SCFI Jumps 4.7% as U.S. Coast Rates Surge Even Where Europe Softens

The Analysis

Shanghai to Los Angeles, Ningbo to Savannah, Yantian to Rotterdam. Three lines I have drawn on a map more times than I can count, and this week they are not moving in the same direction. The Shanghai Containerized Freight Index closed at 3,205.97 points on 31 July, up 4.7% on the week. The composite is the number that gets repeated in headlines. The number that changes what I do on Monday is the split underneath it. U.S. West Coast went up 12.5% to 6,229 dollars per FEU. U.S. East Coast went up 12.6% to 9,054 dollars per FEU. North Europe went the other way, down 3.7% to 3,039 dollars per TEU. I have run cargo on all three of these lanes. When two of them rise together by almost the same percentage and the third falls inside the same seven days, that is not noise in the index. That is the market telling you the lanes have stopped keeping the same time.

I read a composite the way I read a national weather forecast. It is accurate and it does not help me plan a drive. What I plan against is the lane, then the gateway inside the lane, then the inland road out of that gateway. This week the useful information is the shape of the divergence rather than the level. Both U.S. coasts moved, 12.5% and 12.6%, which tells me the pressure is not sitting at one terminal or inside one chassis pool. If Oakland were having a bad fortnight, or if Savannah yards were full, you would see one coast move and the other hold still. Two coasts moving together by nearly identical amounts points at how much space is being offered across the whole trade. You do not fix that with a phone call to a terminal. You fix it by deciding again where your boxes should land.

There is a second thing in these readings that I want to put on the table before anything else, because it is the expensive one. The transpacific numbers in this index are booking-level. Drewry's spot assessments over the same days read softer. Two thermometers in the same room giving different answers. One measures what you are quoted when you ask for space. The other measures what cargo actually changes hands at. When they separate by this much, the gap is not a methodological quarrel you can wave away. It is the distance between the number you sign and the number the market settles at, and in a contract season that distance is counted in dollars per container.

Start with the 2,825 dollar difference between the two U.S. coasts. To me that is not a rate statistic, it is a toll booth. Pay it and your box stays on the water all the way to Savannah, Charleston, Norfolk or New York. Do not pay it and your box gets off at Los Angeles, Long Beach or Oakland and finishes the trip on a train. Every other part of your routing decision hangs on which side of that toll booth your inland cost sits.

The comparison most people run is the wrong one, because they run it at the water's edge. You do not consume a container at the water's edge. You consume it at a distribution centre in Chicago, or Dallas, or Columbus, or Allentown. So the arithmetic goes like this. Total cost through the West Coast is 6,229 plus whatever your inland leg costs from that gateway to the door. Total cost through the East Coast is 9,054 plus the inland leg from that gateway to the same door. The West Coast wins whenever your West Coast inland leg exceeds your East Coast inland leg by less than 2,825 dollars per FEU. That is the entire test. Put your own drayage and rail numbers into it, because mine are not yours, and the 2,825 is the only figure in the sentence that comes from the index.

Then put time on the same page, because a cheaper road that takes two extra weeks is not cheaper. Suppose, only to make the arithmetic visible and using your own averages from last quarter rather than any number of mine, that Shanghai to Los Angeles is 14 days on the water and 5 days on rail to a Chicago door, so 19 days. Suppose Shanghai to Savannah all-water is 30 days and 2 more by road to that same door, so 32 days. Thirteen days apart. Thirteen days of inventory standing still has a price. If one FEU of your cargo carries 60,000 dollars of landed value and your cost of capital is 8% a year, then 60,000 x 8% x 13 / 365 comes to about 171 dollars. At 200,000 dollars of value in the box the same thirteen days costs about 570 dollars.

Now the two halves fit together. The West Coast route wins whenever your West Coast inland leg exceeds the East Coast one by less than 2,825 plus that carrying cost. With 60,000 dollar boxes at 8%, your break-even inland gap is roughly 2,996 dollars per FEU. If your West Coast inland leg runs about 3,200 dollars above the East Coast one, the East Coast door is cheaper by roughly 200 dollars and it costs you thirteen days to collect it. If the gap is 2,000 dollars, the West Coast wins by about 1,000 dollars once the carrying cost is counted, and it gets there close to two weeks sooner. In the end it is a road problem, not a rate problem. The rate only decides where the toll booth stands. Your own inland network decides which side of it you live on.

Here is the uncomfortable part of this week's shape. Because both coasts moved by almost the same percentage, sliding volume from Los Angeles to Savannah buys you very little. You are walking from one toll booth to another that is charging the same toll on a bigger base. The diversification that actually helps in a week like this one is inside the coast, not across it. Oakland instead of San Pedro. Prince Rupert or Vancouver instead of a Puget Sound call. On the Mexican side, Manzanillo or Lazaro Cardenas with a rail crossing north, if the rest of your inland map makes that sane.

Do not underestimate those secondary gateways, and do not romanticise them either. I have routed through Prince Rupert when the San Pedro basin was backed up and it took a week out of my transit. I have also watched boxes sit there because the rail ramp ran out of slots and the trucking pool in that town is a small fraction of what Southern California has. A second gateway is a spare tire. It is worth exactly as much as your ability to actually fit it, and fitting it means knowing the ramp capacity, the chassis supply, the cut-off times and the appointment system before the day you need it rather than on the morning of it.

Price the switch honestly while you are at it. A gateway change is not a form. It touches your drayage contract, your chassis arrangement, the rail ramp where you have rates filed, the receiving hours at your DC, and the habits of the people who book your appointments every week. I allow two to three weeks of untidy service on the first handful of boxes through a new gateway. That is the reason you test it on cargo that can swallow two bad weeks, in a month when nothing else in the business is on fire.

Now the other half of the divergence. North Europe at 3,039 dollars per TEU, down 3.7% in the same week the American trades were up by double digits, says the Asia-Europe lane is not fighting for the same space. I am careful about how much I read into a single week, but the direction is clear enough to act on. If you serve Europe and you have any flexibility about where your boxes enter the continent, this is a reasonable week to use it.

European entry is the same kind of map problem as the American one, with more moving parts. A North European base port puts you onto barge and rail networks that have been built out over a century and that reach deep into Germany, the Benelux and Poland. A Mediterranean gateway puts you closer to the industrial north of Italy and the south of France, and it moves your customs point with it. Swapping one for the other is not a rate decision. It is a decision about which inland network you intend to live on and which office will be looking at your entries.

One thing I will not do from a desk is tell you which of the two ways into Europe is running this week. That depends on the waterway on the day you book, and guessing at it is how people lose a month and then blame the carrier. Look at the rotation your carrier is actually publishing for the sailing you intend to book, and make the comparison in door-to-door days rather than on the port pair.

The practical consequence for your negotiations is that you should refuse bundles. If a carrier wants to talk about a multi-trade agreement this month, take it apart lane by lane. One trade is reading hot at the booking desk and another is handing space back. A blended price hides precisely the thing you are trying to use, which is that you can afford to be patient on one and you have to be quick on the other.

Back to the two thermometers, because this is where the contract decision actually lives. If you sign a twelve-month transpacific agreement in a week when the booking-level readings are up 12.5% on the West Coast and 12.6% on the East Coast while the spot assessments read softer, you are signing at the top of the quote cycle. I have done it. By October I was holding a rate card the market had walked away from, and nothing in the contract let me walk with it. The lesson I took from that was not that contracts are a bad idea. It was that the week you sign matters more than the year you sign for.

The item puts the divergence down to the post-front-loading correction landing unevenly across lanes and contract types, and that is consistent with what I see on the ground. Cargo was pulled forward and now the pull-forward is unwinding, but not at the same speed everywhere. If you front-loaded hard, your third-quarter pipeline is full and your marginal container is not urgent. You can take the slower road, or the secondary gateway, and bank the difference. If you did not front-load, you are buying into a booking window that is tight right now, and your problem is not the price. Your problem is getting on the ship.

That is also why the spot-versus-contract question has two different answers this week depending on which side of the front-loading you were on. Spot cargo in a tight window gets rolled. Not in theory, in practice: it gets pushed to the next sailing, and if that next sailing is the one that gets voided, it gets pushed again. Contract cargo with a volume commitment gets protected ahead of it. In a week like this one the value of a contract is not the rate it contains. The value is that somebody owes you a slot.

Which does not mean sign everything up. It means buy the slot and limit the exposure. In a market where the booking index and the spot assessment disagree, the sensible shape is a shorter commitment and a smaller committed share. Ask your carrier to price twelve months, six months and a quarter on the same volume. The premium on the shorter one is the price of not being wrong for a year. Sometimes that is worth paying and sometimes it is not, and you will not know until all three numbers are on paper, which they will produce if you ask in writing and mean it.

The trap inside that conversation is the minimum quantity commitment. A shorter contract usually carries a higher rate and costs you the volume tier you would have earned across a full year. Ask what your committed volume buys in each structure and treat the difference as the price of flexibility. Assume the six-month version comes in 150 dollars a box above the twelve-month one and you run 400 boxes: that is 60,000 dollars for the option to be wrong for two quarters instead of four. Against a market that may be 10% lower by November, that can be cheap.

I run switching rules on numbers rather than on instinct, and I write them down before the market moves so I am not negotiating with myself later. My line on the West Coast reading is simple. Two consecutive weeks under 5,600 dollars per FEU, roughly 10% below this week's 6,229, and I start moving volume back toward spot. Two consecutive weeks above 6,850, roughly 10% above, and I lock more of the fourth quarter into contract. In between the two lines I change nothing structural and spend the time on routing instead. Your percentages will not be mine. What matters is that they exist, that they are written down, and that somebody else in the company can read them off a screen and act without calling you.

The other rule I keep is on the spread between the coasts. Today it is 2,825 dollars. If it widens past roughly 3,500, all-water to the East Coast stops making sense for most of the inland destinations I work with, and I go back to West Coast plus rail. If it narrows below about 2,000, the answer flips. A rule about a spread is much easier to hold than a rule about a level, because spreads move slower than the absolute numbers do and they are far less sensitive to which index you happened to read that morning.

Free time is where a gateway switch usually goes wrong. Move to a new terminal and you move onto somebody else's clock. Free days at the terminal, free days on the box, and the event that starts the clock are three separate things, and they differ by terminal and by carrier. The failure is common enough to be worth naming: the water leg gets cheaper and the invoice gets bigger, because nobody re-read the free time at the new gateway and the boxes sat through a long weekend. Get the number of free days, the clock start event and the per-day rate in writing before the first container moves.

Billing bases deserve the same suspicion. Any charge quoted as a percentage is meaningless until you know what it is a percentage of. Ask whether it applies to the base ocean leg alone or to the base leg plus named accessorials, because the difference across a full invoice is not small. Ask the same question about the charges sitting next to it. The purpose is not to fight every line. It is to know which lines scale and which do not, because the ones that scale are the ones that will surprise you.

Transit promises deserve a harder question than most people ask. A quote that says eighteen days usually means something narrower than the sentence suggests, and in a hot market the window quietly widens while the number on the quote stays where it was. Ask for two numbers separately: port to port, and door to door including the inland leg and the appointment. Then ask what happens to the commitment when the sailing is voided. If the answer is that the transit commitment resets, you do not have a commitment. You have an estimate, and your inventory plan should be built on the estimate rather than on what you were sold.

If you are on spot this month, work out where you stand in the queue. In a tight window the order of protection runs roughly from contract volume with a commitment, to contract volume without one, to spot. If you are in the last group, book further ahead than feels necessary, and ask directly what happens to your booking when the sailing is cancelled. Ask who absorbs the extra dwell. Ask whether the booking survives onto the next sailing or goes back into the pool to be re-sold.

Changing where you buy is the option people reach for too quickly. A new origin can be genuinely cheaper, and the lane out of Haiphong or Nhava Sheva or a Mexican port may price better than the one you use today. But service density is part of the price. A port with one sailing a week gives you a completely different risk profile from a port with four, because one void sailing on a weekly service is a fourteen day slip and one void sailing on a four-times-weekly service is a three day slip. If you are moving origin, count sailings per week as carefully as you count dollars per box.

There is also a version of this problem that is not a shipping problem. If your network already holds inventory in more than one place, moving stock between your own facilities can be cheaper and faster than moving containers between ports, and it sits inside your control in a way no carrier rotation ever will. I have got through more than one tight-booking month by shipping from the DC that had the stock rather than from the port that had the space.

So the week itself. If I were running an import desk looking at these numbers on 2 August, I would do four things before the month closes, and none of them involves tearing up the network. The arithmetic comes first. By Friday 7 August, put a table in front of your team: every major inland destination you serve, with three columns, West Coast gateway total cost, East Coast gateway total cost, and the inland gap between them. Use 2,825 as the spread and your own rates for the rest. What comes out is a recommended gateway per destination plus the number at which that recommendation flips. Keep the flip number. It will still be true in six weeks when the recommendation is not.

The spare tire is the second piece. Before 21 August, get real cut-off times and real ramp availability for one alternative gateway on each coast you use, and book one live shipment through each. Not a rate enquiry, an actual booking, because a rate enquiry tells you nothing about whether the ramp has room. The target for the quarter is that no single gateway carries more than 60% of your U.S. volume, and that on at least one lane you have evidence rather than a promise that a second door opens.

The contract calendar is the third piece. Where transpacific agreements come up for renewal, get the twelve-month, six-month and quarterly price on the same volume, in writing, before you sit down to negotiate. Decide the committed share on paper, in advance, not in the room. My own starting position in a week shaped like this one is contract coverage around 60% of fourth-quarter volume with 40% left to spot, and then the trigger rules move it from there. That is a starting position for my business, not a prescription for yours.

Europe is the fourth piece, and it is the one people skip because it is not the lane that is shouting. A soft North Europe week is a good week to test one alternative entry, whether that is a different base port or a Mediterranean gateway with an inland leg you have never used. Whatever you test, measure it the same way: door-to-door days, not port pair. If the alternative cannot land within five days of the routing you run today, it is not an alternative for the customers you actually have.

None of this requires rebuilding anything, and a week like this one is a poor time to try. What it requires is knowing where the spare tire hangs and whether the bolts fit. The composite will be quoted for a day and then forgotten. The split underneath it, two coasts rising together, Europe easing, and a booking desk running hotter than the spot market, is what tells you which road you are on and what the toll is. Work out your own break-even number, write down the level at which you would switch, and make sure somebody else on your team can read both off a screen and act without asking you first. The road is open. It is simply not the road it was in June, and driving it as though it were will cost you more than the toll.

↑ Back to the key points

— By Ömer Kaya