Supply Chain Action Points
Read this first — the conclusion, and the moves to make:
- Map your origin footprint against the 38 per cent this month: list every lane loading in North Asia and the gateway it uses, and if more than 60 per cent of volume passes through one gateway, qualify a second one before the third quarter.
- For Europe-bound cargo, bring the congestion assumption forward by four weeks and price the 13 per cent European backlog into August arrival dates, even though the headline is about Asia.
- Build your own baseline for the unmeasured 22 per cent: if you load in South Asia, the Gulf, Latin America or Oceania, track your own roll rate weekly, because nobody publishes congestion for your lane.
- Size your position honestly before you ask anyone for priority: 300 FEU a month is 600 TEU, about 0.04 per cent of the North Asian backlog, so spend the effort on routing and timing instead.
- Diarise the day any Cape diversion is formally dropped, and hold three weeks of extra buffer for the six weeks that follow, because the first re-phased arrivals will bunch.
Linerlytica data cited by The Loadstar shows nearly 3.7 million TEU - about 11% of the world's box-ship fleet - was waiting for berths, the highest share in four years. North Asia accounted for 38% of the global backlog, North Europe 13%, and Southeast Asia, the Mediterranean and Africa 9% each. Sustained Cape-of-Good-Hope diversions and vessel bunching are blamed.
The Analysis
Numbers are easy to read as one thing. 3.7 million TEU waiting for a berth, roughly 11 per cent of the world's box fleet, the highest share in four years, and the eye stops there, satisfied that it has understood. It has not. The number that carries the argument is not the total but the split underneath it.
Of that backlog, 38 per cent is sitting in North Asia. North Europe holds 13. Southeast Asia, the Mediterranean and Africa hold 9 each. Linerlytica's data, as reported, attributes the pile-up to sustained Cape of Good Hope diversions and the vessel bunching those diversions produce.
Draw those shares on a map and the shape is unmistakable. The queue is not arranged around the world's consumption centres. It is stacked at the factory gate, which is the one place a queue is most expensive and least visible.
Thirty-eight per cent does not say Asia is congested. It says the export side of the network is congested, and those are different diagnoses. North Asia is where boxes are loaded, not where they are unpacked, so the ships at anchor there are largely full and waiting to work rather than empty and waiting to discharge. That distinction decides how the queue clears. A queue at the consumption end is cleared by demand falling away. A queue at the production end can only be cleared by schedules coming back into phase, and schedules are re-phased at the service level, not at the port level. Nobody fixes this by adding a crane.
The mechanism is phase, not luck. A liner rotation is a closed circle: a fixed sequence of calls repeated on a fixed interval. The Cape routing lengthens one arc of that circle, and when the ship comes back around it is no longer at the right point in the sequence. The error does not wash out. It propagates through the next loop and the one after, until several ships are re-imported into the same berth windows. The bunching was not decided by anybody. It is arithmetic. The port records it as congestion because it sees a cluster of arrivals; the network records it as capacity withdrawn because the circle no longer closes. Same event, two names, and only one of them is on the front page.
The layered bill differs by contract terms. An FOB seller holds the box until it is loaded, so he pays the yard, the detention and the cost of the missed vessel; the freight risk is not his, but the timing risk is entirely his. A CIF or DDP buyer owns the freight risk and yet sees only an ETA, with no window into how long the line at origin actually is. Large shippers can negotiate at both ends. Mid-sized shippers can only accept at both ends. Small shippers learn they were rolled after the fact.
Direction decides who feels it when. North Asia's 38 per cent is today. North Europe's 13 per cent is next month, because the European queue is fed by the same circle from further downstream and takes three to five weeks to fill. Anyone importing into Europe should already be carrying that lag inside the August arrival plan rather than treating it as a separate event.
Here is what the coverage left out. The five named regions sum to 78 per cent, which leaves 22 per cent unaccounted for. That remainder is spread across South Asia, the Gulf, Latin America and Oceania, none of them large enough individually to appear in a headline, and all of them invisible in the total. For an importer loading in Latin America or Oceania, the whole of his exposure lives inside that unmeasured residual. Africa is the sharper case: the continent handles nothing like 9 per cent of world container volume, so a 9 per cent share of the global backlog means its congestion is far heavier relative to its own trade than the figure suggests. Small markets absorb congestion out of proportion and then vanish from the summary.
Run a number and it gets uncomfortable. Thirty-eight per cent of 3.7 million TEU is about 1.4 million TEU sitting off North Asia. Assume you load 300 forty-foot containers a month from northern China, which is 600 TEU, and that figure is my assumption rather than anything in the report. Six hundred divided by 1.4 million is 0.04 per cent. What that proportion means is that you hold no negotiating position in this queue whatsoever. The only levers left are structural: book earlier, split gateways, move what can wait outside the window. At this scale, asking for priority produces nothing but a longer email thread.
The trap is reading 38 per cent as a signal to route around Asia. You cannot, because the alternative gateways sit inside the same 38 per cent; you would be changing the terminal, not the queue. The other trap is assuming the queue disperses once the Cape diversions stop. It does not. In the first four to six weeks after the circle shortens, vessels that were spread across a long rotation come back together, and arrivals get denser than they are now. Relief arrives on the second loop. So on the day it is announced, do not translate it into a delivery promise.
The map did not get bigger. It changed shape, and the shape now has a knot in the middle of the manufacturing world. A knot is not a wound. But you plan around a knot differently than you plan around an open road.
↑ Back to the key points