Drewry World Container Index Surges 12% to $3,969, Hitting 18-Month High as Transpacific and Asia-Europe Rates Spike

The Analysis

Nine days ago the composite index printed 3,433 dollars. For the week ending 18 June it printed 3,969, up 12 per cent, the highest level in eighteen months. Shanghai to New York went to 6,769, up 15 per cent. Los Angeles to 5,142, up 10. Rotterdam to 4,342, up 15. Genoa to 5,756, up 12. Six blank sailings were announced on the Transpacific for the following week, three on Asia-Europe. Carriers are putting surcharges through to manage front-loading ahead of expected July tariff changes and the 1 July bunker adjustment. Drewry expects further increases, and reckons the US-Iran interim agreement should stabilise oil and bunker prices and take some of the cost pressure off.

Two things moved in those nine days and they are not the same thing. The index went up 536 dollars a box, about 15.6 per cent. And the Transpacific blank sailing count went from three to six. The first is demand. The second is capacity being pulled back. Nine days ago the story was a demand spike against nearly full deployment. This week it is a demand spike and the carriers beginning to manage it. That difference is the whole piece, because it changes what I should be doing with a pen in my hand.

Where an eighteen-month high breaks next is Vivian Zhao's call and I am not going to guess at it; she has the filings and the surcharge notices in front of her every week. My problem runs on a different clock. My annual contract is up for renewal, and 3,969 is the level it gets signed at if I sign it this month. So I have exactly one decision in front of me: how much of next year's volume do I commit at the top of an eighteen-month range, and how do I structure the rest so that being wrong is survivable.

Start with the mistake, because I have made it. The instinct at a high print is to lock everything, on the reasoning that if it is expensive now it will be worse later. The opposite instinct, stay entirely on spot and wait for it to fall, is the same mistake with the sign reversed. A twelve-month fixed rate is not a view on the market. It is insurance, and insurance is priced at the moment you buy it. Buying a full year of cover on the highest weekly print in eighteen months is buying at the top of the range and calling it prudence. I did exactly that once. It looked responsible in the meeting and it cost me for four quarters.

Going all spot is not braver, it is just unhedged. Drewry expects further increases in the coming weeks and the blanking list supports that reading: six on the Transpacific and three on Asia-Europe is capacity being withdrawn, and capacity withdrawal in a peak season is not a signal that rates are about to be kind. If I run a whole year on spot into that, I have no floor under my unit cost, and I will be explaining a variance to finance every month instead of once. Volatility in the P&L is a real cost even when the average works out fine.

Before I sign anything, I go back to my own ledger, because the index tells me about the market and my ledger tells me about me. Three numbers. How much of next year's forecast demand is already covered by booked orders or confirmed programmes; if it is under half, I am signing against a forecast and the commitment should be sized smaller. What my stock-to-sales ratio is running at on the affected categories; if it is already above plan, part of my peak exposure is inventory I do not actually need to move at peak rates.

And how many days of safety stock I hold against my lead time. On a sixty-day ocean lead time, ten extra days of cover is a genuine substitute for a slice of expensive space, and it is often cheaper per unit than buying capacity on the spot market in the week the index hits an eighteen-month high. Inventory and contract capacity are substitutes. Most procurement teams run them on separate budgets and then wonder why they paid twice.

Now the arithmetic, and let me be explicit about which numbers are mine. Assumption: my annual Transpacific plus Asia-Europe book is 1,200 FEU, which is 100 a month; replace it with your own. Assumption: I can sign the fixed tranche at the current 3,969 level. Assumption for the downside: the index reverts to roughly 2,791, which is where the arithmetic puts it before the 23 per cent move at the start of June; that figure is derived from the reported numbers, not a forecast.

Assumption for the upside: spot averages 4,500 across the year. On those assumptions: all-fixed costs 1,200 times 3,969, about 4.76 million dollars. All-spot at 2,791 costs about 3.35 million, so all-spot saves about 1.41 million if the market falls back. All-spot at 4,500 costs about 5.40 million, so it costs about 640,000 more. Split 60 per cent fixed and 40 per cent spot: 720 boxes at 3,969 is about 2.86 million, plus 480 at 2,791 is about 1.34 million, so about 4.20 million in the downside case, which is roughly 565,000 better than all-fixed.

In the upside case it is 2.86 million plus 480 at 4,500, about 2.16 million, so about 5.02 million, roughly 255,000 worse than all-fixed. That is what insurance costs and that is what it buys. I give up about 1.41 million of the good case against going all spot, and I cap the bad case at about 250,000 instead of 640,000.

Which brings me back to the nine days, because they change the sizing. On 4 June the Transpacific showed three blank sailings, and the report read that as confidence in demand. On 18 June it is six, against three on Asia-Europe. The carriers went from running nearly everything to pulling sailings in a week and a half. I do not read that as a plot against shippers; it is what any carrier does when it wants to hold a rate through a peak, and on their side of the table I would do the same. What it tells me is that the capacity argument for waiting has weakened again. Nine days ago I could say the space is there and the price is the problem. This week the space is being managed as well, and there is less reason to assume a full book gets cheaper by itself.

The first structural thing I would insist on this year is the bunker line. There is a 1 July bunker adjustment in the report, and Drewry expects the US-Iran interim agreement to stabilise oil and bunker and perhaps ease some cost pressure. I have no view on oil and I am not going to pretend to one. What I do know is that a fuel adjustment is a pass-through of a cost the carrier does not control, and it has no business sitting inside my fixed rate. If I fix it, I am underwriting oil risk for twelve months, which is not my business and not a risk I am paid to carry.

So I want it separated: base ocean freight fixed or index-linked, and the bunker element adjusted monthly against a published index with the formula written into the contract. Assume the bunker element is 12 per cent of my all-in ocean cost; that is my figure, take it off your own quotation. On 1,200 boxes at 3,969, 12 per cent is about 476 dollars a box, around 571,000 dollars a year of exposure. Whether the interim agreement holds or not, I want that number on its own line where I can watch it move, not buried inside a rate I cannot reopen.

Second structural item: the discharge port, and this week's report hands me the spread twice over. New York at 6,769 against Los Angeles at 5,142 is 1,627 dollars a container. Inside Europe, Genoa at 5,756 against Rotterdam at 4,342 is 1,414 dollars a container. Those are the two biggest levers in this week's numbers and neither of them gets mentioned in a sourcing meeting. The test takes an afternoon: take my inland rate from the alternative port to my distribution centre, plus the cost of the extra transit days, both of which are my own numbers and not in this report, and compare the total to the spread. If inland plus the time penalty comes in below 1,627, the west coast routing wins on money and the only question left is whether the schedule can carry it. On a 1,200-box book, moving a quarter of it, 300 boxes, is about 488,000 dollars. That is worth an afternoon.

The batches, then. What I am actually going to do with the third quarter is split it into three and give each batch its own rule instead of making one blanket call. Batch one is anything that has to clear before a July tariff date. That is the front-loaded book from earlier in the month; the decision on it is already made and the only open question is whether the arrival dates still hold, which is a tracking job, not a strategy job. Batch two is July and August volume, and this is the part I am willing to flex: earlier, later, or through a different coast, depending on what the next two weekly prints and the next two blanking lists say. Batch three is September onward, and that is the piece that goes into the annual contract. Treating those three as one decision is how people end up signing a whole year at the worst possible moment.

On timing, then. I do not sign a twelve-month fixed rate in the week the index prints an eighteen-month high. That is not a market call, it is just not buying insurance on the day it is most expensive. I wait for the next print, which is roughly a week out, and I write my two triggers down before it arrives. If the next print rises 10 per cent or more, or the Transpacific blanking list goes to eight or more, I sign the base tranche immediately at 50 per cent of annual volume and stop waiting, because that is a market that is not coming back this quarter.

If the index falls for two consecutive prints and blankings come back to three or fewer, I push the signature to the week of 15 August, after the peak has broken, and hold the volume on a monthly extension in the meantime. Writing the triggers down before the number appears is the entire point. Otherwise I will rationalise whichever way it goes.

Inside the contract, four things. A volume band rather than a hard commitment: I want to commit a floor and hold an option on the upside, because a minimum quantity commitment that is too heavy going into a peak season is me paying for space I cannot fill. Symmetry on any index link: if it goes up by formula it comes down by formula, with a floor and a ceiling that I name rather than discover. And let me be plain that the numbers I open with, a ceiling around 4,400 and a floor around 3,200, are negotiating positions, not a view on where the market is going. The bunker element separated on a monthly formula, as above. And change control on surcharges: a notice period, a stated basis, and an expiry date, so a peak-season surcharge does not quietly become next year's base rate. Those four are worth more to me than fifty dollars a box on the headline.

Then there is the substitute I keep coming back to, which is inventory. Before I commit another 300 boxes of annual capacity at this level, I want to know what ten or fifteen days of extra cover costs me on the categories in question, because on some of them it is cheaper than the capacity and on some of it is not. If I am carrying thirty days of safety stock against a sixty-day lead time and the peak is going to be unreliable, pushing to forty-five days on the fast movers may buy me more protection per dollar than a fixed-rate commitment at the top of the range. The two compete for the same budget and they are rarely compared in the same meeting. My rule of thumb is that I would rather hold inventory I can sell than capacity I have to use, but only up to the point where the carrying cost on the extra days crosses the premium I am paying for the space.

One more pass from the demand side, because that is where this finally lands. If I cannot absorb the increase and have to put it into price, I need to know what that does to volume before I commit to a year of capacity. Take the increase as the gap between 3,969 and the 2,791 level the arithmetic puts the index at before this run, which is 1,178 dollars a box, or about 1.41 million dollars across a 1,200-box year. Assume I pass two thirds of it through to price on a line running at 35 per cent gross margin.

I do not need a model to know which question to ask: at what price does my volume fall far enough that the capacity I committed becomes a minimum quantity I cannot fill. That is the real risk in signing big at the top, and it is not a freight risk, it is a demand risk. It is also why the volume band matters more to me than the headline rate, and why I would rather sign a floor with an option than a number that flatters me this week.

And on sourcing, the honest time scale. Multi-sourcing is the answer everyone reaches for when a cost line runs away from them, and on a seasonal-to-annual horizon it usually is the right answer. It is also the slowest one. A second qualified source on the categories I buy means audit, sampling, a quality ramp and usually one season of lower yield, and none of that can be compressed by wanting it badly in June. What I can do this quarter is narrow: put a number on how much annual volume would have to move for a second source to be worth the switching cost, and start qualification on the two categories where freight and tariff exposure per unit is highest. That work pays off in a year. It does nothing for the signature sitting in front of me this week, and pretending otherwise is how a company ends up with a half-qualified supplier and a worse cost structure than it started with.

The traps this time round are mostly about the signature, not the sailing. Signing the full year because the number frightened you. Signing with a minimum quantity that assumes demand you have not booked. Leaving the bunker adjustment inside the fixed rate so that you own oil risk you cannot hedge. Taking an index link with a ceiling but no floor, which is a one-way ratchet with better paperwork. Signing in the same week as the highest print in eighteen months because the renewal date said so, when the renewal date is a diary entry and not a deadline, and most carrier contracts can be extended month to month while you wait. Comparing the new fixed rate to last year's fixed rate instead of to the spot alternative, when last year is gone and the only two things on the table are this contract and the market.

And the case against me, because there is one. If the index keeps climbing through July and into the peak, then waiting costs more than signing, and my 60 per cent fixed tranche turns out to have been too small. Drewry does expect further increases, and six blankings is consistent with that. If I am wrong, the cost is bounded: on the assumptions I wrote down, the 60/40 structure in the bad case is about 250,000 dollars worse than having signed everything, and I can live with 250,000 in exchange for 565,000 of the good case. What I could not live with is the version where I committed 1,200 boxes at the top of an eighteen-month range and watched the market fall in September while my competitors re-priced. Bounded regret is the product here. Not the best outcome, the survivable one.

One more thing I would say to my own team before this goes to signature. The 536-dollar move in nine days is not a freight department event. It is a unit-cost event, and it will show up in the gross margin line of every SKU that moves on these lanes, for as long as the contract runs. Somebody has to own that number by name, and it should be the person who signs the contract rather than the person who books the container. If procurement treats a renewal in a peak week as an administrative chore, the cost of that attitude is about 4.76 million dollars on my assumptions, which is roughly the whole annual ocean bill.

If I were signing it this month, here is the shape. I would not sign the full year this week; I would wait one print and I would write my two triggers down before the print arrives. I would split the book 60 per cent index-linked with a floor and a ceiling I name myself, 25 per cent on a six-month mini-tender to keep a second carrier honest, and 15 per cent on spot so that I have a live read on the market all year. I would pull the bunker element out and put it on a monthly formula against a published index.

I would rerun the discharge-port decision on the 1,627-dollar transpacific spread and the 1,414-dollar European spread, and move roughly a quarter of the book if my inland numbers hold. I would hold the third-quarter flex volume in its own batch with its own rule rather than folding it into the annual. And I would tell finance in advance, in writing, that on my assumptions the bad case is about a quarter of a million dollars worse than the safe-looking decision, because I would much rather explain that than explain a year locked in at the highest print in eighteen months.

↑ Back to the key points

— By Grace Yang