DHL Air Freight Update: Global Demand Up 4% Year to Date, Rates 48% Above Prior Year

The Analysis

The first thing I do when an air freight update lands is find the line it moves in my budget, not the line it moves in the market. This one, DHL Global Forwarding's May 2026 air freight update, came with three numbers worth keeping: global volumes up 4% year to date, April demand up 5% year on year, and spot rates 48% above prior-year levels. Everyone will quote the 48%. I would not walk into a negotiation holding it.

I write from the buyer's side of the table, so let me say the obvious part out loud. Air freight is the line nobody in procurement wants to defend. It is small in weight and brutal in cost per kilo, it gets bought in a hurry, and it is where a supplier hides the shipment that is already late. When that line is running 48% above last year, and demand is still growing faster than capacity, the question is not whether air is expensive. It has already been expensive for a year. The question is whether this is a spike I can wait out, or a floor I have to write into a contract.

Those are two very different problems, and I think the popular read of this update confuses them. Everything in the numbers points closer to a floor than a spike. I will build that case from what DHL actually published, not from the mood of the room.

Start with the shapes rather than the totals. Volumes up 4% year to date. April on its own up 5% year on year. Capacity up 2% year on year in May, month to date, led by Asia. Spot rates flat on the month but 48% above prior year. Put those four together and one relationship jumps out before anything else: demand is growing faster than capacity. Four to five against two. And note which number is bigger, the monthly 5% or the year-to-date 4%. The monthly one. A market that is rolling over does not print a monthly figure above its own running average. That single comparison does more work than the headline, because it tells me the direction of travel, and the direction is still up.

There is a second shape in the numbers that the headline flattens. Rates are stable but elevated. Stable is doing a lot of work in that sentence. It means the level did not jump this month; the level was set some time ago and is holding. A spike does not hold flat for a month at 48% above prior year. A floor does. If I am buying, that distinction changes everything about how I negotiate, and I will come back to it.

The way this update is built matters too, because the figures are global and your exposure is not. A global volume number of 4% and a global spot number of 48% are averages across lanes, and an average hides exactly the two things a buyer needs. It does not tell you whether the 4% is spread evenly or concentrated on the lanes you actually use, and it does not tell you whether the 2% of capacity growth landed where the demand grew. The one clue in the release is that Asia led the capacity growth, which is reassuring if your cargo originates in Asia and less so if it does not.

So when someone quotes the global 48% at you, the honest reply is to ask for the lane and the commodity, because the average is a mood and the lane is a bill. I have sat through enough tender rounds to know that a supplier will quote the global number and hope you never ask for it to be broken down. Ask anyway.

Why now, and why this high? Three pressures are stacked on the same lane, and only one of them is cyclical. One is the Middle East geopolitical tension that is still disrupting the Asia-Europe corridors. When a corridor bends, cargo does not simply switch off; it often flies a longer path, or shifts to air from sea, and either way the same shipment consumes more capacity measured in tonne-kilometres even when the box count has not changed. Capacity is not a count of shipments, it is weight times distance, and distance just got longer across a big chunk of the world's trade. Then there are the aircraft delivery delays, which mean the capacity the market was promised has not arrived.

A carrier cannot fly a plane that has not left the factory, and no amount of demand can conjure one. And on top of both sits the expansion of the EU Emissions Trading System, which puts a carbon cost on the carrier's balance sheet that it pays whether demand is strong or weak. Count how many of those three are demand. One. The other two are structural, and structural costs do not disappear when the cycle turns. That is my first reason to distrust anyone calling 48% a demand spike.

Who actually pays, and how unevenly? Air freight is a sliver of the world's weight and a giant of its value, so the 48% lands hardest on whoever ships high-value, low-weight, time-critical goods. Consumer electronics and semiconductors, where the value per kilo justifies the premium. Fashion chasing a season, where a two-week delay turns full-price stock into markdown stock. Pharmaceuticals and anything with a shelf life or a cold chain. And the e-commerce returns that come back by air because no customer will wait three weeks for a refund.

A company that moves almost everything by sea and keeps air for genuine emergencies is barely touched in absolute terms, except that the lane it keeps for the worst week just repriced, which is the cruellest place for a repricing to land. On the other side of the ledger, a mid-size supplier with no mode choice at all, because the product is perishable or the end customer will not accept a sea transit, has exactly one lever left, and it is price.

Big buyers can split a lane across carriers and modes and negotiate the blend; small ones eat the number. That asymmetry is not a detail. It is the whole reason a 48% spot print can coexist with a soft market everywhere else, and it is the first place a procurement team should look when it decides who is really exposed.

How fast does the 48% reach my cost line? Slower than the headline implies, and the gap between the two is where the money lives. Spot moves today. A contract that reprices annually moves at renewal. An air contract that is index-linked moves only as far as the index formula allows, and no further, whatever the spot market does. So two buyers on the same lane, same cargo, same month, can be paying very different rates, and the difference is not luck. It is contract design. The transmission from a spot number to a budget runs on the contract calendar, not the news cycle. If your air book is annual, you are insulated for months and then hit all at once, at a level you can no longer influence. If it is monthly, or all spot, you are already paying the 48% and the only question is how long you keep doing it. That timing gap is not a nuisance. It is the negotiating window, and it is closing.

There is a second consequence of the word stable that I should draw out, because it is the trap for anyone renewing this quarter. If the spot market is flat at a level 48% above last year, the carrier's opening position at renewal is not last year's rate plus a little. It is last year's rate plus the whole gap, because the carrier can point at a flat, sustained market and call it the new normal. A spike invites a conversation about how fast it will fade. A floor does not, and that is why the adjective matters more than the percentage.

If you walk into a renewal arguing about the size of this year's increase, you have already accepted the carrier's frame. The argument worth having is about the level and its durability, and the way to win it is to separate the part of the 48% that is structural, the ETS and the delivery delays, from the part that is cyclical, the demand. Concede the structural floor, then make the carrier defend the rest line by line.

Now the part I think the update buries, and it is the part that decides whether you lock or wait. The headline says rates are 48% above prior year, and the natural read is that this is a demand story. Demand is cyclical, cyclical things mean-revert, so the instinct is to sit still and wait for the number to come back. I think that instinct is half wrong, and here is the half that matters. Two of the three drivers are not cyclical at all. Aircraft delivery delays and the EU ETS expansion do not vanish when demand cools.

They are a floor. Which means the 48% is not symmetrical. It can fall, but a real chunk of it will not fall back to where it was, because the floor underneath it has risen while nobody was looking. If you are waiting for last year's rate to return, you may be waiting for a price that no longer exists, and every quarter you wait, the floor gets a little more permanent.

The second buried point is more practical. Go back to the four numbers. Demand 4 to 5%, capacity 2%. The gap is 2 to 3 points. A 2-to-3-point gap does not, by itself, manufacture a 48% spot premium. What produces 48% is a thin market at the top, where the last few points of capacity are auctioned to whoever has to move this week. And that top slice is exactly where a procurement team's unplanned shipments live: the emergency restock, the missed deadline, the sample that had to be there yesterday. So the number that should concern a buyer is not 48%.

It is this: what share of my air volume is unplanned, and what do I pay for that share. A buyer who plans 90% of their air volume and panics on 10% is in a different business from a buyer who plans 50% and improvises the rest, even if both are on the same lane at the same rate. Measure the unplanned share before you argue about the rate, because the rate is not the variable you control. The share is.

Let me put real arithmetic on this, and I will flag every assumption so you can drop in your own numbers. Assume your air freight spend last year was 100 units, whatever currency you keep your budget in, and your air volume was 100 units as well. This year you ship 4% more tonnage, so 104 units. If your average rate is up 48%, your spend is 104 times 1.48, which is 153.9. That is 53.9% more money on the same product line, and notice how the two numbers compound: 4% more cargo, 48% higher rate, and you pay both at once. That is the fully exposed case, and it is the one a supplier will quote you if you have no contract cover at all.

Now the more realistic split. Assume 40% of your air volume rides on spot and reprices to the full 48%, while 60% sits on annual contracts that, at renewal, capture only half of the move, so 24%. Your blended rate is 0.4 times 1.48 plus 0.6 times 1.24, which is 0.592 plus 0.744, which is 1.336. Spend is 104 times 1.336, which is 138.9, or about 38.9% more than last year. The distance between the fully exposed 53.9% and the covered 38.9% is 15 points of your air spend. That 15 points is not a market fact. It is the size of the gap you are negotiating over, and it is the only dial on this page that is entirely yours.

One more, because it is the lever most buyers never pull. Assume sea or rail runs at about one-fifth of your air cost per kilo, which is a reasonable planning figure for a non-urgent SKU. Take 15% of your 104 units, about 15.6 units, and move it from an air rate of 1.48 to a sea or rail equivalent of 0.296. The saving per unit is 1.184, and 15.6 units times that is roughly 18.5 units, close to 18% of your entire prior-year air budget. And the decision that produced it was not a freight negotiation at all. It was a lead-time decision, made twelve weeks earlier, about which SKUs are allowed to be slow.

Before I get to what to do, the other side of the coin, because I refuse to hand you a one-way bet. The entire case for locking rests on that floor holding, and three things could break it. If Middle East tension eases and the Asia-Europe corridors reopen, the extra tonne-kilometres come out of the system and effective capacity grows without a single new aircraft. If aircraft deliveries catch up, the promised capacity finally lands, and 2% could read 5% or 6% next quarter. If demand rolls over, the April 5% turns into a negative number, and a 48% premium has nothing left to stand on.

Any one of those, and the buyer who locked the whole volume at today's elevated level is the one who overpaid. So the boundary is plain: lock the floor, do not lock the ceiling. Fix the part of the volume that is structurally urgent, and leave the optional part index-linked or on spot, so that a capacity recovery works for you instead of against you. If someone tells you the answer is to lock everything, ask them which of those three risks they are willing to put on your balance sheet. I know which ones I am not.

Let me be blunt about the mode shift, because it is the one lever I said a buyer fully controls, and it is also the one most often done badly. Moving a SKU from air to sea or rail does not change the freight line alone; it changes when the goods land, and everything downstream of a landing date is a cost. Assume a fashion SKU with a twelve-week season. Pulling it off air onto a six-week sea transit can turn a full-price sale into a markdown if the transit eats the front of the season, and no saving on the per-kilo rate recovers that.

Assume a spare part that keeps a production line running, and the same shift turns a freight saving into a line stoppage. So the mode decision is not a freight decision at all; it is a decision about which SKUs can absorb time and which cannot. Do that categorisation first, with product and sales in the room, and the freight saving falls out of it. Skip it, and you will save 18% on the rate and lose multiples of it on the sales floor.

The traps here are specific, and they are where money quietly leaks. The most common one is treating air as a single lane. It is not one product; it is a stack of service levels, express and deferred and everything between, and a 48% market move will not hit them equally. If you renegotiate on one blended average, you will overpay on the express slice and win nothing on the deferred slice, and you will not see either until the invoice is a year old. Another is mode substitution that counts only the freight line.

Shifting a SKU to sea or rail is cheap on the rate and expensive everywhere else if the product is seasonal or the customer contract carries a delivery date and a penalty. Do the arithmetic on the landing date, not the per-kilo rate, or you will save 18% on freight and lose 30% on markdowns. Then there is index and currency mismatch. If your air contract is index-linked in a currency different from the one your revenue arrives in, a headline 48% is not the number you pay. And the last one is the emergency lane itself: the shipments you did not plan are the ones now priced at the top of a thin market, and no contract touches them unless you wrote a clause that does. When the Middle East shifts the corridors and you start re-planning routings, that is a question about the road, not the rate. Ömer Kaya has walked those corridors and I have not, and I would take his read before I move a single Asia-Europe shipment.

So here is how I would sign this, if it were my budget and my name on the line. I would go to my air carriers in the next two weeks, well before renewal, and ask for one thing: a fixed floor on the base rate for the structurally urgent slice of my volume, with everything else index-linked and capped. I would target 60% coverage on the firm slice, because the demand-versus-capacity gap is what sets the premium and I want to be covered while the gap is still open, and I would keep 40% floating so that a delivery-driven recovery lands in my favour. And I would take at least 15% of my tonnage out of the air by design, not by panic, moving the non-urgent SKUs to sea or rail with a two-week buffer. That is the part I control. The 48% I cannot control. The share of my volume that pays it, I can.

↑ Back to the key points

— By Grace Yang