Port of Los Angeles Forecasts 7% Container Volume Decline to 9.3 Million TEUs, China Share Falls to 40%

The Analysis

Forty percent. Out of everything the Port of Los Angeles put on the table when it approved its three point four billion dollar budget for fiscal 2026 to 2027, that is the figure I keep coming back to, and I suspect it is not the one most people will quote. Most people will quote nine point three million TEUs, or the seven percent decline, or the three point four billion. I am not dismissing any of those. But forty percent is the one that redraws my supplier map, and I am the one who has to sign the purchase orders that live on the map it draws.

Let me be plain about which desk I sit at, because the same release reads completely differently depending on where you are standing. I buy. I do not run vessels, I do not set tariffs, and I am not going to tell anyone how to run a port. What I do is decide, for a given category in a given season, which factory gets the order, how many units go on it, and what I can charge for them when they land. China origin accounted for sixty-one percent of containerized imports through Los Angeles in 2020, fifty-three point four percent in 2025, and roughly forty percent now. That is twenty-one points in six years, and thirteen point four of those points came off in a single year. Somebody else will read that as trade policy. I read it as a sourcing book that moved underneath my feet while I was still pricing against the old one.

The piece I want to separate out is the difference between the port's number and mine. The port's headline is seven percent down to nine point three million TEUs. My number is different, and it comes out of the same two figures. If seven percent is measured against a base of about ten million TEUs, then China origin boxes through this port went from roughly five point three four million last year to about three point seven two million this year, because nine point three million times forty percent is three point seven two million. That is a drop of one point six two million TEUs, or thirty point three percent. A seven percent decline at the gateway is what the port sees. A thirty percent decline in the China-origin pool moving through it is what I see. Those are not the same business, and only one of them tells me what to do on Monday.

Start with the book, because none of this means anything until it is sitting next to my own numbers. Assume an importer of home goods running twelve hundred TEUs a year through Los Angeles at an average declared FOB of seventy-eight thousand dollars a container. That is ninety-three point six million dollars of FOB a year. Assume also, and this is my assumption rather than anything in the release, that each box carries seven hundred and fifty units, which puts unit FOB at one hundred and four dollars. Last year, at a China share of fifty-three point four percent, six hundred and forty point eight of those boxes were China origin, carrying forty-nine million nine hundred and eighty thousand dollars of FOB. This year, at forty percent, four hundred and eighty boxes carry thirty-seven million four hundred and forty thousand. The difference is one hundred and sixty point eight boxes and twelve million five hundred and forty thousand dollars of FOB that no longer sits behind a China origin label.

What is that worth? Here I have to be explicit, because nothing in the release gives me a tariff rate and I am not going to invent one. Assume the tariff line applicable to China origin sits twenty percentage points above the line applicable to wherever that volume went. I am not saying it is twenty points. I am saying that if it is twenty points, then moving twelve point five four million dollars of FOB out of China origin saves two million five hundred and eight thousand dollars of duty a year. That is the headline number in this whole exercise, and it is also the number nobody actually collects.

Nobody collects it because the goods did not get cheaper. Assume the alternative origin costs eight percent more on FOB, which is roughly what I have seen when a program moves to a factory that has not run it before and is quoting on smaller volume with new tooling. Eight percent on twelve point five four million is one million three thousand dollars. Net the two and the move is worth one million five hundred and five thousand dollars a year. Twenty percentage points of duty against eight points of FOB premium leaves twelve points, and twelve points of twelve point five four million is one and a half million dollars. That is the static case, and the static case is always the one that gets presented in the meeting. It is not the one that shows up in the accounts.

Here is what eats into it. Assume qualifying a new factory, cutting new tooling, running pilot runs and ramping to full rate adds fourteen days to order-to-shelf on the moved volume, and that the transition takes a year. Take the moved FOB of twelve point five four million and gross it up to landed at one point three five times, which is my assumption, giving sixteen million nine hundred and thirty thousand dollars of annual landed flow on that moved volume. Fourteen extra days of pipeline on that flow is sixteen million nine hundred and thirty thousand times fourteen over three hundred and sixty-five, which is six hundred and forty-nine thousand dollars of inventory that did not exist before. At a carrying cost of nine percent a year, that is fifty-eight thousand four hundred and fifty dollars for the transition year. Real money, and it never appears on a sourcing deck.

The second leak is the one that actually hurts, because it lands on the customer rather than on the finance team. Assume that during the transition year twelve percent of the moved SKUs miss their shelf date by more than ten days. Assume each of those SKUs loses four percent of its season sell-through as a result, which is what happens when a seasonal item turns up after the weather or the campaign has moved on. Take the moved FOB at a retail multiple of two point nine, my assumption, which puts retail value on the moved range at thirty-six million three hundred and seventy thousand dollars. Twelve percent of that range losing four percent of its sell-through is nought point four eight percent of retail, or one hundred and seventy-four thousand six hundred dollars of sales that do not happen. At a gross margin of thirty-eight percent, the gross profit I lose is sixty-six thousand three hundred dollars.

Put the three together. Two million five hundred and eight thousand of duty avoided, less one million three thousand of FOB premium, less fifty-eight thousand five hundred of carrying cost, less sixty-six thousand three hundred of lost gross profit. One million three hundred and eighty thousand dollars. That is what a two and a half million dollar duty saving turns into by the time it reaches me. I keep about fifty-five percent of it, and I want to be clear that the rest is not taken by anybody. It is spent on the transition, and whether it gets spent well or badly is entirely inside my own organisation.

Now run the same arithmetic on a thinner assumption, because I do not actually know that the differential is twenty points. Assume it is ten. Ten percent of twelve point five four million is one million two hundred and fifty-four thousand, less the same eight percent FOB premium of one million three thousand, leaves two hundred and fifty-one thousand dollars. Take off the fifty-eight thousand five hundred and the sixty-six thousand three hundred and I am left with one hundred and twenty-six thousand dollars. At a ten point differential I am doing a year of qualifying, tooling, pilot runs and shelf risk to earn one hundred and twenty-six thousand dollars on a ninety-three point six million dollar book.

That is not a sourcing decision any more, it is a hobby. The break-even is simple enough to hold in your head: the move pays only while the tariff differential in points stays above the FOB premium in points. Twenty against eight is comfortable. Ten against eight is not. Eight against eight is a rounding error with a year of work attached to it.

It is worth going back to the sixty-one percent from 2020 as well, because the distance between then and now is the part a one-year comparison hides. On my twelve hundred box book, sixty-one percent would have been seven hundred and thirty-two boxes carrying fifty-seven million ninety-six thousand dollars of FOB. Today it is four hundred and eighty boxes carrying thirty-seven million four hundred and forty thousand. Two hundred and fifty-two boxes and nineteen million six hundred and fifty-six thousand dollars of annual FOB has walked out of China origin in six years.

Walked is the right word, because it did not leave all at once and it did not leave on a plan. It left a category at a time, usually after somebody in a sourcing meeting did the same arithmetic I just did and found it positive on the duty line alone. What that leaves behind is a supplier base that is broader than it was, shallower in each place than it was, and more expensive to administer than it was. Nobody puts the administration cost on the slide either.

And there is a piece of the moved volume that people consistently underweight. When you shift a program to a factory that has never run it, you are not buying a price, you are buying a capacity commitment from an organisation that does not yet know what your peak looks like. Assume the new origin can commit to ninety percent of your peak season requirement in year one. That is my assumption and it is a generous one. The missing ten percent is precisely the part that goes absent in your busiest month, and it is precisely the part your shelf dates depend on. So the eight percent FOB premium is not the whole cost of the new factory. The premium plus the shortfall is the cost, and the shortfall is measured in lost sell-through, not in dollars per unit.

I want to be careful with the seven percent, because it is the number that will get quoted most and it is the one I can use least. Nine point three million TEUs is a forecast about a gateway, not about me. If the gateway handles seven percent less and I hold my volume, my share of it goes up, and that is not a problem for me at all. What the figure does tell me is something about the port's own posture. It approved a three point four billion dollar budget and pushed capital spending to a ten-year high in the same breath as it told everybody volume is falling.

Three point four billion against nine point three million boxes is three hundred and sixty-five dollars of budget per box, and the seventy-four million dollar rail expansion at Berths 302 to 305 works out to under eight dollars a box if you spread it across a single year of throughput. Those are my sums on published figures, not a fee schedule, and I am not suggesting they get billed to me per container. What they tell me is the scale of the bet. This is a ten-year asset decision made in a down-volume year.

Which brings me to Pier 500 and the rail work, and to the only place where a port's capital plan touches my margin at all. The port is building the first new container terminal in decades and putting seventy-four million into rail at Berths 302 to 305. If my distribution centre sits on the rail side, that money eventually shows up as a better inland option and I should be positioning for it now rather than when it opens. If my network is truck-based out of the basin, none of this reaches me and I should stop reading it.

The mistake buyers make with port capital plans is treating them as a signal about next quarter's volumes. They are not. A terminal that takes years to build is a statement about where the port wants to be in the next decade, and the honest response from my side of the table is to ask what it does to my inland leg and my delivery calendar, and to leave the rest of it alone.

On alternatives, let me take them in the order a buyer actually considers them. Changing origin is the one this release is really about, and the trap is that it gets sold internally on the duty number alone. The duty number is the gross figure. The net figure is the duty differential less the FOB premium, less carrying cost on the longer pipeline, less the sell-through you lose while the new factory learns the program. If you cannot put a number on the third and fourth of those, you do not have a business case, you have a slide. The threshold I would hold people to is this: if the net after all four is under six percentage points of the FOB being moved, do not move it. On my book that is six percent of twelve point five four million, or seven hundred and fifty-two thousand dollars, and it is a test most of the proposals I see would fail.

Changing gateway is the second alternative, and I am going to measure it in a way that will look odd to anyone in transport. I do not measure a gateway change in freight cost. I measure it in total days to my distribution centre, because days are the unit that converts into safety stock, into working capital and into shelf dates. A gateway that costs less and takes four more days is not cheaper, it is four days of inventory I now own and finance at my own carrying cost.

The third alternative is changing inventory rather than anything else, which means pulling volume forward ahead of a policy change you believe is coming. Assume you pull ninety days of cover forward. Ninety over three hundred and sixty-five times nine percent carrying cost is two point two two percent of the goods value. Against a twenty point tariff differential that is cheap by an order of magnitude, and against a ten point differential it is still cheap. What makes it dangerous is not the carrying cost, it is being wrong.

If the change does not come, you have spent two point two two percent to own goods you would have bought later anyway, and you are carrying them into a season you have not sold through. I am not going to tell you whether the policy lands. I am going to tell you that two point two two percent is the price of being early and unsold inventory is the price of being wrong, and that those are two different budgets.

The traps are mostly boring and they are mostly paperwork. Assume nothing about origin: moving final assembly to a third country does not by itself change origin, and that is a question for the compliance function and the customs authority, not for a sourcing meeting with a margin target attached to it. Watch the FOB premium over time, because the eight percent I assumed is a first-year number and it is the one most likely to be quietly wrong by month nine, when the new factory has found out what the program actually costs to run. And watch the length of the transition, because the fourteen days I assumed is a plan, and plans of that kind have a habit of becoming twenty-one. At twenty-one days the carrying cost goes from fifty-eight thousand five hundred to eighty-seven thousand seven hundred, and the case gets thinner by twenty-nine thousand dollars without anybody making a decision.

There is also the half of the book that does not move, and it deserves more attention than it gets. Four hundred and eighty boxes and thirty-seven million four hundred and forty thousand dollars of FOB stays China origin. That is still the largest single block in my book, and the temptation is to treat it as the residue after the interesting work is done. It is not. It is the part where I have six years of cost history, known quality, and a factory that already holds my tooling.

For those categories the ask is not a lower price. It is a shorter committed lead time and a smaller minimum order quantity, because those two are what let me hold shelf dates on a smaller base without holding more stock. A supplier who will take a fifteen percent smaller order and ship it nine days sooner is worth more to me than one who takes two percent off the unit price, and I have yet to see a sourcing scorecard that scores it that way.

One thing I am deliberately not covering is what happens in the days after a box clears, when it is the inland leg and the delivery appointment that decide whether the shelf date actually holds. Nadia Pryce lives in that part of the week and reads it better than I do. My job stops at the point where I decide what to buy, where from, and how much of the delay I have already paid for in the price.

If I were signing this as a buyer, I would not sign a blanket shift away from China. I would sign a staged one with a gate in it. Stage one moves only the categories where the differential is widest and the program is least engineering-dependent, capped at a defined share of China-origin FOB. Stage two does not start until stage one has held its shelf dates for a full season. Every stage carries its own transition inventory budget and its own sell-through floor, and a stage that misses the floor stops. The reason is the arithmetic above: the duty saving is real and it is large, and it is also the easiest number in the business to lose in the eighteen months between signing the order and seeing the goods on a shelf. My margin does not care which factory made the unit. It cares whether the unit was there on the day the customer came in for it.

That is what forty percent means from where I sit. Not a trade story and not a port story. It is a number that says my supplier map changed faster than my cost model did, and the gap between those two is where the money goes. I can close that gap this quarter or I can discover it in the margin line next year. The port has already published its numbers. Ours are the ones still missing.

↑ Back to the key points

— By Grace Yang