Drewry's composite World Container Index closed at $4,465 per 40ft on 3 September, unchanged on the week. Look underneath that and nothing about this market is flat: Shanghai–Los Angeles rose 5% to $7,185, while Shanghai–Genoa fell 10% to $4,368. Two trades, one index, opposite directions.
The composite is doing what composites do — averaging away the only thing worth knowing.
The transpacific is being priced by congestion, not demand
US import volumes are not what is holding transpacific rates up. The National Retail Federation's tracker has July at 2.21m TEU, down 7.6% year on year, with August through November all forecast lower again. Full-year 2026 lands around 25.5m TEU, essentially flat against 2025. Retailers pulled their peak forward into May and June ahead of the late-July tariff changes, and the back half of the year is paying for it.
So the strength is on the supply side. Roughly 3.92m TEU — about 11% of the global container fleet — was sitting in congestion queues at the start of this month, and North Asia accounted for 54% of it. Ships waiting to berth are ships not carrying anything. When a tenth of the fleet is parked, effective capacity tightens without a single vessel leaving the water, and carriers price accordingly.
Shanghai–New York at $9,587 tells you the same story from the other end. That is not a demand number.
Asia–Europe has the opposite problem
Europe has had two consecutive months of soft demand and, unlike the transpacific, it is getting capacity back rather than losing it. About 19% of Asia–Europe capacity now transits Suez after moves by Maersk, Hapag-Lloyd and MSC, and every loop that comes off the Cape route shortens its round trip by a week or more. CMA CGM's INDAMEX went back through the canal and cut two weeks off a 77-day rotation, freeing two ships in the process.
That is the mechanism people keep missing. A returning loop does not add ships to the trade. It adds sailings, because the same ships complete more round trips a year. The Cape diversion was, in capacity terms, a two-year-long absorption of surplus tonnage, and unwinding it releases that surplus back into a market that never actually needed it.
Shanghai–Rotterdam at $4,092, down 5% on the week, is what the first stage of that unwind looks like.
Which of these two moves is temporary
The transpacific one. Typhoon Saudel is the immediate cause of the North Asian queues, following Bavi, Noul, Dolphin and Narra earlier in the season. Weather-driven congestion clears; berth waits at Shanghai and Ningbo hitting ten days is not a structural feature of those ports, it is a backlog. Once the queue drains, the capacity comes back and the pricing power goes with it.
The Asia–Europe move is the structural one, and it has further to run. If the Suez return completes, one analysis has container demand contracting 8.7% year on year in the first half of 2027 on a cargo base that is still growing 6.6% — the gap being the artificial tonne-miles the Cape route manufactured.
I would treat that 8.7% carefully. It assumes a full, sustained return, and carriers have shown they will move a service back through the canal one week and quietly reroute it the next if the security picture shifts. But the direction is not really in doubt.
What to do with a booking this month
On the transpacific, do not lock length right now. You would be pricing in a weather backlog as though it were a demand recovery. If you can hold, hold.
On Asia–Europe, the temptation is the reverse — grab the falling rate for as long as you can get it. Resist that too, at least for anything past Q1. The market you are trying to lock is the one that has not oversupplied itself yet.
And whichever side you sit on, check what your quoted transit actually means at the moment. With ten-day berth waits in North Asia, the delay is now longer than the buffer most quotes carry. A cheap rate on a ship that cannot get alongside is not a cheap rate. The live picture is on our sailing schedules if you want to see which strings are still holding their windows, and the Shanghai to Rotterdam lane page has the current transit spread.
What I am watching next is the October general rate increase announcements on the transpacific. If carriers pull capacity before they announce, they believe in the level. If they announce first and blank later, the queue has already started clearing and they know it.
