Drewry World Container Index Falls for Third Straight Week to $4,255

The Analysis

$4,255 is the number pinned to my screen this week. That is where Drewry's composite World Container Index landed in the week to August 2, 2026, down 3% on the week, and it is the third consecutive weekly decline. Three prints, one direction. Two down weeks is noise. Three is a trend you build a plan around. Four is a trend your finance department comes looking for.

The composite hides the interesting part, so pull it apart by lane. Shanghai to Los Angeles slipped another 2% to $5,739 per FEU, and that lands on top of a 6% fall the week before. Shanghai to New York did nothing at all. Flat at $7,578 per FEU. Same origin, same week, two completely different behaviours, and the gap between them is worth more to your books than the headline.

Drewry puts the softening down to two things: demand weakening as new U.S. tariff measures take effect, and the May to July front-loading surge having drained the pool of cargo available to pull forward. One is policy, one is arithmetic, and both of them bite. The rate is the symptom. The condition underneath is that your forward booking book is thinner than it looks, and every carrier on the lane is staring at the same quiet September.

Let us run the numbers, starting with the boring one. A 3% move on $4,255 is roughly $128 per box. The same 3% on a $1,500 market would have been $45. Identical percentage, three times the money. Read the dollars, not just the decimals, or you will misjudge every cycle you ever trade through.

Back-solve the composite and last week's print comes out around $4,387, because $4,255 divided by 0.97 lands there. There is nothing clever in that; Drewry published both the percentage and the level, so the earlier level is implied. It matters because a single number in isolation has no direction in it, and direction is the only thing you can make a decision on.

The direction is much clearer on Los Angeles. If $5,739 is a 2% fall, the previous print was about $5,856. If that was itself a 6% fall, the print before it was roughly $6,230. Put those two together and Los Angeles has given back about $491 per FEU in fourteen days, which is about 7.9% off the box. Write all three weeks on one line in your notes. Most people record only this week's print, and that is exactly why they negotiate from memory instead of from evidence.

Scale it the way your own operation runs. Assume you tender 40 FEU a month into Los Angeles. Assume each box carries 1,200 units of a product that retails at $20. Then the freight content of one unit went from about $5.19 ($6,230 divided by 1,200) to about $4.78 ($5,739 divided by 1,200). Forty-one cents a unit. As a share of retail, freight dropped from roughly 26% to roughly 24%. On monthly volume of 40 boxes, the two-week slide handed you something close to $19,640 back, which annualises past $235,000 if it holds. That is the whole size of the prize, and it is worth knowing precisely, because people routinely argue about hundreds of dollars of accessorial charges while ignoring nineteen thousand dollars of base rate sitting right there.

Now the number that deserves a second look. Shanghai to New York held flat at $7,578 while the composite fell 3% and Los Angeles fell 2%. Flat is not a rounding error in a week like that. Flat is a statement. The New York to Los Angeles spread is now $1,839 per FEU, which means you are paying about 32% more box cost to land on the East Coast before a single inland dollar is counted. When did you last rebuild that comparison honestly?

Here is how it plays out with a number you can replace with your own. Assume that reaching the same distribution centre costs you $1,400 more inland from Los Angeles than from New York. Then today Los Angeles totals $7,139 ($5,739 plus $1,400) against New York at $7,578, so Los Angeles is $439 per FEU cheaper. Two weeks ago the same arithmetic gave Los Angeles $7,630 against New York's $7,578, so New York was cheaper by about $52. The whole decision flipped by roughly $491 a box in fourteen days. If your allocation model has not been re-run since June, it is running on stale input and it is quietly giving you the wrong answer every week.

Finish the arithmetic before you touch a routing plan. Assume the Los Angeles routing costs you twelve more days door to door. Assume the cargo in one box is worth $60,000 and your carrying cost is 20% a year, which is $32.88 per day per FEU, so twelve days is about $394.50. Against a $439 advantage, you net about $44 per FEU, call it $1,760 a month on 40 boxes. Real money, not decision-changing money. This is why I do not reroute anyone on a headline spread. You reroute on a spread that survives the carrying-cost line, and this one barely limps across it on my assumptions.

Notice what the deciding variable actually was. It was not the ocean rate. It was my assumed inland differential of $1,400. Run it again at $1,050 instead and Los Angeles totals $6,789 against New York's $7,578, a $789 advantage, which nets about $395 per FEU after the same twelve days of carrying cost and suddenly looks like a genuine decision. The lever you need this week is not a better freight quote. It is real inland numbers from your trucker and your rail provider, in writing, for the lanes you actually move. Get those two figures and the whole question answers itself in about ten minutes.

The reason behind the slide tells you how long to expect it to last. New U.S. tariff measures taking effect is a demand story, and demand stories can run for months. The exhaustion of front-loading is a stock-flow story, and those are always shorter than people think. There was a finite amount of cargo you could pull forward, you pulled it forward between May and July, and that inventory now has to be sold through before normal ordering resumes. You cannot front-load the same shipment twice. So the rate softness has one leg that persists and one leg that unwinds, which is another way of saying this is a correction inside a high base rather than a return to cheap freight.

That second leg has a second consequence that rarely gets mentioned in rate reports. Assume you pulled four weeks of third-quarter volume into June and July so you could beat the tariff measures. Assume your annual contract carries a minimum quantity commitment of 600 FEU a quarter, and your realistic tender this quarter is 420. You are 30% short by your own choice, and your contract does not care whose choice it was. Depending on how the language reads, you are either absorbing a shortfall exposure or quietly losing the tier pricing that made the contract worth signing in the first place. Check the clause this week rather than discovering it in October when somebody re-quotes you at a different level.

If you sell rather than buy ocean freight, this week looks different but not unfamiliar. Assume you quoted a U.S. buyer CFR when the Los Angeles box was around $6,230 and you are now buying it at $5,739. That is $491 per FEU of unearned margin. You can keep all of it and leave your buyer to notice on their own, or you can hand back $150 per FEU to protect next year's volume and still bank $341. The first version wins the quarter, the second version wins the account. Buyers benchmark freight faster than they benchmark anything else, because it is the one line printed on somebody else's index every week.

The same visibility cuts the other way too, which is why I treat the index as a compass rather than an invoice. The World Container Index is Drewry's own assessment of a composite 40-foot box across its lane set. Your quote is your quote, with your own billing basis, your own surcharge stack and your own free-time terms. Use the composite for direction and your own all-in quotations for money. Comparing your carrier's base rate against the composite and declaring victory is how people end up congratulating themselves and then paying more than last month.

So here is what that looks like as an actual conversation rather than a feeling. Your forwarder quotes $6,100 all-in to Los Angeles. The lane just printed at $5,739, and the print two weeks ago was about $6,230. You are not asking for $5,739; that is Drewry's composite-related lane level, not your buy rate. You ask for $5,984, which is the midpoint back to the pre-slide level, plus 21 days of destination free time, plus written confirmation that the quote is all-in rather than subject to named surcharges. If it comes back at $6,050 with ten days of free time, then the real gap is not the sixty-six dollars. It is the eleven days of demurrage exposure you just accepted at somebody else's discretion.

Budgeting needs the band, not the point estimate. Take this week's $4,255 and run three honest cases. If the slide continues for three more weeks at 3% a print, the composite lands near $3,883. If it stops here, it stays at $4,255. If it rebounds 6% because capacity gets pulled faster than demand falls, it prints around $4,511. Give finance the band from $3,883 to $4,511 and a plan that survives the top of it, because plans built only on the bottom always get rewritten at the worst possible moment. The width of that band, by the way, is about 16% of the index, which is the honest measure of how much you actually know right now about freight in November.

Then come the rules, and mine are deliberately mechanical, because mechanical rules do not panic on Tuesdays. Until further notice, no fixed-rate commitment longer than six months while the index moves more than 1.5% a week in one direction. On this composite that is about $64; on the Los Angeles lane about $86. The moment two consecutive prints land inside that band, whatever the direction, that is the plateau signal, and that is the moment a twelve-month fixed rate becomes a sensible conversation instead of a gamble.

Beyond that, write your lock lines down now, while nobody is frightened and nobody is euphoric. Mine for the balance of this year, and they are mine rather than a market forecast: Shanghai to Los Angeles prints at or below $5,500 per FEU in any week, put 60% of fourth-quarter volume onto a six-month fixed. It prints at or above $6,500, run 80% spot and stop chasing fixed allocation you will not need. Between the two lines, hold your current mix and go do something more useful with your afternoon. Write it in your team channel so nobody has to improvise in the middle of a booking week.

Inventory deserves its own warning. Do not add weeks of cover because the box got cheaper. Assume you hold nine weeks today. At 20% annual carrying cost on a $60,000 box, one extra week costs about $231 per FEU, and two extra weeks cost about $462. Two weeks of additional cover wipes out essentially the entire $491 per FEU you just won on the freight line, and leaves you holding inventory bought for a demand assumption nobody has tested. Add cover if your demand signal genuinely says so. Never add it because the freight got cheaper; those are two different decisions that only look alike from a distance.

Booking hygiene is unglamorous and it is where most of the leakage happens in soft markets. Before a quote is accepted this month, three things go into writing: free time at destination stated in days, the per-diem schedule once free time expires, and whether the quote is all-in or subject to named surcharges and how each of those is billed per container. Ask the effective date of any peak season charge and get the answer in an email, because "we will sort it out later" is a line item with your name on it. Any verbal answer means the quote is incomplete, and an incomplete quote is not a price.

Then the alternatives and their real costs. Changing ports, on my assumptions above, nets $44 a FEU, which is not worth the operational churn unless your true inland differential is closer to $1,050, in which case it is worth a proper look. Changing modes into the air market during a soft ocean market only makes sense if your stockout cost per day is extreme, so price your own stockout cost before you even open that conversation. Changing sourcing country is not a freight decision at all; it is qualification, tooling, sampling and lead time, and it should be counted in weeks and months rather than in dollars per box. Changing your inventory policy is the one lever people pull hardest and measure least, so it is the one I would least trust myself to pull on a rate headline.

One habit worth building while the market actually moves: keep your own weekly sheet rather than somebody else's newsletter. Four columns and nothing else. The composite print. Your lane print. The best all-in quote you were offered on that lane in the same week. The free time attached to it. Then manage off the four-week average of each column instead of reacting to single prints, and decide your review rule in advance. Mine is blunt: revisit the tender calendar only when that four-week average has moved more than 5% away from where it sat when you last renegotiated. Ten minutes a week, and it stops the whole team re-litigating the same decision every time a new number appears in somebody's inbox.

Then think about what the carrier does next, because their next move becomes your next cost line. Three down prints is exactly the point at which carriers stop cutting and start protecting, and protecting usually means withdrawing capacity rather than publishing lower numbers. I am not going to claim any August schedule changes, because I cannot see what anyone has actually filed. What I can do is ask two cheap questions in writing: send me your published August and September sailing plan for this lane, and tell me which weeks are firm today. Those two answers together tell you whether the rate on your screen is even bookable next week, which is worth more than the fourth decimal place of the quote.

The buying side of your own supply chain needs the same discipline. Assume your supplier priced you at June levels and you have now identified about $491 per FEU of movement, roughly forty-one cents a unit at 1,200 units to the box. Asking for the full forty-one cents is a losing conversation and a waste of leverage you may need later. Asking instead for a freight adjustment clause that revisits quarterly, naming this index as the reference and setting the trigger at a 5% move off the agreed level, costs the supplier nothing today and hands you a mechanism instead of a fight. Put it into the next purchase order rather than trying to reopen the last one.

Watch what soft rates quietly do to habits nobody remembers agreeing to. Cheaper boxes tempt people to consolidate less aggressively, to take slower routings for a few dollars less, and to book later in the cycle because space has been easy. All three work perfectly until the week they stop working. Keep your consolidation targets where they are, hold your transit requirements, and do not let a 7.9% rate move talk you into a booking process that only functions while nobody else wants the space either.

Timing the next tender round deserves its own pasted rule. Reopen the annual tender when either of two things happens: the composite sits inside a 1.5% band for two consecutive prints, or it has fallen more than 10% cumulatively from the level that was live when you signed the current deal. Ten percent off $4,255 is about $426, which puts that second trigger near $3,830, and it lands suspiciously close to the three-more-declines scenario at roughly $3,883. When two rules that have nothing to do with each other point at nearly the same number, that number deserves your attention.

Send finance one page this week before they come looking for it. Three derived weekly prints for your main lane, the monthly swing on your own volume, and the planning band from $3,883 to $4,511. Add the two lock lines and ask for sign-off on them while nobody is frightened. The alternative is explaining a four-week-old decision to somebody holding a print they found before you did.

One test I run every soft cycle tells you more about your own buying than about the market. Line up your actual buy rate against the printed lane level for the last eight weeks and look at the gap rather than the rate. Assume your August buy is sitting around $6,100 to Los Angeles while the lane printed about $5,739. That is roughly $361 per FEU of daylight, which is about $14,440 a month on 40 boxes, and it is not the carrier taking it from you. It is lag. Nobody in your own process repriced anything, so you are buying at a price that stopped existing two weeks ago. Measure it once, then set a rule: any week the gap between your buy and the print exceeds 3%, somebody has to explain it in writing. Three percent on this lane is about $172 a box.

The East Coast deserves its own small worry. New York held flat while everything else softened, and the flat ones are usually the ones where space quietly goes tight first, because nothing in the pricing told anybody to move cargo away. Test it cheaply this week. Ask your East Coast carrier for a written space commitment for August, stated as slots per sailing, and ask what happens to your allocation if you tender above it. If the answer comes back as a general assurance rather than a number, you have learned something, and it costs nothing to learn it in July instead of September.

Three prints down and every inbox I have got is planning a party. My own rule is simple and it has saved me money more than once: sentiment changes on the third print, budgets change on the sixth, and print number four usually has a very firm opinion about print number three.

↑ Back to the key points

— By Vivian Zhao