Transpacific box rates defy the season, Asia-US East Coast holds near $9,600/FEU ahead of Golden Week

The Analysis

Asia to the US East Coast printed at roughly $9,600 per FEU in the week of 25 September. Week on week, essentially flat. In a normal year that sentence is the least interesting thing on the page, because the last week of September is exactly when trans-Pacific pricing starts to leak ahead of the holiday lull. It did not leak.

The same week, Asia to the US West Coast rose 4% to above $8,100 per FEU. Asia to North Europe fell 15% to about $3,700. Asia to the Mediterranean fell 7% to about $3,900. Four prints, one week, and only one of them is behaving the way the calendar says it should.

So the question from shippers this week is not whether rates are high. Everyone can see they are high. The question is whether to book at a number that has refused to fall, or wait two weeks for a holiday lull that may or may not arrive. I have a view, and it starts with the print itself rather than with the forecast.

A flat print in a week that should have been down is not a flat market. Read it properly. If the seasonal norm for the last week of September is a decline of five to eight percent into the holiday lull, then a print that holds at $9,600 has, on a seasonally adjusted basis, gone up by that same five to eight percent. Call it $10,100 to $10,400 against where the calendar said you would be. Nothing in the announcement says this, because the announcement compares this week with last week and last week was already abnormal. The base you should compare against is not last week. It is the same week in a year that behaved normally, and that comparison is the one that quietly costs you money if you skip it.

The second problem is what the $9,600 actually contains. A container rate quoted on a trans-Pacific lane is a base ocean freight plus a stack of separately invoiced items: bunker adjustment, terminal handling at both ends, documentation, and whatever peak season surcharge the carrier has on file for that week. When a forwarder says 9,600, ask which of those are inside the number and which are added at invoice. The gap between a quoted base and an all-in figure is routinely the difference between a shipment that hits its landed cost target and one that misses it by a margin nobody budgeted. I have watched shippers model a rate and then discover at invoice that the number they modelled was three components short. Model the invoice, not the quote.

Then there is the clock on the quote. Most spot quotes on this lane are good for 24 to 72 hours, sometimes less when the market is moving. That makes the quote an option with a very short expiry, and an option you did not exercise is not a price you can plan against. Ask for the validity window in writing. If the forwarder will not put a window on it, that is itself information about how confident they are in the number. A quote with no expiry is a number with no owner. While you are at it, put an escalation trigger into your own customer quotes: if the band's pessimistic end is touched, you have a pre-agreed mechanism to pass part of it through rather than silently absorbing the whole move. The mechanism has to exist before the move, otherwise it is not a mechanism, it is an argument.

Two things are holding this market up, and they are different in kind. The first is demand timing. Shippers are pulling cargo forward ahead of China National Day Golden Week in early October, so bookings that would normally sit in the October book are being compressed into the last week of September. That is not extra demand. It is demand that arrived early, and it will leave a hole behind it. The second is supply management. Carriers are withdrawing capacity through blanked sailings, which removes the slack that would normally let the rate slide once the front-loaded demand passes. One of those forces is temporary by construction. The other is a decision that gets remade every week, which is precisely why it is worth tracking separately.

This is why I keep telling people to watch the blank sailing count rather than the rate. The rate tells you what happened. The count tells you what is about to happen. Every blanked sailing removes one string's weekly allocation from the market, and allocation is what you are actually buying when the market is tight. A published rate of $9,600 with twelve blanks announced over the next four weeks is a different market from $9,600 with four blanks, even though the headline number is identical on both days. Carriers publish these weekly and they are not secret. If your forwarder is not sending you the count without being asked, you are paying for a service you are not receiving.

Here is the part that gets lost in every rate story. Your capacity is not the market's capacity. It is your forwarder's allocation on the strings they actually hold. If you need 40 FEU inside a fixed two-week window and one of the three strings your forwarder uses gets blanked, the market rate is irrelevant to you, because there is no space at any price on the string that had your booking. That is the failure mode this season, and it is a booking failure rather than a budget failure. Budgets can be revised on a Friday afternoon. A missed sailing in the last week of September cannot.

Now look at the other side of the same week, because it carries the forecasting value. North Europe came off 15% to about $3,700 and the Mediterranean came off 7% to about $3,900, while the Pacific held. A fifteen-point spread between two trades in the same week is not noise. It is a statement about where tonnage wants to go. Carriers do not leave assets sitting on a lane that is falling 15% when another lane is flat at a level that pays. If the European slide continues, redeployment into the Pacific becomes the single largest bearish input on this market, and it arrives with a lag of four to eight weeks, because a string has to be re-slotted, re-marketed and re-contracted before it physically moves.

Which brings me to the strangest link in this week's chain. A possible strike at German ports is, for a trans-Pacific shipper, a bullish input. Not because it touches your cargo, but because it would put a floor under European rates, slow the fall, and delay exactly the redeployment described above. So the thing an importer trying to buy space cheaply in October should be rooting against is a quick settlement at the German ports. That is an uncomfortable sentence to write and it is the correct one. Watch German port news for the same reason you watch the blank count: it moves your number, just from further away.

The impact lands differently depending on what is inside the box, and the differences are wide enough to plan against. Seasonal retail cargo with a fixed Black Friday shelf date has zero tolerance for a missed week, because the shelf date does not move and the markdown is total. For that cargo the rate is a rounding error against the cost of arriving late. Lock the space and stop negotiating. A shipper who saves $400 a FEU and lands on 28 November has not saved $400, they have destroyed the margin on that container and probably on the programme that was counting on it.

Industrial inputs and raw materials are the opposite case. A one-week slip usually costs a production schedule adjustment rather than a lost sale, and the rate spread between now and mid-October is real money at volume. This is the category that should be allowed to wait. It is also the category where waiting is most often done badly, because the same team that decides to wait forgets to keep the allocation warm. Waiting is a decision that requires a booking you can activate on notice, not the absence of a booking. There is a difference between optionality and having nothing.

Cross-border e-commerce on fulfilment deadlines sits in between and is the most exposed to one specific mistake. The US West Coast is cheaper and faster, at $8,100 against $9,600, so the routing logic pushes everything west. That logic is correct until a West Coast string blanks, at which point the inventory plan has no answer. Keep a written fallback to the East Coast or to interior point intermodal routing, with the transit delta and the cost delta both on the page, so the call can be made in an hour by whoever is on shift rather than in a week by committee. The fallback that lives in someone's head is not a fallback.

By origin, the difference between Shenzhen and Ningbo this week is less about rate than about which strings each gateway can reach, and that is a forwarder question rather than a market question. By destination the split is wider. The East Coast direct call at $9,600 against the West Coast at $8,100 is a spread of roughly $1,500 a FEU, and that spread is the price of a landbridge you do not have to buy. Add the inland leg out of the West Coast plus the dwell at the rail ramp, and the spread narrows by an amount that varies by destination city. For Chicago or Columbus the gap can close to a few hundred dollars. For the Southeast it rarely closes at all. Run that comparison for your own door, not for the port.

One more number deserves a second look, and it is the one nobody quoted this week. The 4% move on the West Coast is smaller in dollars than the 15% move on North Europe, but it is the more informative print, because the West Coast is where a carrier adds capacity most cheaply when it wants to chase volume. A rising West Coast against a flat East Coast means the cheap lane is filling first. If you are routing through the West Coast on the assumption that it is the soft option, that assumption was true in August and is getting less true every week. The $1,500 spread between the two coasts is not a fixed feature of the market. It is a price, and prices move.

Timing matters more than the level, so let me put dates on it. The first wave is immediate and it is already running. Spot quotes on this lane last 24 to 72 hours, so anything you have not booked this week is being priced in a market you cannot see. If you have cargo that must sail before the holiday, the decision window has effectively closed and you are now buying whatever is left on someone else's allocation.

The second wave is contract repricing, and it lands in the four to six weeks after the holiday. Carriers use the post-holiday period to reset the level for the next quarter, and the argument they will make is that the market held through a period when it should have softened. That argument is stronger than usual this year, because a flat September print is evidence they can point at in a meeting. If you have a contract renewal in October or November, expect the opening ask to be built on the September print rather than on October reality, and prepare your counter with the blank count and the Europe print rather than with opinion.

The third wave is structural and it decides Q4. If the post-holiday lull is real, the blanked capacity returns and rates ease through November. If it is not, carriers hold the blanks and the market sits at this level into the pre-Chinese New Year build. Those two outcomes differ by more than a thousand dollars a FEU, and the tell is visible in the second and third week of October, not in any forecast published today.

We should do the arithmetic, because percentages hide the size of this. Assume a shipper moving 40 FEU a month into the US East Coast with an average cargo value of $30,000 a FEU, so about $1.2 million of goods a month. At $9,600 the freight bill is $384,000. That is 32% of the value of the goods. At the pessimistic end of a band, say $10,500, the freight bill is $420,000, or 35% of the goods. A three-point swing in freight as a share of landed value is not a rounding error. For a distributor running low double-digit gross margins, it is a large fraction of the year's profit. Every one of those figures is an assumption I have set out loud, so replace them with your own before you use the conclusion.

Now the same arithmetic at volume, because the shape changes. Assume 200 FEU a month at an average cargo value of $25,000 a FEU, so $5 million of goods. At $9,600 the freight is $1.92 million, or 38% of the goods. At $10,500 it is $2.1 million, or 42%. A four-point swing on $5 million of monthly goods is roughly $210,000 a month of margin moving for reasons that have nothing to do with your selling price and nothing to do with your operations. At that scale the answer is not a better forecast. It is a contract structure: a published base with an index linkage, so that a spike is shared rather than absorbed, and a defined band inside which nobody reopens the negotiation. The band is the product. The forecast is decoration.

The point I have not seen anyone make this week is about what a flat print means mechanically. A price that stops moving in a market with fixed supply and rising demand is a price that has stopped clearing the market. When price stops clearing, something else has to, and on a container lane that something is allocation. That is the regime change worth planning for. You can no longer solve a space problem with money at the margin, because the marginal buyer is not being served at any price. Shippers who are still budgeting their way through this are solving last month's problem with this month's money.

The second thing the headline hides is that an average is the wrong statistic for a market in this state. The published number blends bookings that closed weeks ago on allocated or contracted space with spot bookings made this week, and those two populations are moving in opposite directions. A flat average sitting on top of a rising spot is exactly what a market looks like in the week before it gaps. If your forwarder can tell you what the spot-only number did this week, that is worth more than the published print. If they cannot, that is worth knowing as well.

Here is the case against my own read. If the German port situation resolves without a strike, European rates keep sliding from $3,700, and carriers start moving tonnage into the Pacific earlier than the four-to-eight-week lag I assumed, then the Pacific flat print breaks downward in the first half of October, well before any holiday lull would explain it. Under that scenario, shippers who locked this week paid a premium for insurance they did not need, and the correct action was patience. The conditions that would confirm it are visible and weekly: North Europe below roughly $3,500, the blank sailing count falling for two consecutive weeks, and the East Coast to West Coast spread narrowing rather than holding.

The case in the other direction is simpler and I weight it more heavily. If the front-loaded demand does not leave a hole behind it, the market holds through October, carriers keep the blanks in, and the next repricing event is upward into the pre-Chinese New Year window. Under that scenario the cost of having waited is not the difference between $9,600 and $9,000. It is the cost of the cargo that did not sail, which has no ceiling. Asymmetric bets should be priced as asymmetric bets. This one is: the upside of waiting is a few hundred dollars a box, and the downside is a season.

Four things to watch, all published weekly, none of them requiring a forecast. The blank sailing count, because it leads and the rate follows. The East Coast to West Coast spread, because a widening spread means carriers are shifting tonnage coast to coast while a narrowing one means the East Coast is softening. The North Europe print, because it is the earliest read on redeployment. And any carrier notice restricting heavy or out-of-gauge bookings, because restrictions on the awkward cargo are usually the first sign a string is full and the last thing anyone tells a small shipper.

None of this needs a view on where the market goes next month. It needs a view on which of your containers can afford to be wrong, and pricing the ones that cannot accordingly. The teams that come out of this quarter intact are rarely the ones with the best forecast. They are the ones who wrote the pessimistic number into the customer quote, kept the space, and then went to lunch. The forecast, as usual, will be revised on Monday.

↑ Back to the key points

— By Vivian Zhao