Drewry WCI Hits a 15-Week High as Carriers Pull 56 Sailings From Weeks 44-48

The Analysis

Drewry's week-44 note leads with a number that sounds like a verdict. The World Container Index rose 2 per cent in the week of 22 October, a second straight weekly gain and its highest print in 15 weeks. Behind it sit four lane numbers and one withdrawal count. Shanghai-Rotterdam added 3 per cent, Shanghai-Genoa 5 per cent, Shanghai-Los Angeles 2 per cent and Shanghai-New York 1 per cent. Across weeks 44 to 48 carriers have pulled 56 sailings out of 714 advertised voyages, an 8 per cent blank rate, with Asia-North Europe and the Mediterranean taking 54 per cent of those cancellations and transpacific eastbound 32 per cent.

And then the line most people read straight past: next week transpacific blanks fall from six to two, while Asia-Europe blanks rise from two to eight.

That last sentence is the whole story, and it is printed in the same note as the headline number. Two trades, one index, opposite directions. Read the 2 per cent and act on it and you have acted on the average of two things moving apart.

Four things below. Whether the 2 per cent has a base under it. What 56 out of 714 converts into when you are the one booking. Which lane still has fuel and which one has just been handed its capacity back. And the number in this note nobody quoted, which decides whether you lock this week or wait two.

Every figure is used as printed. Where the note gives a percentage without the level behind it I say so.

Let me start with the number everybody quoted, because it is the least usable one in the note. Two per cent. It is a percentage on a composite index, and the note prints no level behind it. I cannot tell you whether that 2 per cent is 30 dollars a container or 90, and neither can anyone who only read the headline. That gap is not academic. Two per cent on a 1,500 dollar lane is 30 dollars a box. The same 2 per cent on a 4,500 dollar lane is 90. Same headline, three times the money. Before you decide whether the print is worth a phone call, get the level: ask whoever sends you the index to send the rate underneath it with it.

Now the second claim, the one with the drama in it: highest in 15 weeks, and a second straight weekly gain. Read the two together and they say more than either says alone. Fifteen weeks before the week of 22 October is roughly the week of 9 July. 'Second straight gain' means the week before was also up. So the shape is a high around early July, thirteen weeks that did not beat it, then two weeks up. The 15 weeks is the headline. The two weeks is the evidence. Two observations is the smallest sample that can be called a direction, and it is being asked to carry a 15-week frame. The note has not shown you anything you could bet on.

Here is the test that settles supply against demand, and it takes one line. A demand recovery lifts lanes together. Boxes do not care which ocean they cross, so if European and American buyers are restocking at the same time the lane prints move the same way and in the same order of magnitude. What the note prints is 5 per cent on Shanghai-Genoa and 1 per cent on Shanghai-New York, in the same week. A five-times spread, on broadly similar cargo, out of the same origin. Demand does not make that shape. Capacity withdrawal does, because withdrawal is a decision taken per string and it lands unevenly. Rotterdam 3 and Genoa 5 are the two ends of the trade where the cancellations sit. Los Angeles 2 and New York 1 are the two ends of the trade where they are being handed back. The spread is the signature of supply management.

Now the arithmetic on the withdrawal, because 8 per cent is a rate and rates hide counts. Fifty-six blank sailings out of 714 advertised voyages across weeks 44 to 48. That is 7.84 per cent, rounded to 8. Split it. Asia-North Europe and the Mediterranean take 54 per cent, about 30 sailings. Transpacific eastbound takes 32 per cent, about 18. Thirty plus eighteen is 48, so eight sailings, 14 per cent of the programme, sit in trades the note does not attribute. A seventh of the withdrawal being unlabelled is worth knowing: if some of it sits on your lane and nobody counted it in public, your lane's real withdrawal is worse than the split says. The other conversion matters more in practice. Eight per cent of voyages is roughly one voyage in thirteen. If you have been assuming the ship you booked exists, one in thirteen of those assumptions is now wrong.

Why now. Weeks 44 to 48 run from late October into the first days of December. The peak push is behind us. Volume off Asia normally comes off through November, before the pre-holiday and pre-New-Year bookings rebuild it. So carriers do the only thing a carrier can do when the boxes stop arriving and the rate is sliding: remove ships until the remaining ones fill. That is not a recovery. It is a defence, and it happens at the same point every year. The tell is in the size of the action. Carriers do not blank 56 sailings while they still believe demand is coming to save them. They blank when they have decided the rate is theirs to hold rather than the market's to set. Which means the 2 per cent is a policy outcome, and a policy outcome can be reversed on a Tuesday.

Two things are wrong with treating that 8 per cent as a capacity number, and this is the part nobody quoted. The first is a denominator problem. The 714 counts voyages, not slots. A very large ship on a main Asia-Europe loop and a much smaller one on a secondary string each count as one. So 8 per cent is 8 per cent of departures, and it cannot be turned into a box count without the nominal capacity of each of the 56 sailings, which the note does not print.

If the cancellations skew to the big mainline loops, the TEU removed is materially more than 8 per cent of the TEU; if they skew to small strings, it is less. The 54/32 split tells you where the cancellations sit by trade, not by ship size, and the gap between those two readings is the gap between a tight market and a soft one. Use the 8 per cent for direction, not to work out how many of your boxes get rolled.

The other problem is worse, and it is the one that decides whether you lock this week. Fifty-six blank sailings is an announcement, not an outcome. It is what carriers said, in Week 44, that they intend to cancel between now and early December. Carriers reinstate blank sailings when the bookings fill, and they add new ones when the bookings do not, and both happen inside a fortnight. The note proves this against itself in its last sentence. Next week's transpacific blanks are revised from six down to two, and next week's Asia-Europe blanks are revised from two up to eight. The programme moved, in both directions, between one publication and the next.

So the number being used to explain the index is a number that changes after the index prints. You cannot build a hedging decision on a denominator its own author revises weekly. What you can measure is the revision behaviour, and that is a real number: count how many of the 56 actually sail blank, against how many were announced. If the actual count comes in under 60 per cent of the programme, under about 34 of the 56, the withdrawal is softer than advertised and the 2 per cent has lost most of its support.

Which brings me to the sentence I would build the whole decision on. Next week: transpacific blanks from six to two, Asia-Europe blanks from two to eight. Set that against the five-week programme of 56. Eight Asia-Europe blanks in a single week is 14 per cent of the entire five-week cancellation programme landing in seven days. Two transpacific blanks is 3.6 per cent. These two trades are not drifting apart by a little. Asia-Europe is taking four times the weekly cancellation rate it had last week, and the transpacific is cutting its own by two-thirds. The index printed both trades in the same week, added them together and handed you 2 per cent. Then, in the next line, it told you the two halves are heading in opposite directions. Act on the 2 per cent and you act on the average of two opposite moves, and you will be wrong on both trades at once: too late on Europe, too early on the Pacific.

Who this lands on, and how hard. Split by how you buy, not by what you ship. If you hold an annual contract on Asia-North Europe, the 3 per cent and the 5 per cent do not touch your invoice this quarter. What touches you is the 8 per cent, because a blanked sailing on your service means your boxes roll to the next departure and your cargo lands a week after the plan your warehouse was built around. For a contract shipper this note is not a price story at all. It is a service story, and the price number is decoration. If you buy spot on Asia-Europe, or you are small enough that the carrier protects its contract customers ahead of you in a tight week, you get both halves: the rate step and the roll.

The roll is the expensive one. A box rolled one sailing on Asia-North Europe is a box arriving roughly a week late, and on a 40-box-a-month programme that is the difference between a stockout and a call nobody enjoys taking. On the transpacific the weighting flips: blanks go down to two a week and the rate pressure is about to ease rather than tighten. Same week, same index, opposite exposure.

Timing, because 'up 2 per cent' has no date on it and your business runs on dates. Three clocks. The rate clock is fast: a spot quote moves within days of a withdrawal announcement, so if you are re-tendering in the next fortnight you are already paying some of it. The rollover clock runs one to three weeks behind the blanked week, because a sailing is cancelled after bookings are taken and the rebooking scramble lands on the following departures. The arrival clock is slowest: a rolled Asia-North Europe box loses its sailing, then loses the transit, and turns up four to six weeks after the week you originally booked. So if your November Europe sailings are already booked, watch the rollover clock, not the index. And if you are booking December now, you are booking into the weeks the programme says are heaviest, weeks 44 to 48, which means get the carrier to name, in the booking confirmation, the alternative sailing if yours is blanked.

Money, with the assumptions stated because the note will not give me the level. Assume you ship 40 boxes a month. Assume your last North Europe spot was 2,000 dollars a container, and put your own quote in its place. Rotterdam's 3 per cent is 60 dollars a box, 2,400 dollars a month on 40 boxes. Genoa's 5 per cent is 100 dollars a box, 4,000 a month. Los Angeles at 2 per cent is 40 dollars a box, 1,600 a month. New York at 1 per cent is 20 dollars a box, 800 a month. Now read those four against the next-week revision. The lane where you would pay 4,000 a month is the lane whose cancellations go from two to eight.

The lane where you would pay 800 is the lane whose cancellations go from six to two. The expensive lane has fuel in the tank. The cheap lane has just been handed its capacity back. A buyer reacting to this week's print does both things wrong at once: locks the 800-dollar lane out of caution, and haggles over the 4,000-dollar lane because it looks steep. At 40 boxes a month that gap is 3,200 dollars a month, and 9,600 across the three months the programme covers.

One more sum, because the roll is usually bigger than the rate. Same 40 boxes a month to North Europe, and assume your monthly volume sits on roughly three departures. At a one-in-thirteen blank rate the chance that at least one of those departures disappears inside a five-week window is not small, and each one costs you a week of arrival. On a lean book with a week of cover that is a stockout, and a stockout is not priced in dollars per container. It is priced in a customer.

What would make me wrong. Two conditions, both measurable, which is the point of naming them. The first is reinstatement. If carriers put back a meaningful share of the 56, if actual cancellations come in below about 34, which is 60 per cent of the programme, then the supply story is weaker than the announcement and the 2 per cent loses its cause. In that world the Asia-Europe prints flatten or come off inside a fortnight, and anyone who took a month on the back of a 5 per cent Genoa print is paying for a withdrawal that never sailed.

That counterfactual is not exotic: cancelling a sailing costs nothing and reinstating one earns money, so carriers do it every November. The second condition is demand. If the index keeps climbing into a fourth and a fifth week while the blank count falls, then it was not supply after all, and the right move would have been to lock in Week 44. The way to tell the two apart is to watch which variable moves first. Supply-led gains lead the blank count.

Demand-led gains follow volume. Both sit on one sheet: the weekly lane print, and announced blanks against actual blanks. Two columns, four weeks, and you will know which story you are in before your broker does. There is a third reading I cannot settle here: part of the 56 may be network consolidation rather than rate defence.

So which lane you lock this week, and which one you let run. Asia-North Europe and the Mediterranean: lock, now, for anything sailing in weeks 44 to 46. The reason is not the 3 per cent or the 5 per cent. The reason is that the cancellation count on that trade goes from two to eight next week, and Genoa's 5 per cent is the lane print that has already moved furthest, which means it has the least room left to be negotiated down. Take a month, not a quarter. A month is long enough to clear the heaviest cancellation weeks and short enough that you are not still holding a supply-driven print after the supply comes back.

Transpacific eastbound: do not lock. Two-thirds of that trade's withdrawal disappears next week, from six blanks to two, and its prints, 2 per cent to Los Angeles and 1 per cent to New York, are the two smallest moves in the note. Whatever is holding that rate up stops holding it up in about seven days. Re-tender in a fortnight, and use the fortnight to get a written commitment to the rotation and to the fallback sailing if the rotation changes.

Thresholds, because 'lock' without a number attached is a feeling, and I do not trade in feelings. On Asia-Europe, a single-week step of 5 per cent or more on your own lane, which Genoa has already printed, is the line. Cross it and take the month. A step of 3 per cent or less is not a lock yet: cap the agreement at two weeks and re-tender. On the transpacific the trigger runs the other way: two consecutive weeks with blank sailings at two or fewer, and the lane print still rising. That combination means the gains are no longer supply-driven, and it is the only point at which a transpacific fix is worth signing. And a floor under all of it: never sign a fixed-duration agreement benchmarked to a composite index. The composite is 2 per cent because two trades are being averaged, one rising and one falling. Index to the lane you ship, or do not index at all.

Three ways to get this wrong, and I have watched all three. The common one is booking by sailing week instead of by vessel. In a week carrying eight Asia-Europe cancellations, the sailing week is a fiction; the only thing that exists is a named ship on a named rotation. Another is treating the 8 per cent as a box count: it is voyage share, not slot share, and it moves. And the one that costs the most while hurting the least at the time: locking the transpacific because the index went up, then watching the print come off a fortnight later because the six cancellations you were quoted became two. That is 2 per cent paid for a withdrawal that lasted a week. At a 2,000 dollar base and 40 boxes a month, 1,600 dollars for the privilege of being early.

Two weeks up and a 15-week high. The 15 weeks is the part they lead with, and the two is the part they would rather you did not count. The sentiment content is running ahead of the evidence. What the note actually contains is a withdrawal of 56 sailings, split about 30 to Asia-Europe and 18 to the Pacific, and a revision that takes the Pacific's share down to two next week while taking Europe's up to eight. That is the trade. Lock Europe this week for November sailings, one month. Re-tender the Pacific in a fortnight. And then do the one thing nobody does: count the cancellations that actually sail, and compare them to the 56. If that number comes in under 34, the whole episode was a conversation rather than a withdrawal, and you will be holding the only confirmed figure in the market while everyone else is still arguing about the 2 per cent.

↑ Back to the key points

— By Vivian Zhao