Against the annual forecast

Every cargo airline sets an annual forecast in the autumn, builds a budget on it, and then spends the following year explaining the difference. By February the number is already wrong. By December it is still the number everyone is measured against.

This summer is a clean illustration. IATA's June figures had global demand up 8.5% year over year. A month later, July came in at 3.9%. Capacity growth halved over the same stretch, to 1.7%, and jet fuel was up 57% on the year. Underneath the headline, the lanes moved in opposite directions: Asia–North America grew 9.2% while Europe–Middle East fell 16.1% and Middle East–Asia fell 14.1%. A single number for the year cannot contain all of that, and it was never going to.

The instinct is to treat this as a forecasting-accuracy problem — buy a better model, hire a better economist, forecast harder. That misreads what went wrong. The annual number is not bad because the analysts were careless. It is bad because one figure is being asked to do three different jobs, and each of those jobs needs a different horizon.

The budget needs a year. The fleet needs longer than that — a freighter decision made this quarter flies for a decade. Capacity allocation and allotments need a month, because that is how fast lane mix moves. Price needs a week, sometimes a day. When all four are read off the same annual line, the fast decisions get made with stale information and the slow decisions get made with noise.

The commercial teams that handle this well stop defending the annual number and run separate clocks instead. A rolling twelve-month view, by lane, for anything that touches the fleet. A monthly re-forecast that actually moves capacity and allotments, rather than a variance report that explains why they didn't move. And a weekly read — bookings, materialization, spot versus contract — that sets price. The annual figure still exists. It becomes the frame the year is judged against, not the plan the year is run on.

There is a cultural cost to this, and it is worth naming. A forecast that changes monthly looks less confident than one that holds for a year. Boards prefer a number that doesn't move. But the market moves whether or not the forecast does, and a plan that stays put while demand swings from 8.5% to 3.9% in thirty days is not stable. It is simply late.

A forecast is not a promise about the future. It is a decision about how often you are willing to change your mind.