Drewry World Container Index jumps 9% to $4,530 as peak-season rates hit 2024 high

The Analysis

Nine percent is the headline, and the headline is the least useful number in the file. The Drewry World Container Index closed the week to 2 July at 4,530 dollars per 40ft container, up 9% week on week, up 61% year on year, and at its highest level since September 2024. Underneath it sit four lanes that I actually care about. Shanghai to Los Angeles, 6,349 dollars, up 10%. Shanghai to New York, 7,902 dollars, up 11%. Shanghai to Rotterdam, 4,682 dollars, up 7%. Shanghai to Genoa, 6,360 dollars, up 10%. And Drewry counted eight blank sailings on the transpacific for the following week, against one on Asia-Europe.

This number has a problem. Not that it is wrong, but that it is a weighted blend across trades that hardly any single shipper sails on in those proportions. Say 9% in a budget meeting and you under-forecast the East Coast by two points and over-forecast North Europe by two. Both errors cost money in opposite directions: one leaves you short of space you should have bought, the other has you panic-booking space you did not need. I stopped quoting the composite to my own team a long time ago. I quote the lanes.

The 61% year-on-year figure does more work than the 9%, so let us do the arithmetic on it before anything else. Divide 4,530 by 1.61 and you get roughly 2,814, which is where the same index sat a year ago. If you signed an annual contract in the summer of 2025 anywhere near that level, the spot market is now about 61% above your contract and the renewal conversation is going to be unpleasant. If you are fully on spot, you have already paid it and there is nothing left to discuss. Which of those two you are decides what this week's print means to you, and it is the first thing to establish before anybody touches a booking screen.

Before any of that, I run the week backwards, because week-on-week percentages are worth more than people think. Divide each lane by one plus its change and you have last week: Los Angeles 5,772, New York 7,119, Rotterdam 4,376, Genoa 5,782, composite 4,156. The week therefore cost 577 dollars a box into Los Angeles, 783 into New York, 306 into Rotterdam and 578 into Genoa. Four numbers, no forecast, no broker email, no view about what happens next. Just this week's print divided by one plus the published percentage, which is arithmetic rather than opinion, and which gives me something I can actually put into a model.

One index is a methodology wearing a number, and I want a second one before I move money. The Freightos Baltic Index readings for the week to 27 June had Asia-US West Coast and Asia-US East Coast both up 8%, at roughly 6,200 dollars per FEU on the West Coast and roughly 8,000 dollars per FEU on the East Coast, with West Coast rates up about 120% since mid-May and East Coast rates up 85% over six weeks. Run those backwards too: 6,200 divided by 2.20 is roughly 2,820 for the West Coast in mid-May, and 8,000 divided by 1.85 is roughly 4,325 for the East Coast six weeks earlier. Those two derived numbers are the most useful things in this file, because they describe the slope of the market before anybody started writing about a peak season.

Freightos also put the East Coast roughly 1,000 dollars a FEU above the front-loading high of the summer of 2025, the West Coast just above its 2025 peak, and Asia-North Europe and Asia-Mediterranean roughly 1,300 and 3,000 dollars above their respective 2025 peak-season highs. Now read the caveat that came attached, because the publication said explicitly that Drewry and Freightos use different lane definitions and methodologies and that their rates and weekly changes should not be compared directly. I am not going to subtract 1,300 from Drewry's Rotterdam print and announce where the 2025 peak was. What I will say is that two datasets built differently are pointing the same direction on all four lanes, and that agreement is worth more than either one alone.

The number I trust least in this file is one I produced myself, so let me say so. My 2,814 comes from a percentage that was published rounded, and a rounded percentage moves the base. If the true year-on-year change was 60.5%, the base was 2,822. If it was 61.5%, the base was 2,806. Sixteen dollars of spread on a number I am about to put into a budget discussion. That is fine for a renewal conversation and not fine for a contract escalation clause, where you want the published index value rather than my long division. The same rule applies to the 5,772 and the 7,119 I derived above: good enough to size a decision, not good enough to invoice against.

Here is the part that makes this expensive rather than merely irritating. Drewry counted eight blank sailings on the transpacific for the week after 2 July and one on Asia-Europe. Eight voids take a real slice out of scheduled capacity, which is why the transpacific printed 10% and 11% while Asia-Europe printed 7% and 10% with its fleet still largely running. At the same time, Drewry's five-week outlook has carriers restoring capacity and blank sailings easing. Both sentences come from the same publication and neither reconciles the other. Near term, less space. Five weeks out, more. That combination is precisely the market in which you should not sign a six-month fixed rate at this week's print, and it is also the market in which you cannot assume the space will still be there if you wait.

What is holding the near term up is not only the voids. Freightos pointed at congestion at major hubs in South Asia, the Far East and Europe, which keeps vessels and boxes tied up longer and reduces the capacity that actually turns up on schedule, plus demand pulled forward into May and June ahead of July BAF increases, manufacturer price hikes and the US tariff deadline. Middle East disruption stayed in the background as an operational risk rather than a rate driver, and Freightos identified surging peak demand rather than oil prices as the main driver of the container rate move.

Pulling demand forward has a consequence that the same publication flagged, and it is the most under-priced sentence in the whole file: an early start can also produce an early unwind, possibly sometime in July. Freightos framed that as a possibility rather than a forecast, and I will be equally careful. If you are deciding whether to commit all of your July volume at 6,349 into Los Angeles, that sentence is the argument for not committing all of it.

Money, then. Every assumption is labelled, and you should change every one of them to your own. Assume you are a US importer moving 60 forty-foot boxes a month, 30 into Los Angeles and 30 into New York. Assume you are buying spot rather than sailing on an annual contract. Assume you could have booked at last week's readings and did not, because the quote in your inbox still said valid until the end of the month.

Last week, 30 boxes into Los Angeles at 5,772 is 173,160. This week at 6,349 it is 190,470, a difference of 17,310. Last week, 30 boxes into New York at 7,119 is 213,570. This week at 7,902 it is 237,060, a difference of 23,490. Together that is 40,800 dollars a month on a 60-box book, for the single act of letting one week go by. Per box that is 680 dollars, a little over 10% against a blended 6,445 average, and 489,600 annualised if both the level and the volume hold. If your freight budget had any slack left in it for 2026, this is probably where it went.

The other branch, because I do not get to pick only the scary one. Assume the unwind arrives in the second half of July and takes 10% off both lanes. Los Angeles goes to 5,714 and New York to 7,112, so the same 60-box book costs 171,420 plus 213,360, which is 384,780, or 42,750 less than booking today. The two branches are close to symmetric: roughly 40,800 at risk if you wait and rates keep climbing, roughly 42,750 forgone if you lock and they fall. Two similar numbers with different probabilities is not a decision, it is a coin toss. The way you stop tossing coins is a threshold, and I set mine below.

If you are on contract rather than spot, the arithmetic looks different and it is worse in a slower way. Assume you have 40 boxes a month sailing on an annual contract signed around the level the index sat at a year ago, call it 2,814. Assume renewal lands on 1 October and you roll onto a spot-linked rate at 4,530. That is a step of 1,716 dollars a box, or 68,640 dollars a month on 40 boxes, or 823,680 across a full year. Nobody announces that number. It arrives as a renewal letter. The only defence is to start the conversation before the letter exists, which means this month, while you still have a quarter of negotiating runway and before the carrier has finished testing its July filings.

And if your 2026 budget was built on the assumption that rates would be flat year on year, here is the size of the hole. Assume 720 boxes a year, which is the 60 a month from the example above. At the year-ago level of 2,814 that is 2,026,080. At 4,530 it is 3,261,600. The difference is 1,235,520, and it did not arrive as one event. It arrived as 61% spread over twelve months, most of it in increments small enough that nobody escalated them. That is the actual argument for circulating a weekly freight number inside the finance team rather than leaving it inside the logistics team.

Space risk is a separate calculation and most people fold it into the rate calculation, which is how they get surprised twice. Assume your 60 boxes sit on four strings, roughly 15 boxes each, and one of those strings is among the eight voids. That is 15 boxes re-booked into a week in which everybody else with a voided string is also re-booking. Assume the re-booking costs you one week. Your inland appointment moves, your free-time clock starts later than your plan assumed, and the exposure turns up about a month from now as detention and demurrage, the line nobody budgets and everybody pays. Then assume congestion at the Far East and European hubs adds four days to container availability on those 15 boxes, which is 60 box-days of exposure. Assume your contracted detention rate is 150 dollars per box per day, and note that this one is my illustration rather than a published figure. Sixty box-days at 150 is 9,000 dollars, for one rolled string, in one month.

The other question I always get asked in a week like this is whether to pull fourth-quarter volume forward, and the honest answer is that the arithmetic usually says no. Assume you have 200 boxes of Q4 merchandise you would normally ship in September. Assume the goods are worth 40,000 dollars a box and your carrying cost is 20% a year, so two extra months of holding is 40,000 x 0.20 / 12 x 2, roughly 1,333 dollars a box, or 266,600 on 200 boxes. Pulling forward only pays if you expect September spot to be more than 1,333 dollars a box above today, which means above 5,863, a rise of 29% from 4,530.

Nothing in this file supports a 29% rise. The counter-argument is space rather than price: eight voids mean July space is genuinely tight, and if your goods have a fixed shelf date, the value of guaranteed space can exceed 1,333 a box. That is a merchandising decision, not a freight decision, and it belongs to whoever owns the shelf date rather than to whoever owns the rate.

Thresholds, because this is where the rule I repeat every summer earns its keep: once it crosses the line, lock it. Shanghai-New York at 7,902 sits 98 dollars under 8,000, while the Freightos East Coast reading is already around 8,000, so the two datasets disagree by roughly 100 dollars about where the East Coast actually is. That disagreement is not a reason to wait for a cleaner number. If Drewry prints above 8,000 on the East Coast, I commit the third-quarter allocation within 48 hours of the print, because 8,000 is roughly 1,000 above the 2025 summer high and round numbers are where escalation clauses, approval limits and everybody else's stop orders live. On the West Coast, 6,349 is already above the Freightos 6,200; my line there is 6,500. And if the East Coast prints below 7,500 twice in a row, I stop committing and let it ride, because that is what an unwind looks like in the data long before anybody is willing to call it one.

Two jobs fall out of that, and they belong to two different people, which is usually why both get half done. The pricing job is to re-price every quote in the July file issued before 2 July against the four lane prints, and to throw out any quote that presents the composite as if it were a lane rate. A quote that predates the July filings is not a quote, it is a historical document. The booking job is to establish in writing, for every booking reference, whether it carries protected space, confirmed equipment, or only a committed allocation, because those three behave completely differently when a string is voided. I want at least 80% of July volume on references with protected space or confirmed equipment before the month is out, and I want the exceptions listed by name rather than discovered on a Friday afternoon.

Alternatives, priced properly. Routing East Coast cargo through the West Coast is the first thing everybody suggests, so here is what it actually costs. The gap between 7,902 and 6,349 is 1,553 dollars a box. Assume your inland move from a West Coast port to an Eastern distribution centre is 1,200 dollars a box and adds 14 days of transit; both are my assumptions, check them against your own inland tariff. Assume the goods are worth 40,000 dollars a box and your carrying cost is 20% a year, so 14 extra days is 40,000 x 0.20 / 365 x 14, roughly 307 dollars. All-in West Coast routing is 1,200 plus 307, which is 1,507 against 1,553. That is a 46 dollar advantage wrapped in two weeks of extra transit risk, an extra handling and one more appointment to miss. It is not an advantage, it is a wash with more ways to go wrong. Do not switch on the 1,553 headline alone.

Asia-Europe is the other side of the ledger and it needs a different playbook. One blank sailing against eight means the space risk there is materially lower, and I would not pay transpacific-style premiums or accept transpacific-style peak season surcharges on a lane running a single void. But the lane is not cheap against its own history: Rotterdam up 7% to 4,682 and Genoa up 10% to 6,360, with the Mediterranean sitting roughly 3,000 dollars above its 2025 peak-season high, which makes it the most stretched lane in this file relative to where it has been. European hub congestion is in the Freightos read as well. Lower space risk, higher price risk, and a five-week outlook that says capacity is coming back: I would fix less volume there and hold more of it on spot than I would on the transpacific.

The traps, in the order I have personally fallen into them. Do not bill on one index and quote on the other; the publication warned about it and the warning exists because people do it every week. If you sail on an annual contract, the composite does not touch you until renewal, but BAF, PSS and GRI filings do, and July filings were being tested, so re-read the surcharge schedule rather than the base rate. A blank sailing is not only a rate event: it produces rolled cargo, revised ETAs, missed connections and downstream free-time exposure that becomes detention and demurrage a month later, long after whoever made the booking has moved on.

Cap the length of anything you fix at this week's print, because a market that is tight for one week and looser five weeks out is not a market for a twelve-month signature. And when you renegotiate, negotiate the surcharge schedule and the free time, not just the base rate; the base rate is the part you can see.

What I watch next, in this order. Whether the July increases hold, which is the clearest read on how much pricing power the carriers actually have left. Whether the eight transpacific voids turn into nine or fall back to four, because the five-week outlook says they should fall and if they do not, the capacity story has changed. Schedule reliability and vessel queues at the Asian and European hubs, because a softer rate print means much less if delayed rotations are still eating the capacity that arrives on time. And the East Coast against 8,000, which is the number that decides whether I lock or wait.

Sixty-one percent in twelve months, nine percent in one week, and the market treated the nine percent as the story. The useful thing about a weekly index is that it only lets the freight market upset you once every seven days. The dangerous thing about it is exactly the same, and it is how a 61% move arrives as twenty small surprises and one large one, by which point the renewal letter has already been typed.

↑ Back to the key points

— By Vivian Zhao