Schedule reliability isn’t recovering – it’s splitting in two

Alt. headline: Why the 39% on-time average hides the real story

Alt. headline: Port congestion just broke shipping’s reliability rebound

Global on-time reliability climbed for three months, then slipped in June as port congestion bit back. But the headline average is the wrong number to watch – the real story is a chasm.

There is a comforting number doing the rounds in container shipping, and it is 39%. That was the global on-time reliability figure for May in Xeneta’s Global Reliability Scorecard, the third monthly gain in a row – 36% in March, 37% in April, then 39% in May – with average delays falling to 3.4 days, the lowest since July 2025. Xeneta’s Destine Ozuygur called it the closest the market has come to the 40% ceiling since December 2023. It reads like a recovery. Then June arrived and undid the momentum: on-time performance fell two points to 37%, delays widened again to 3.6 days, and not a single trade lane improved.

Figure 1 – A three-month climb toward the 40% ceiling reversed in June, with delays widening back to 3.6 days.

Congestion is the binding constraint

The cause is not a mystery, and it is not capacity. Blanked sailings actually fell in June to around 890,000 TEU, or 9% of planned capacity – the first month this year with fewer blanks than a year earlier. The problem is what happens when the ships arrive. As ShiftX’s Keith Gaskin puts it, deployed capacity only matters if a box can be discharged, handled and loaded on time, and this year that is precisely where the system jams. North European ports are the worst offenders; congestion at Rotterdam is cascading disruption across the network, and planners are now baking in large buffers just to absorb it.

The cost lands on the blameless shipper

Here is the part that should worry anyone moving cargo. Some alliances are chasing better schedule-adherence numbers, and to hit them, a delayed vessel will sail on time even if congestion means it has not finished processing its boxes. The consequence is perverse: a shipper who delivered and cleared its container on time can watch that box roll to the next voyage through no fault of its own. The carrier’s reliability metric improves; the customer’s cargo is left on the quay. Punctuality, measured at the vessel, can be bought at the shipper’s expense.

“Punctuality measured at the vessel can be bought at the shipper’s expense.”

An average that spans a 54-point gap

This is why the 37% composite is close to meaningless as a planning input. Underneath it sits a chasm. In the second quarter the Gemini Cooperation delivered 69% on-time with average delays of just 1.1 days, whilst Wan Hai – the only major carrier to go backwards – fell six points to 15%. At carrier level Maersk (58%) and Hapag-Lloyd (57%) led, more than ten points clear of CMA CGM at 46%. Ocean Alliance posted the biggest jump, up fifteen points to 43%. Reliability, in other words, is no longer a market condition you can read off a single line. It has become carrier-specific alpha.

Figure 2 – From Gemini’s 69% to Wan Hai’s 15%, on-time performance now varies more between carriers than the market average conveys.

The strongest case against me

The bull will say I am over-reading one soft month. The three-month trend was up, blank sailings are down, most trades improved – South America’s east coast leapt seventeen points to 54%, Far East–Europe fourteen points to 39% – and Ozuygur herself insists the global recovery is real. All true, and worth respecting. But a recovery whose average runs from a 25% Middle East trade to a 69% alliance is not a market converging on a reliability level. It is a market where the spread, not the mean, is the signal.

Figure 3 – The rebound is real but wildly uneven: South America’s east coast at 54% against a Middle East trade stuck at 25%.

“The spread, not the mean, is the signal.”

What I would watch

Stop reading the composite. Watch three things instead. The gap between the best and worst carriers on your specific lane – that is your real exposure, not the global average. Congestion at North European hubs, Rotterdam above all, because that is where reliability is currently made or lost. And rollover incidence, the quiet tax that a carrier’s pursuit of its own punctuality can pass straight to you. In a market this uneven, the operators who plan around the spread will out-execute the ones who trust the headline.