Kuehne+Nagel Q3: Turnover Up 22% to CHF 7.4bn as Air Logistics EBIT Soars 84%

The Analysis

I have spent more nights than I care to count standing on the cargo apron at Frankfurt waiting for a main deck to close, and the thing that apron teaches you is that you never read a freight forwarder's quarter from the press release. You read it from the deck. When the deck is full of server racks and wafer equipment, when the trucks outside are running an hour late because there are not enough of them, somebody's margin is about to look extremely good. Kuehne+Nagel published third-quarter figures on 22 October, and the number travelling around the industry is the eighty-four per cent: Air Logistics EBIT up eighty-four per cent to CHF 169m. I read that line twice, because EBIT cannot grow eighty-four per cent against turnover that grew thirty-eight per cent unless the price of a kilo did the work.

Here is the pile as it stands. Group net turnover CHF 7.4bn, up 22 per cent. Group EBIT CHF 435m, up 53 per cent. Air Logistics net turnover CHF 2.4bn, up 38 per cent, EBIT CHF 169m, up 84 per cent, conversion rate 32 per cent, with volumes accelerating from 3 per cent growth in the second quarter to 10 per cent in the third on cloud and semiconductor demand. Sea Logistics net turnover CHF 2.7bn, up 20 per cent, EBIT CHF 172m, up 55 per cent, 3.3m TEU handled. Road Logistics EBIT up 40 per cent. Full-year recurring EBIT guidance lifted to CHF 1.5bn to 1.6bn.

What I want to do with those numbers is not admire them. It is locate them. Profit shaped like that is never earned evenly across the map; it is earned on a handful of lanes, on a particular kind of cargo, and it is paid for by a particular kind of shipper. So before we get anywhere near what anybody should do, let us walk the corridors: where the yield came from, which node jams first when cargo keeps moving this way, what it costs to step off the road, and the one line in that announcement that no headline has touched yet.

Begin with the arithmetic, because the arithmetic is the story. Air Logistics net turnover went up 38 per cent to CHF 2.4bn. Air volume went up 10 per cent. One divided by the other says net revenue per tonne went up roughly 25 per cent, and EBIT per tonne went up roughly 67 per cent. So the 84 per cent everybody is quoting is not a tonnage story at all. It is a yield story with some mix inside it, and the difference between those two matters to anyone who signs the invoice.

Mix means the average rose because high value-per-kilo cargo pushed low value-per-kilo cargo off the same deck, and not one shipper paid a cent more. Rate means everybody paid more per kilo. Kuehne+Nagel is not obliged to publish that split and did not. Both halves, though, point the same direction for the rest of us: the freight that wins a scarce main deck slot is the freight carrying the most value per kilo, and if your freight is not that freight, you are the one who gets moved to a later rotation.

Next, the 32 per cent. This is the figure most readers get wrong, and it deserves thirty seconds. The conversion rate is EBIT over gross profit, not EBIT over turnover. CHF 169m at 32 per cent implies gross profit of roughly CHF 528m, which is about 22 per cent of the CHF 2.4bn net turnover. On an air freight bill of one hundred francs, roughly twenty-two stay as gross profit and about seven become operating profit. The business is thin. It is also geared to yield in a way that should worry you rather than reassure you: that is how a 25 per cent move in revenue per tonne produces an 84 per cent move in EBIT.

Read the gearing from your side of the table and it says the holder of scarce space will protect rate by pulling capacity, because pulling capacity costs him a few tonnes of contribution while cutting rate shows up at once in a number his own guidance now commits him to. In a softening market, expect rolled bookings long before you expect discounts.

Why this quarter, and on which roads. Cloud is not a service, it is hardware: racks, switches, power shelves, accelerators, cooling units. Most of it is assembled in Taiwan, Korea, Malaysia, Vietnam and northern Mexico, and it flies because a datacenter that slips a month costs more per day than the freight bill. Semiconductors move out of Taipei, Hsinchu, Incheon, Shanghai and Singapore into Amsterdam, Frankfurt, Leipzig, Chicago, Dallas and Penang, and they fly because a stopped assembly line is worse than any air rate. Put that beside the sea side of the same report: turnover up 20 per cent on 3.3m TEU.

Treat 3.3m as a nine-month figure and the quarterly run rate is about 1.1m boxes, which means the sea growth came from price per box far more than from boxes. So the third quarter is not a picture of world trade growing 10 per cent. It is a picture of one narrow, very valuable slice of world trade deciding, inside about ninety days, that it would rather fly.

The acceleration is the part I would watch rather than the level. Volumes went from 3 per cent growth in the second quarter to 10 per cent in the third. A level tells you where the water is; a change in the rate of growth tells you which way it is moving, and it moves months before any rate index admits it. That jump did not come from a new factory opening in September. It came from planners moving existing cargo between modes inside a quarter, which is the fastest kind of demand shift there is, and also the most fragile, because anything that can be moved from sea to air in ninety days can be moved back in ninety days.

Who absorbs what. Sort shippers by value per kilo rather than by turnover and the damage becomes legible. At the top sit semiconductor equipment, datacenter hardware and pharmaceuticals: freight is one to three per cent of goods value, they hold annual block-space agreements, and a 25 per cent rate move is an annoyance that lands on a project budget. In the middle sit consumer electronics, automotive components and medical devices: freight runs five to ten per cent of goods value, delivery dates are contractual, and a 20 to 25 per cent air increase takes one to two points off gross margin unless the sales price is re-quoted. At the bottom sit apparel, homeware, furniture and hardware, where freight is a heavy share of goods value and the rate rise genuinely hurts, but the rate is not what actually finishes them.

What finishes them is that 300 kg of cushions does not compete with 3,000 kg of servers for the same main deck, so it gets rolled, and a rolled booking in November is a missed selling window rather than a cost increase. Size runs the same way: a shipper with block space and five tonnes a month is on the allocation list, and a shipper calling in with three hundred kilos is the buffer the list is built out of.

When it reaches you, in order. The yield move is already behind us; it happened in August and September and is now sitting in a quarterly report. What reaches you next is allocation rather than price. Rolling starts two to four weeks after the yield peak, which puts the worst of it in the back half of October and the first half of November. After that comes your own quoting rhythm: if you buy air on spot, re-price every seven days through November, because a rate held open for a month in this market is a one-way option you are handing to the seller for nothing.

Then comes the tender. First-quarter 2027 talks open in late November and run through December, and the other side will open them with the CHF 1.5bn to 1.6bn full-year guidance sitting in front of them as proof that the third quarter is the base case rather than a spike. Walk into that meeting armed with nothing but last year's average rate and you are negotiating against a number that was fixed before you sat down.

Here is what nobody put in a headline, and the reason I sat down to write this instead of forwarding the release. The company lifted full-year recurring EBIT guidance to CHF 1.5bn to 1.6bn on 22 October. Nobody lifts guidance in the fourth week of October on optimism. By then a group that size has most of the fourth quarter committed: block space and charter capacity signed, peak allocations pre-sold to the shippers who bought early, and whatever is left priced to clear. So the main deck you go shopping for in the last week of October is not an open market. It is the residue of a book somebody filled two or three weeks ago, and the person answering your email knows to within a few tonnes how much is left while you are guessing. That asymmetry is worth more to you than the 84 per cent, because the 84 per cent is about money already spent and the asymmetry is about money you are about to spend.

It also tells you to change the question. Most shippers answer a tight deck by pushing for a better rate, and rate is the one thing the holder of scarce space will never concede, because conceding it to you means conceding it to the whole market. Ask instead for a fixed weekly allocation in kilograms at the rate on the table, with a written penalty if the allocation is not honoured. A guaranteed 500 kg a week at a rate you think is high beats a rate you think is fair with no guarantee of moving at all, and in a quarter like this one the distance between those two outcomes is a week of production.

Now put the divisions next to each other. Group EBIT is CHF 435m. Air is CHF 169m and sea is CHF 172m, so those two trading divisions account for CHF 341m, or a bit over three quarters of the group, and the remaining CHF 94m or so covers road, contract logistics and everything else. On turnover the two are close in margin, about 7.0 per cent for air against 6.4 per cent for sea, but on growth they are not close at all: 84 per cent against 55 per cent. Capital follows growth, which means the chartering desk and the block-space budget of that company will be pointed at aircraft for the next two quarters. If you are a sea shipper who assumed your ocean contract was the relationship that mattered this year, this is the quarter that quietly demoted it.

Then there is the piece that catches people out because it does not appear in the air numbers at all. Road Logistics EBIT went up 40 per cent. If you fly into Frankfurt or Amsterdam and truck onward into Poland, Czechia or Hungary, that inland leg is repricing on top of the air leg, for reasons that have nothing to do with aircraft. Assume, conservatively, that the European road portion of your door-to-door air cost rises eight to twelve per cent off the back of that kind of margin pressure. The 25 per cent you derived on the air leg is therefore not your real increase; your door-to-door increase is that plus a road leg moving on its own logic. I have watched shippers fight hard for two days over an air rate and then wave through a trucking surcharge worth half of what they just won.

Where this breaks. Everything above rests on one assumption: that cloud and semiconductor capital spending keeps running at this pace into next year. If the hyperscalers trim capex guidance, the 10 per cent volume growth goes back toward 3 per cent within a quarter or two, and the yield comes off faster than it went on, because freighter capacity is contracted months ahead and does not evaporate when demand does. In that world the 84 per cent reverses just as violently: a 20 per cent fall in yield takes EBIT down by more than half, and the same people who were rationing space in October will be calling you in February.

The supply response is the other half of the same clock. Charters ordered on the back of an 84 per cent quarter take three to six months to show up as metal in the air, so the capacity relief that eventually breaks this yield is already being ordered and will land somewhere in the first half of next year. The signal to act on is not a press release. It is two consecutive weeks of falling Asia-origin air rates on the published indices together with charter enquiries drying up. When both appear, stop signing twelve-month agreements and go to quarterly blocks with a volume band, because a counterparty offering you a twelve-month fixed rate in a falling market is selling you his hedge, not his goodwill.

The smaller risk on the other side: if the deck stays full through November, the pre-Lunar-New-Year rush starts earlier than usual and the squeeze runs into the middle of January instead of stopping at Christmas. Anyone whose Q1 tender closes late should assume the first three weeks of January are priced like December.

Now the ledger, with every assumption written out so you can argue with it. Assume you ship semiconductor test equipment from Taipei to Frankfurt. Assume one consignment of 1,200 kg gross and 1,680 kg volumetric, so chargeable weight is 1,680 kg. Assume your current air rate is USD 4.20 per chargeable kilo, which puts the consignment at USD 7,056. Apply the 25 per cent per-tonne revenue move derived from the report and the rate becomes about USD 5.27, putting the consignment at USD 8,854, so USD 1,798 more each time and USD 21,576 more over a month of twelve consignments.

That is the hole. Here is part of a plug for it. Assume 400 kg of that consignment is accessory cartons, cables, brackets and paperwork, low value per kilo and not needed on day one. Move those 400 kg off the aircraft onto the Eurasian rail service via Xi'an or onto a sea-air routing, at an assumed USD 1.60 per kilo all-in: USD 640 instead of USD 2,108, which is USD 1,468 saved per consignment and USD 17,616 over twelve consignments. The price of that saving is time. Rail adds roughly eighteen to twenty-two days door to door.

Assume the goods in those cartons are worth USD 60 per kilo and your carrying cost is 12 per cent a year; holding 4,800 kg of accessories twenty extra days costs about USD 1,894 a month. Net of carrying cost, trimming the low-value tail saves roughly USD 15,700 a month against a hole of USD 21,576, and it leaves your exposure on the 1,280 kg that actually matters completely unchanged. It is not a fix. It is a third of a fix, and it is the third you can action this week.

If you are thinking about stepping off the air road entirely, price the exit properly. Rail from central China to Duisburg or Lodz runs roughly eighteen to twenty-two days against three to five days by air, and it will not take your lithium-containing lines or your temperature-controlled lines without a separate conversation and separate documentation. Sea-air via Jebel Ali or Colombo lands in the middle on both time and money and adds a transfer handling, which is exactly where high-value and temperature-controlled cargo gets damaged or goes missing. The mistake I see most often is switching mode without moving the order point: a planner moves cargo from air to rail, banks the freight saving, and then misses two delivery windows because procurement is still ordering on an air lead time. Move the order point first, then move the cargo. And never switch a SKU that carries a contractual delivery date without a written amendment beside it, because the freight saving is smaller than the penalty clause.

On contract shape, take the lesson from the conversion rate rather than from the headline. A business earning seven francs of operating profit on a hundred francs of billing will not sell you a twelve-month fixed rate while its own yield is rising, so asking for one wastes a meeting. What you can usually get is a six-month block with a collar: a floor and a cap set a fixed percentage either side of the rate you agree today, with the allocation guaranteed inside it. That structure gives the seller the yield protection his guidance commits him to and gives you the certainty of moving, which in a quarter like this one is worth more than the last five per cent of price.

Ravi would tell you to find the single variable that moves the whole system, and on this quarter that variable is yield per tonne rather than tonnage, which is a tidier way of saying the 38 per cent is not what produced the 84 per cent. For those of us who have to move goods rather than model them, the practical version is short. Work out which of your own SKUs are competing with server racks for the same deck, price those separately for the next eight weeks, and do not let them ride on a general air tariff somebody negotiated in June. The road is still open. It is tolled now, and the toll is highest for the cargo that did not need to be on it.

↑ Back to the key points

— By Ömer Kaya