Supply Chain Action Points
Read this first — the conclusion, and the moves to make:
- Before 12 June, pull every open air booking on China-origin lanes and record the real cutoff hour, not the one in the routing guide; where it has moved earlier than 16:00, tell the warehouse in writing the same day.
- Re-quote all June air volume within 48 hours and accept nothing with a written validity under seven days; anything quoted before 1 June is already stale at $3.40.
- Split June volume 70 percent blocked allotment and 30 percent spot, with a trigger at 15 percent above $3.40: above that line the shipment does not book until a second carrier has priced it.
- Apply a two-roll rule: any consignment rolled twice moves to another carrier within 12 hours, without waiting for the next available from the same forwarder.
- By 15 June, identify the 20 percent of SKUs and accounts that carry the promise and pre-position three days of buffer at the destination warehouse for the early peak, rather than waiting for the peak to prove it is early.
Global air cargo spot rates rose 41% year-on-year in May 2026 to an average of $3.40 per kilogram, driven by sustained e-commerce cross-border volumes, semiconductor supply chain restocking, and pharmaceutical cold-chain demand out of Asia and Latin America. Overall air cargo demand increased approximately 4% year-on-year, while global capacity remained constrained by ongoing Middle East airspace closures that have forced carriers onto longer routings, reducing effective available tonne-kilometers. The Asian export lanes saw the strongest rate gains, with China-origin air freight to Europe and North America trading at significant premiums above the global average. Forwarders expect the traditional summer air freight peak to arrive early in 2026, compounding already-tight ocean-to-air conversion demand as shippers seek alternatives to congested and expensive ocean freight lanes.
The Analysis
You know the call. It lands a few hours before cutoff, and the forwarder is not ringing to confirm the booking. The allotment is gone, the number on the screen is not a number he can hold, and the only flight with space leaves the day after tomorrow. That call is what a 41 percent year-on-year increase feels like at four in the afternoon, and it is the part of the figure nobody puts in the headline.
Global air cargo spot rates averaged $3.40 per kilogram in May 2026, up 41 percent on the same month last year, on demand growth of about 4 percent. Read those two numbers side by side and the story is not demand. Demand moved by a rounding error compared with the price. The price moved because the capacity did not.
Air freight is the one mode where the promise is quoted in hours, so let us walk the clock instead of the chart: cutoff, pickup, linehaul, customs, delivery, and find the hour that gives way first.
The first hour that gives way is not the flight. It is the six hours before cutoff, and it gives way quietly, because nothing fails. The booking still confirms. What changes is that the confirmation stops being a commitment. A forwarder holding $3.40 on Monday comes back on Thursday with a price he cannot repeat, or with space on a flight a day later. The 41 percent is not a wall you hit at the airport. It is a run of small retreats, each a few hours wide, and by the time the cargo is late nobody can point at the moment it went wrong.
Here is what the May average does not tell you, and it is the part I would want on the table in any budget meeting. A spot rate average is built from transactions that were priced and flown. Cargo that got rolled, or re-quoted after the first quote expired, or pushed onto a charter because the scheduled lane had nothing left, is not inside that $3.40. It sits outside the sample. So the print is an average of the cargo that cleared, which means the market faced by anyone without a blocked allotment is worse than the headline, not equal to it. When your forwarder says the market is $3.40, ask what he paid for the shipment he rebooked on Wednesday afternoon. That is the real number, and it never reaches the monthly average, because it was never the first price.
Capacity explains why the retreats cluster in the afternoon. Demand rose about 4 percent. Rates rose 41 percent. If demand were doing the work, those two numbers would sit closer together. What moved is effective capacity. Aircraft flying longer routings around closed Middle Eastern airspace still burn the block hour, still carry the same payload and still land, but each tail delivers fewer tonne-kilometres per week than it did on the old routing. The fleet did not shrink. Its output did. Airlines do not publish that as a capacity cut. They publish it as a schedule that departs twenty minutes later and arrives four hours later, and the cutoff moves with it without an announcement.
Along the clock, the damage lands in a fixed order. The booking desk feels it first, in quotes that expire inside a day. The warehouse feels it next, when the cutoff that used to be 18:00 becomes 14:00 and nobody tells the pickers. The customer feels it last, at the delivery scan, which is also the only place it becomes visible and the only place it becomes expensive. Who is insulated: anyone on blocked space or an annual allotment, because that capacity was contracted before the spot market moved, and anyone whose product can wait, because a two-day slip on a six-week ocean routing is noise. Who is not: the shipper whose promise is counted in days, and whose cargo is e-commerce replenishment, semiconductor line-feed or cold-chain pharma. Those three appear in the same sentence as the rate rise for a reason. Each has a hard date on the other end.
Forwarders expect the summer peak to arrive early, and that is the sentence with a date attached. A peak that lands in June instead of late July does not merely lift the average. It compresses the window in which you can still buy space anywhere near the quoted level. The transmission is neither instant nor even. A rate agreed this week is charged at uplift, days later on Asia-Europe, and lands in your landed cost at month end. The customer promise breaks earlier than the invoice does. The delivery scan is where a shipper first learns that the transit time he sold in April was built on April's block hours.
Put a number on it, with the assumptions on the table. Take 6,000 kilograms a month on a China-origin lane to Europe. At $3.40 that is $20,400 a month. Back out the 41 percent and last May's equivalent is about $2.41, or $14,460 on the same weight, so the increase is roughly $5,940 a month and about $17,800 a quarter. That is the visible half, and it is the easy half, because it is a budget line you can see coming. The invisible half: assume 40 air consignments a month and a roll rate of one in twenty, which is two shipments. If a roll costs 24 hours and one of those two is a line-feed component for a plant running three shifts, the cost is not the freight difference. It is the two days of output that stopped while the rebooking email went back and forth. Assumptions stated: 6,000 kilograms, 40 consignments, 5 percent roll, 24-hour roll. Change any of them and the arithmetic changes. The shape does not.
The trap is treating a spot rate as a planning number. Spot is the price of a specific kilogram on a specific flight, and its validity is counted in hours. A budget built on $3.40 assumes you can buy at $3.40 all quarter, which in a market climbing 41 percent is an assumption about luck. The other trap is the cheap quote. When space is this tight, the quote that is thirty cents under everyone else usually has the latest cutoff and the weakest allotment, and you find that out at the hour when you can do least about it. So the order of work is not negotiating. It is protecting what is already moving: know which shipments are riding a promise you can no longer keep, call those customers before the scan does it for you, and buy the space you need for the next fourteen days before you sit down to argue about August.
↑ Back to the key points