A modern container ship arrives at the breakwater doing four or five knots and immediately becomes the least maneuverable object in the harbour. That’s the part people miss when they watch the process. The ship isn’t being helped along. It has effectively stopped being able to steer itself, and three small boats are now doing the job its 60,000-kilowatt main engine can’t. The reason is that a rudder is a wing.

The number that produced the slowdown headlines was an export gain of roughly 24% year on year in July, down from about 27% in June. Customs also reported the trade surplus narrowing to $112.5 billion from $125.6 billion, with imports up 27.5% against a 36% jump the month before. Both prints came in ahead of what analysts expected. In almost any other year, a 24% export print would have been the story by itself rather than the evidence of decline.

The headline everyone will carry is the 7% move and the beat: underlying EBITDA of $3.0bn for the second quarter against a company-compiled consensus of $2.12bn, revenue up 20% year-on-year to $15.8bn, EBIT of $1.6bn versus $845m, group EBIT margin at 10.0%. The number that actually describes the year is buried in the guidance table. Maersk has now raised full-year underlying EBITDA three times — $4.5-7bn on 7 May, $8-10bn on 29 June, $10.

A mid-size container ship with working bow thrusters, a competent master and a pilot on the bridge does not, on a calm afternoon, need two tugs to leave a harbour. It takes them anyway. The reason is not seamanship. It is that the port’s marine services regime priced a low-probability failure and decided the shipowner would pay for it on every movement, every day, whether the wind blows or not.

Every physical object in the historic centre of Venice arrived by hull. Bread, cement, hotel linen, restaurant deliveries, scaffolding, refuse going the other way, the coffin at a funeral. There is no road network behind the facade doing the real work. The canal in front of you is the road network, and the moored boats in this frame are the freight fleet at rest. The system has three legs. Goods reach the edge of the lagoon by road at Mestre and Marghera, transfer to flat-decked motor barges at the port or at Tronchetto, and then run into the city by water.

A port announces it handled ten million TEU last year. The natural reading is that ten million boxes crossed the quay. They didn’t. The real number was probably closer to six million, and depending on the port, a large share of those boxes never entered the country at all. TEU stands for twenty-foot equivalent unit. It is a normalising unit, invented so that ports and carriers with mixed box lengths could compare capacity on one scale.

The honest answer is that container traffic is further gone than tanker traffic, and for reasons that have almost nothing to do with how the war ends. Crude can move on a shadow fleet with the transponders off. A 19,000 TEU boxship on a fixed weekly rotation cannot. What is being lost is not the waterway. It is Jebel Ali’s position as the default first port of call for the Middle East.

A cruise ship at a commercial berth looks like a nice piece of civic branding. Passengers on deck, cranes in the background, the whole port working at once. What the picture does not show is that for the length of that call, a container terminal has been operating with one hand tied behind its back. Most mid-size ports that handle both passengers and boxes never built a separate cruise facility. The cruise business arrived later, or seasonally, or as a political initiative, and it got slotted into whatever quay had the depth and the road frontage.