hull takes the per-box charge from 240 dollars to 600 dollars without a single word of the regulation changing. Nobody has to announce anything. The arithmetic does it for them.

The other tier has no ratchet and no branch that can be argued downward. USD 50 per net ton, full stop, for a vessel owned or operated by a Chinese entity. Same 40,000 net ton hull: 50 times 40,000 is USD 2,000,000 on a single call. Over 3,000 boxes that is 667 dollars a box. Over 1,200 boxes it is 1,667 dollars a box. Fifty is 2.78 times eighteen, so the ownership tier costs nearly three times the build tier for identical steel, and nothing in the 50-dollar tier offers a per-container ceiling that a shipper can hold the carrier to.

Bulk has no container branch at all, which makes it both the cleanest case and the ugliest. A coal cargo discharges no containers, so the 120-dollar alternative computes to zero and the entire per-net-ton figure lands on the cargo. Assume a China-built Panamax of 33,000 net tons lifting 60,000 tonnes of coal from a US terminal. The fee is 18 times 33,000, which is USD 594,000, and across 60,000 tonnes that is 9.9 dollars per tonne. Nine dollars ninety on a bulk commodity is not a rounding error, and it lands whether or not the cargo ever touches a container.

Layer it by who is standing where. A US importer on a Chinese-operator service faces the 50-dollar tier. A US importer on a non-Chinese operator running Chinese-built hulls faces the 18-dollar tier with a 120-dollar floor. A US importer on tonnage that is neither Chinese-owned and operated nor Chinese-built pays nothing at all under this action. A US coal exporter pays through the chartering market rather than through the fee. Four sets of facts, four different numbers, and the distance between the cheapest and the dearest is not a matter of a few percent. On the assumptions above the range runs from zero to 1,667 dollars a box.

Sorting which tier a hull sits in takes three pieces of information: the IMO number, the entity that actually operates the vessel, and the yard that built it. The IMO number leads to the build yard. The operator is a question about the management contract and the commercial control of the vessel, not about the flag painted on the stern, and those two are frequently not the same party. Assume a vessel flying a non-Chinese flag, operated by a non-Chinese company, built in a Chinese yard: that is the 18-dollar tier. If the same hull's operator is treated as a Chinese entity, it jumps to the 50-dollar tier, which is 32 dollars more per net ton on the same steel. Nobody in your organisation is automatically doing that classification. Ask.

Transmission runs slower than the headline and faster than the invoice. The rule revives on November 10 if nothing is published. Broker pricing moves on the day the expectation moves, and the expectation already moved. A booking made this week sails in seven to fourteen days. Asia to the US West Coast runs twenty-eight to thirty-five days. The charge is assessed when the vessel calls. Assume the carrier issues its surcharge notice ten to thirty days after the call, an assumption you should check against your own service contract, but it is the range most shippers recognise. Stack those intervals and the cash leaves your account somewhere between forty-five and seventy days after November 10, which puts it in the second half of December and into January. Your Q1 tender is being negotiated in exactly that window, and those two facts have not yet been put in the same sentence anywhere.

Here is the point the coverage has not made, and it is the one that will actually cost money. The charge is built on net tonnage, which means it is built on the vessel. It is not built on your bill of lading. A per-net-ton fee produces one number per vessel call and says nothing whatsoever about how that number is divided among the three thousand boxes underneath it. The rule sets the total. The allocation is a commercial decision, and it lands in your contract and on your paperwork, not in the regulation. Per container discharged, per container loaded, per TEU booked, per bill of lading, or spread across every call on the string: each basis produces a different number for identical cargo, and on a part-load the gap between a per-box basis and a per-bill basis is not small.

Derek Xu lives at this end of the problem, and the answer he would give is the one I am giving you: if your service contract does not state which basis applies, the basis is whatever appears on the invoice, and you will be arguing about it in January against a vessel that sailed in November. Get the allocation basis into the contract before the sailing, because after the sailing you are negotiating a number somebody else has already computed.

The invoice end deserves the same attention. When the surcharge shows up it usually arrives as one line and one total, and no carrier is obliged to hand you the allocation working. Take three checks to it. Confirm the IMO number on the invoice is the vessel that actually carried the cargo, which on a service with substitutions is not a formality. Confirm the net tonnage against the classification society record. Confirm the allocation basis is the one written in your contract. If any one of the three fails, raise the dispute before payment rather than after it, because a fee dispute that crosses the payment terms stops being a commercial conversation and becomes a finance one, and the room to move shrinks considerably.

The contrary case is easy to state and it is not a weak one. If USTR publishes the extension to January 10, none of this bites before January 10, and anyone who locked a capped passthrough in late October bought insurance that never paid out. That is a real cost and I am not going to pretend it away. What makes me comfortable carrying it is the asymmetry. The extension is worth sixty-two days of clean arrivals. The non-extension is worth a per-box charge running from 120 to 667 dollars on modest assumptions, plus a tender season negotiated while the invoices are landing. Exactly one condition flips the conclusion: publication. The moment a dated notice appears in the Federal Register, the exposure window moves to that date and this analysis restarts from there. Until it does, the paper rule is the rule, and the paper rule says November 9.

Two more conditions can move the number without any notice at all. Box count at discharge moves it, as shown: below 0.15 times net tonnage the per-box cost rises as boxes fall, so a carrier dropping a US port from a rotation makes the remaining ports worse for everyone on the hull. Fleet mix moves it as well. If Chinese-linked tonnage withdraws from US calls, the hulls replacing it may sit outside both tiers entirely, which takes the fee to zero and puts the cost back into freight as scarcity. That is the coal exporters' problem restated in container terms, and it is why a freight quote and a fee exposure should be read as one number rather than two.

So where is the threshold, and when do you lock. Take a figure most importers will recognise: assume your all-in Asia to US West Coast rate is USD 1,800 per box. Under the 40,000 net ton and 3,000 box assumption, the build tier puts 240 dollars on that box, which is 13.3 percent of the freight. At 1,200 boxes it is 600 dollars, or 33.3 percent. My own trigger sits at ten percent: once a single regulatory passthrough can move more than a tenth of the freight bill, I stop treating it as a surcharge and start treating it as a contract term. At 1,800 dollars a box, ten percent is 180 dollars, and the 3,000-box case is already past it at 240. You cross the trigger on the plausible case, not on the worst case.

Scale it to a shipping programme. Assume you move 40 boxes a month on a service sitting in the build tier, with 3,000 boxes discharged per call on a 40,000 net ton hull, which puts the fee at 240 dollars a box. That is 9,600 dollars a month. The exposed arrival window between November 9 and January 10 is roughly two months of arrivals, so call it about 19,200 dollars across the gap. Now assume the operator on that same service is treated as a Chinese entity and the same 40 boxes move at 667 dollars: 26,680 dollars a month, roughly 53,000 across the gap. The distance between those two outcomes is 34,000 dollars on a 40-box-a-month account, and it turns on one classification question you can settle this week with three emails.

On what to actually do: the move is not to rebook, it is to bound. Ask the carrier for three things in writing before the next sailing: which tier the vessel sits in, what its net tonnage is, and which allocation basis will be used. Those three answers collapse a range running from zero to 1,667 dollars into a single number you can plan against. Then cap it. A passthrough clause carrying a dollar ceiling per box and a named basis survives both outcomes, because if the extension lands the ceiling never binds, and if it does not you are holding a number you already budgeted.

The traps are where the money actually leaves. Budgeting 120 dollars a box because that is the number in the headline is the expensive one, since the plausible case under these assumptions runs two to five times that. Waiting for the notice is the second, and it fails on timing: the notice, when it comes, arrives after your booking window has closed. Accepting a passthrough clause with no basis and no ceiling is the third, and it is the same contract you had before you asked. Treating the fee as separate from freight is the fourth. If Chinese-linked tonnage exits the trade, the freight increase and the fee are substitutes, and a deal that caps one while leaving the other open has capped nothing at all.

Rules are the only counterparty in shipping that keeps its promises by default. This one expires on November 9 unless somebody turns up and stops it, which is more than can be said for most rate filings. Nobody at USTR is going to phone you on November 8 to check whether you were ready. The day to lock is not November 9 and it is not January 10. It is the day before your next booking closes, and for most people reading this that is this week.

↑ Back to the key points

— By Vivian Zhao