Transpacific Freight Rates Surge 2026: China–US Rates Near Pandemic Highs Into Golden Week

The off-contract rate to move a 40-foot container from China to the US East Coast reached $10,948 on 17 September 2026, according to freight pricing platform Xeneta — more than four times the level on the day the Iran war began on 28 February, and the first time this lane has traded back in pandemic territory since 2022. That is the direct answer to the question every China-based exporter and US importer is asking this week: the summer peak did not end and rates did not soften. They went up again, hard, even as Asia–Europe freight rates fell for the eighth consecutive week.

This article explains what is actually moving the transpacific into Golden Week — the bunker spike, the capacity carriers are withholding, congestion that has swallowed more than 4 million TEU at Asian ports, and a US import peak arriving three weeks later than any forecast expected — and what it does to your landed cost, your tariff exposure and your booking calendar between now and January. This is the biggest lane in container shipping: US ports are forecast to take 2.31 million TEU in September 2026, the busiest month of the year, and China-origin cargo still moves through the same corridors that Shipping from China to USA sets out port by port. If you read only two sections, read Section 5 (the Golden Week date math) and Section 6 (the tariff clock); those two decide what you do this week.

1. The twelve numbers that define this week

• $10,948 per 40ft — China to the US East Coast on 17 September 2026, the highest reading on the lane this year and a four-fold increase since 28 February (Xeneta, via Reuters).
• $10,373 per 40ft — Shanghai to New York on 24 September, flat week on week after a 7% jump the week before (Drewry World Container Index).
• $7,838 per 40ft — Shanghai to Los Angeles on 24 September, up 2% week on week and up 9% in three weeks.
• $4,468 per 40ft — the WCI composite on 24 September, down 1% — because a rising transpacific is being cancelled out by a falling Asia–Europe.
• 15 versus 9 — transpacific blank sailings announced for the week ahead against the week before, the sharpest one-week step-up since May (Drewry Container Capacity Insight).
• ~14% — share of planned sailings withdrawn on Asia–North America services ahead of Golden Week, with some market trackers putting pre-holiday withdrawals at 18–25%.
• $4,000 per 40ft — CMA CGM's peak season surcharge on Far East cargo to both US coasts from 1 October, up from $2,500 on Shanghai–Los Angeles — a surcharge filed on top of the base rate and any GRI.
• 4.3 million TEU — container capacity tied up in port congestion worldwide, above the roughly 4.0 million TEU stranded at the peak of the pandemic (Linerlytica).
• 12 days — the longest vessel waiting time recorded at Shanghai and Ningbo in September 2026, with delays spreading to South China and Southeast Asia.
• 2.31 million TEU — forecast US container imports for September 2026, up 9.6% year on year and the busiest month of the year (NRF and Hackett Associates Global Port Tracker).
• $835 per tonne — Singapore VLSFO on 17 September 2026, still more than 60% above pre-conflict levels while Brent is up about 40%.
• 10 January 2027 — the new expiry date of the US–China trade truce, extended from 10 November 2026 at the Washington summit on 24 September — while the 178 Section 301 product exclusions still expire on 9 November unless USTR publishes a separate notice.

2. What actually changed in the seven days to 28 September 2026

This was not a slow drift. The transpacific repriced, and the policy calendar moved underneath it.

Two data sources, two methods, one direction. Drewry's assessed spot rate and Xeneta's market average differ by a few hundred dollars on the same week, which is normal — Drewry prices named port pairs from Shanghai, Xeneta's index blends contracts and spot. Take the direction, not the decimal.

3. Why the transpacific is the most expensive lane out of China again

Four forces are working at the same time, and three of them are not going to disappear after Golden Week.

Source: Drewry World Container Index assessments for 3, 10, 17 and 24 September 2026. Xeneta put the China–US East Coast market average at $10,948 per 40ft on 17 September.

First, fuel. This is the driver most shippers under-price. Singapore VLSFO — the benchmark grade for most of the fleet — has been assessed between roughly $820 and $910 per tonne through September, against $433.50 on 1 January 2026. That is a rise of more than 100% in nine months, and fuel-oil inventories across Singapore, Rotterdam–Antwerp and Fujairah have run around 30% below three-year seasonal averages, with a projected global fuel-oil deficit of about 218,000 barrels a day in the third quarter. On a China–US East Coast rotation, the fuel bill on a 14,000 TEU vessel moves by several hundred thousand dollars a voyage when bunkers move like this. Carriers recover it through BAF and emergency bunker surcharges — line items that sit outside the base rate, and outside most of the quotes importers compare.

Second, capacity withheld on purpose. Carriers announced 15 transpacific blank sailings for the week ahead on 24 September, up from nine. That is capacity management aimed squarely at the pre-holiday booking window, and it is working: the rate rose into the withdrawal rather than through it. This is the structural difference between 2026 and the post-pandemic collapse. There is a large orderbook, but the injection is being metered by cancellations rather than allowed to flood the lane.

Third, congestion is eating capacity that carriers never withdrew. Linerlytica counts more than 4.3 million TEU of container capacity sitting in port congestion globally, above the level stranded at the pandemic's worst point. Shanghai's seven-day average vessel waiting time was measured above four days through early September, with individual terminals past nine days and Shanghai and Ningbo waits reaching 12 days at the peak; a day of waiting at origin is a day of lost voyage capacity, and it arrives as unreliable ETAs rather than as a line on your invoice.

Fourth, demand did not fade on schedule. NRF and Hackett Associates revised September US container imports up to 2.31 million TEU, making it the busiest month of 2026 — three weeks later than the peak any model expected, and after an August that had already been written off as the top. Port throughput confirms it: Los Angeles handled 955,907 TEU in August and Long Beach 919,992 TEU, both records for the month.

Put together, the arithmetic is straightforward: demand above forecast, capacity below deployment, and fuel that costs more than twice what it did in January. None of those three reverses on 8 October.

Then there is the fourth-order effect nobody plans for. When capacity is tight, carriers prefer the highest-yield boxes, which are the ones with the shortest inland leg and the cleanest paperwork. Cargo that needs an IPI rail move to the Midwest, cargo with a late documentation file, cargo from a supplier who delivers on the last working day before the cut-off — that is the cargo that gets rolled. In a market where a rollover plus a delayed departure adds close to two weeks, the practical cost of a bad file is now larger than the difference between two competing quotes.

4. What a $7,838 box actually costs you

The published base rate and the invoice you pay are two different numbers, and on this lane in 2026 the gap has widened to the point where comparing base rates is close to meaningless.

Base FAK rates to the US West Coast can still be quoted as low as $2,500–$3,500 per 40ft on some services. Nobody is moving a peak-season box at that number. Here is the stack that sits on top of it.

Now put real numbers against a container. Take a 40ft high cube of consumer goods, FOB value $40,000, about 700 units, moving Shanghai to Los Angeles in the current market:

Read the bottom two lines carefully, because they change how you should be managing this. Duty is now the largest single cost line on a China–US container — 53% of everything above the goods value, and larger than the ocean freight. Freight and inland together are about 43%.

That is also why the freight number still matters so much. Run the same shipment on a service priced at $4,000 ocean with an $800 surcharge, and the cost above the goods falls to $23,324 — $33.32 a unit, a saving of $10.05 per unit, or a bit over $7,000 on the container. At 40 containers a year, that is $280,000. Nobody negotiates that away by arguing over 5% on a base rate; they get it by booking the right service at the right time with the paperwork complete.

Three practical conclusions follow:

1. Compare all-in quotes, not base rates. Ask for a written breakdown of base, GRI, PSS, BAF, EIS, origin, destination and inland. On this market the surcharge stack can equal the base rate.
2. Treat quote validity as a risk line. Spot quotes in a market moving 2–7% a week are typically valid 7–14 days. If your quote says "subject to surcharges at time of shipment", you have bought nothing.
3. Ask what the quote excludes. Most published rates exclude destination free time, detention and demurrage, chassis, and any terminal storage after free time expires. With US terminals running high utilisation, a week of avoidable demurrage is a four-figure number per container.

If you are moving FCL, the service selection and port-pair economics for your destination are handled as a sea freight planning exercise rather than a rate auction; if you are shipping LCL, the equivalent calculation is per CBM, and the LCL surcharges move with the same market.

5. Golden Week 2026: the double-holiday squeeze and the date math

Everything above now collides with the Chinese holiday calendar, and 2026 is worse than usual because the Mid-Autumn Festival and National Day fall a few days apart.

Carriers have already pulled the sailings out of the schedule. Maersk has blanked transpacific voyages with Eastbound departures of 9 October out of Busan (TP8 640E) and 9 October out of Ningbo (TP12 641E), with inducement calls on alternative sailings to keep port coverage; Hapag-Lloyd has cancelled the WC2 Guthorm Maersk 640E out of Shanghai on 10 October and the US2 Navios Unison 641E out of Ningbo on 7 October, adding a Norfolk call on US1 to cover it. In the week to 24 September, transpacific cancellations jumped to 15 from nine, within a broader pattern that market trackers put at roughly 14% of planned Asia–North America sailings, and some estimates as high as 18–25%.

Note the shape of the problem. The sailings that were cancelled are the ones sitting immediately after the holiday — which is exactly where everyone who could not load before Golden Week now wants to be. Space on the 8–16 October window is the single most contested capacity on this lane for the rest of the year.

Our practical guidance for US-bound cargo over the next three weeks:

• If cargo is ready in the last week of September, gate it early in the 28–30 September window rather than on the last working day, and confirm the vessel and voyage number in writing before cutoff. Anything presented late in that window is a rollover candidate in a market where a rollover plus a slipped departure costs close to two weeks.
• If cargo cannot load before the holiday, stop aiming at the first post-holiday sailings and plan against the second half of October, then re-forecast your arrival dates instead of assuming them.
• If it is Amazon FBA replenishment, work backwards with the full chain: 28–30 days of ocean transit to the West Coast, 3–5 days to deconsolidate and dray, and then 7–14 days of fulfilment-centre intake lead time on top. Our Amazon FBA Shipping from China guide sets out the intake windows, and in a peak season this late, inventory that misses Black Friday does not simply arrive late — it arrives into a promotion you can no longer run.
• Book two to three weeks ahead, even for cargo that is not ready. Space secured early costs nothing; space found late costs rate plus rollover.
• Assume the usual post-holiday relief is delayed. Drewry expects transpacific rates to ease slightly into the holiday, but Linerlytica's warning is the one that matters: the backlog of cargo that could not load before Golden Week should keep vessels full through the holiday, which means the seasonal dip may not arrive until late October.

The mechanics behind this pattern, and the 2026 version of it, are set out in our Golden Week blank sailings analysis; the difference this year is that the withdrawals are landing on a lane that is already short of effective capacity.

6. The tariff clock running beside the freight clock

Freight is what is expensive this month. Tariff policy is what is dangerous this quarter, because two deadlines sit inside the booking window you are about to fill.

The headline from Washington helps, and less than the headlines suggest. At the 24 September summit, the Treasury confirmed the Busan trade truce extended by two months, from 10 November 2026 to 10 January 2027. What that extension covers is the suspension of new retaliatory tariffs, the pause on the Section 301 shipbuilding and maritime investigation, and the suspension of new export controls. What it does not do is move a single tariff rate you are paying today, and it does not automatically move the two dates that matter at entry level.

Three consequences you should act on now.

First, the 9 November exclusion cliff is a real date and it is close. If any of your HTS lines have been claiming the exclusions, the underlying Section 301 list rate returns on entry or warehouse withdrawal on or after 10 November — 25% on Lists 1 to 3, 7.5% on List 4A. The exposure is concentrated in industrial inputs and capital equipment: pumps, motors, compressors, specialty chemicals, wear parts and the solar-manufacturing tooling lines. Ask your broker to pull every entry line filed under 9903.88.69 since November 2025, sorted by entered value. Multiply by your list rate. That number is your exposure, line by line.

Second, bonded storage does not beat this deadline — it can trigger it. Duty is assessed at the rate in effect on the date of withdrawal for consumption, not the date the goods entered the warehouse. Covered inventory sitting in bond on 10 November loses the exclusion at withdrawal. If you hold covered goods in bond, put a withdrawal on the calendar for the first week of November with the broker's filing lead time built in. For containers that land after the deadline, the cleaner route is a bonded warehouse entry with partial withdrawals, so that a late or retroactive USTR notice can still be claimed and you are never forced into a full-rate entry during the weeks when USTR is most likely to publish.

Third, check your continuous bond, and check your entries generally. The 2026 changes are mostly compliance costs rather than headline duties: the de minimis exemption is gone for all countries including China, so every consignment — including a $5 sample — needs a proper entry, and commercial shipments up to $2,500 now move on a non-formal Type 11 entry while anything above that needs a formal Type 01 entry and a continuous bond. MPF resets on 1 October 2026 to a floor of $34.58 and a cap of $670.86 for fiscal 2027. ISF filing penalties have been raised to $5,000–$10,000, so a late or wrong 10+2 filing is no longer a nuisance cost. A 7.5-point increase on your China-origin volume raises annual duties, and under-bonded importers receive insufficiency letters that stop entries outright.

One scheduling point ties the two clocks together. 10 January 2027 sits inside the pre-Lunar New Year rush — Chinese New Year falls on 6 February 2027 — which is the worst possible week to be pulling volume forward: factories are completing year-end orders, space is tight, and rates rise. If your plan is to front-load ahead of a tariff date, the cheap window to do it is October and November, not the second week of January.

7. Air freight: the pressure valve, and what it costs to use it

Air is where US-bound shippers go when the ocean lane stops being predictable, and this quarter it is both tighter and more expensive than usual.

China–US air spot rates have moved above $7 per kilogram, roughly 20% higher than before the conflict began in late February, while China–Europe air rates have gone past $4.8 per kilogram, an increase of more than 75%. The reasons are the same ones pushing ocean rates, plus one that is unique to air: Middle East airspace disruption has removed an estimated 12–20% of global air cargo capacity through longer routings and cancelled frequencies, and aviation fuel costs have risen more than 90%. Carriers have responded with fuel surcharges across Asia that have moved up by 15–100% depending on the lane, and the surcharge table published by UPS puts the Shenzhen–North America fuel surcharge at a minimum of CNY 87 per kilogram as of mid-September, before the base rate.

Do the arithmetic before you assume air is the answer. On our earlier example — 700 units in a 40ft high cube, about 15 kg per unit — the ocean package costs roughly $17 a unit in freight terms, and air at $7–9 per kilogram costs $105–135 a unit, six to eight times more. That is a rational choice for launch stock, high-value electronics, a spare-part line, or inventory that has to be on a shelf before a specific promotion, and an irrational one for anything with retail margins below about 40%.

Where air genuinely pays right now is recovery. If your ocean cargo is stuck behind a cancelled 9 October sailing and your buyer needs stock, the comparison is not air versus ocean — it is air versus a lost season. Book air space five to seven days ahead of the intended uplift, expect five to ten days door to door into a major US airport, and confirm whether the quote is a spot rate or a blocked-space agreement, because in a peak window the difference is who gets on the aircraft when a freighter is oversold. Our air freight team quotes both the express and consolidated options with the surcharge structure stated up front, which is the only way to compare them honestly.

8. Where the congestion actually sits — origin, ocean and destination

Rate headlines hide the operational constraint that will actually decide whether your cargo arrives on time. In the current market it sits in three places at once.

At origin. Asian port congestion has absorbed more than 4 million TEU of capacity, with Shanghai and Ningbo waits reaching 12 days at the peak and the backlog from the August and early-September typhoon sequence still clearing. Shanghai's seven-day average waiting time was measured above four days through early September, with individual terminals past nine days and 57 Shanghai call omissions in one fortnight's sample as carriers moved calls to Ningbo. The practical effect is not that your container does not load; it is that your container loads on a different vessel than the one on your booking confirmation, and your ETA moves by a week without anyone telling you.

On the water. Panama Canal restrictions are still capping daily transits at 32 vessels, including nine Neopanamax slots, and carriers have added canal surcharges: CMA CGM from 10 September and MSC from 12 September, at around $500 per TEU. That is one reason US East Coast and Gulf rates are holding above the West Coast, and one reason to check whether a Panama-routed quote has the canal surcharge inside or outside the number you were given.

At destination. US West Coast gateways are handling record volumes — Los Angeles took 955,907 TEU in August and Long Beach 919,992 TEU, both records for the month — and the operational stress shows up as landside costs rather than berth queues. Port of Los Angeles dwell time for local import cargo has been averaging around 4.7 days, on-dock rail dwell around 4.2 days, and import units on the street four to six days. Free time is typically seven days at the terminal and five for the container, after which the meter is brutal: carrier detention of roughly $255–575 per box per day on a 40ft dry container, terminal storage of $52–211 per day according to the daily tier, and chassis charges of $22–55 per day.

Two planning decisions follow, and both are cheaper than a demurrage invoice:

• Pick the gateway by your inland destination, not by the headline ocean rate. An IPI rail move to the Midwest runs about 20–25 days and is one to two weeks faster than an all-water routing to the East Coast followed by trucking, even when the ocean leg to the East Coast looks more convenient on paper.
• Settle destination free time and chassis before the container sails. If your warehouse cannot take delivery inside the free window, negotiate extended free time into the rate, or book chassis separately. That conversation costs nothing in September and costs four figures in October.

9. Three scenarios for Q4 2026

We are planning on the base case and pricing the risk case. The reason is asymmetry: the downside of being wrong about the base case is a few hundred dollars a container, and the downside of being wrong about the risk case — with cargo mid-Pacific and a retail commitment in December — is a missed season.

10. A booking plan for the next 30 days

The single highest-value action in that table costs nothing: put your cargo readiness dates and your promotional calendar in front of your forwarder this week, and ask which sailing each shipment is intended to catch. In a market where the difference between a good and a poor booking is $10 a unit and two weeks of transit, the information flow between importer, supplier and forwarder is now worth more than any single rate negotiation.

11. How to quote this market: five questions, in writing

Because the gap between the published base rate and the payable rate is now several thousand dollars a container, what you ask determines what you pay. We recommend putting these five questions to every forwarder on every transpacific quotation this quarter.

1. What is the all-in figure, itemised — base, GRI, PSS, BAF or emergency bunker, EIS, origin charges, destination charges and inland? And which of those lines are fixed for the validity period, and which are "subject to change at time of shipment"? A quote that does not separate them is a quote you cannot budget.
2. Which carrier, service, vessel and voyage is this rate on — and has that voyage already been blanked? With 15 transpacific cancellations announced in a single week, a rate without a voyage number is an intention, not a booking.
3. How many days is the rate valid, and what happens if my container rolls? Ask for the validity in days, the rollover policy, and whether a rolled container keeps its original rate or is repriced at the next GRI.
4. What are the free time, detention and demurrage terms at destination, and what does an extra day cost? US terminals are running high utilisation, chassis are tight, and the meter after free time is $255–575 a day for the container. This is a cost you can negotiate in September and cannot in October.
5. If the routing changes — a Panama restriction, a rotation moving back to Suez, a discharge port omitted — what happens to my transit time and my rate? The 2026 network has been rewritten twice in six months. The answer tells you whether your forwarder is monitoring the lane or just quoting it.

There is a sixth question for anyone importing goods covered by a Section 301 exclusion, and it is a November question being asked in September: if this container is withdrawn for consumption after 9 November, who is watching the exclusion claim? The answer should be your broker, in writing.

12. The AllBestShipping view

We are telling our clients three things this week.

Budget the all-in number, not the base rate. The peak season surcharge CMA CGM filed for 1 October is $4,000 per 40ft; other carriers are filing $300–1,000. That spread is not a negotiation signal, it is a routing and service signal — and it is invisible if you compare headline rates. We quote US-bound cargo with the surcharge stack itemised line by line, including origin, destination, inland and the tariff costs that land at the other end, so the number you approve is the number you pay.

Plan from the sailing, not the schedule. In a market with 15 weekly cancellations and a two-week rollover cost, the useful question is not "what is the rate to Los Angeles this week" but "which vessel is my cargo actually on, and does that vessel still exist on the day it sails". We book across multiple carriers on the transpacific, which is what allows us to move a client's order to a surviving sailing instead of explaining why it did not move.

Treat the tariff calendar as part of the booking file. The 9 November exclusion deadline and the 10 January truce date are not customs-team trivia; they decide whether you pull volume forward, how you use a bonded warehouse, and which month your cash needs to be ready. We handle that calendar alongside the ocean booking, because a container that arrives on the wrong side of a deadline costs more than the freight on it.

If you want a same-week read on your own shipments — current all-in rates by service, the sailings that survived the Golden Week blankings, transit times to your specific gateway and the tariff dates that touch your HTS lines — send us your supplier list, your cargo readiness dates and your promotional calendar. We will come back with the sailing plan and the landed-cost math, including a straight answer on which of your shipments should fly. Start with AllBestShipping or send the details through our contact page and we will run the lane for you.

FAQ

1. Why are transpacific freight rates so high in September 2026?

Four forces are working at once. Bunker fuel: Singapore VLSFO has been assessed between roughly $820 and $910 per tonne through September against $433.50 on 1 January, so fuel surcharges have roughly doubled their contribution to the voyage cost. Capacity: carriers blanked 15 transpacific sailings for the week ahead of the 24 September assessment, up from nine. Congestion: more than 4 million TEU of capacity is tied up in port congestion, above the pandemic peak, with Shanghai and Ningbo waits reaching 12 days. And demand: NRF and Hackett Associates revised September US container imports up to 2.31 million TEU, the busiest month of 2026 and later than any forecast expected.

2. How much does it cost to ship a 40ft container from China to the US right now?

As of the Drewry assessment on 24 September 2026, Shanghai to Los Angeles was $7,838 per 40ft and Shanghai to New York $10,373, with Xeneta's market average for China to the US East Coast at $10,948 on 17 September. Expect the payable figure to sit above the base rate you are quoted: GRIs, peak season surcharges (CMA CGM filed $4,000 per 40ft from 1 October), bunker adjustments, equipment imbalance and congestion surcharges typically add 25–40%, and more where a carrier has filed an aggressive peak surcharge. Always ask for the all-in number with each line itemised.

3. Will rates fall after Golden Week 2026?

Probably, but later and more slowly than in a normal year. Drewry expected transpacific rates to ease slightly into the holiday, and the classic post-holiday demand dip still follows. The counterweights are real: the backlog of cargo that could not load before 1–7 October should keep vessels full through the holiday fortnight, carriers are managing capacity with blank sailings rather than releasing it, and bunker costs remain far above their January levels. Our working assumption is rates firm into mid-October, then a 10–20% drift down into December, with a floor set by fuel and capacity discipline.

4. What is the last date to ship from China to arrive in the US before Black Friday 2026?

Work backwards from the retail date, not forwards from the factory. Black Friday falls on 27 November 2026. With a 28–30 day transit to the US West Coast, the last sensible sail date is around 28 October; for the East Coast, on a 35–40 day transit, it is around 18 October. Then subtract three to five days for deconsolidation and inland delivery, and, for Amazon FBA sellers, another 7–14 days of fulfilment-centre intake lead time — which pushes the effective cut-off for FBA replenishment into the first half of October.

5. Did the US–China trade truce extension change the tariffs I pay on Chinese goods?

No. The 24 September summit extended the Busan framework by two months, from 10 November 2026 to 10 January 2027, which keeps existing suspensions in place: no new retaliatory tariffs, a pause on the Section 301 shipbuilding and maritime investigation, and no new export controls. The rate you calculate on a Chinese import today is unchanged — base MFN duty plus the Section 301 forced-labor tariff at 12.5%, plus any product-specific Section 301 rate (commonly 25%), plus Section 232 metals duties at 50% where they apply. Separately, the 178 remaining Section 301 product exclusions still expire at 11:59 p.m. ET on 9 November 2026 unless USTR publishes its own notice, and the pending Section 301 excess-capacity action, reported at 7.5%, is on a different track entirely.

6. Should I switch to air freight while ocean rates are this high?

Only for cargo where speed pays for itself. China–US air spot rates are above $7 per kilogram — roughly six to eight times the per-kilogram ocean cost on the same goods — and air capacity is 12–20% tighter than normal because of Middle East airspace disruption, with aviation fuel up more than 90%. Air is the right answer for launch stock, high-value electronics, spare parts and inventory that has to be on a shelf for a dated promotion. For anything with retail margins under about 40%, the correct response to a tight ocean market is earlier booking, better documentation and a gateway chosen for your inland destination — not a modal switch.