Air cargo tonnages ease 1% in week 39 while worldwide rates climb 2% to $2.37 per kilo

Begin with the apparent contradiction, because if you get this wrong you will make the wrong call on price. Volume fell 1 percent while rates rose 2 percent. The explanation is that these two numbers are measuring different populations of cargo. The 1 percent is the total tonnage on the market, and it is falling because the last few days of September are when Chinese factories wind down and new production stops. The 2 percent is the average rate on the cargo that is actually moving, and the cargo that is actually moving in the last week before Golden Week is the cargo that has to move, which is the cargo that pays whatever the rate is. Volume is being pulled down by the factory shutdown, and the rate is being pulled up by the composition of what remains. Both numbers are true and they are not in conflict.

What that means for you is specific. If your cargo is ready now and must fly, you are in the expensive population, and the price you get this week is a pre-holiday price, not the market price. If you can wait, you are in a position to buy a cheaper week later. That is the whole decision, and it is a calendar decision more than a market decision.

Put numbers on it. Assume you move 20 tonnes a month by air from Shanghai to Europe. At 2.37 dollars per kilo, that is 2,370 dollars per tonne, so 47,400 dollars a month in freight. A 2 percent move up is about 950 dollars on your monthly bill, and a 3 percent two week move is around 1,420 dollars. That is not a catastrophe, and I want to be honest that for most shippers the rate move this week is a rounding error compared to the booking and cut off mistakes that cost ten times more. A missed uplift before Golden Week means a week of delay, and a week of delay on air cargo defeats the entire purpose of paying for air.

Now the part that matters more than the rate, which is capacity. The two week number shows capacity flat, and September shows capacity up 10 percent year on year, with Asia Pacific outbound capacity up around 30 percent. That is a lot of belly and freighter capacity chasing the market. When capacity is that loose, you should not be paying a premium rate for guaranteed space, because the space is not actually scarce. What is scarce in the Golden Week window is space on the specific dates you want, out of the specific airport you want, with the specific cut off you can actually meet. That is a scheduling scarcity, not a market scarcity, and it is bought with planning rather than with money.

So the specific thing I would do this week is book early and book flexible. Instead of paying a premium for a fixed uplift, I would book two or three alternative dates and accept whatever the forwarder can firm up, because with capacity flat and rates only up 2 percent, the forwarder has room to move. The mistake people make is treating a pre-holiday week like a peak season week and overpaying for certainty they could have gotten for free with two days of planning.

Then there is the lane mix, which is where I think the real opportunity sits. The strongest lane increase was Middle East and South Asia to Asia Pacific at 12 percent, followed by Asia Pacific to Europe at 7 percent. Meanwhile North America to Asia Pacific fell 4 percent. If you buy air freight in both directions, or if you have flows that can be routed differently, there is an obvious arbitrage. Rates going one way are up 7 to 12 percent and rates going the other way are down 4 percent. If you have any flexibility on the return leg, or on shifting volume onto lanes that are not the hot ones, this is the week to look at it.

Let me make that concrete. Assume you ship 10 tonnes a month Asia Pacific to Europe at 2.37 dollars per kilo, which is 23,700 dollars. A 7 percent increase is about 1,660 dollars a month, call it 20,000 dollars a year. If you can move even a third of that volume onto an India or Southeast Asia origin, where the lanes have not moved the same way, you might recover several hundred dollars a month. I want to flag the assumption clearly, which is that this only works if your sourcing or your customer allows an origin shift. If it does not, the lane arbitrage is not available to you and you should ignore it rather than chase it.

On the destination side, the 4 percent fall from North America to Asia Pacific tells you something useful about the return leg. If you have anything moving from the US back to Asia, whether that is returns, samples or components for rework, this is a cheaper moment than the last quarter, and I would not assume that persists once the Golden Week disruption clears and volumes reset.

Now the number I would actually structure a contract around, which is the 2.37 dollars per kilo. It is 31 percent below last year and 38 percent above September 2019. Read those two comparisons together and you get the real picture. Rates are not high in historical terms, they are normal, and they are low relative to the last year. That means this is not a market where you should be signing a long term contract at a high fixed rate to protect against a spike, because the spike is not the base case. If anything, the risk in the next two quarters sits on the downside, because capacity is up 10 percent and September volume was only down 2 percent, which is close to flat. When capacity growth outruns volume growth, the pressure on rates is down, not up.

So I would refuse any annual contract that prices in a permanent premium, and I would push for a shorter commitment with a rate review at a fixed date. Assume you have 500 tonnes a year and you sign at a fixed rate 5 percent above the current market to protect against a spike that does not come. That is around 59,000 dollars a year of extra freight on a base of roughly 1.185 million dollars, and it is money you cannot get back when the market moves your way. I would rather take the market rate with a quarterly review than buy insurance against a spike in a market with 10 percent extra capacity.

There is also a seasonal pattern I would plan around rather than react to, and it is the holiday calendar. Golden Week takes Chinese capacity and Chinese production offline, and the weeks after it typically see a bulge of cargo that was held back, which then competes for the first available uplifts. So the quiet week is often not the week after the holiday, it is the week of the holiday itself. If your cargo can move during the holiday window, you may find space easier and rates softer on the specific dates the market is not watching. Most shippers never test this because their warehouse is also closed, and the ones who plan around it get an advantage that costs them nothing.

The trucking and warehouse end of that deserves a mention, because a holiday is not only a flight problem. Your export warehouse closes, your trucking capacity thins out, and the handlers at the airport run reduced shifts. So even if you find a flight with space during the holiday week, you need to have arranged the pick up and the export clearance before the holiday starts, which means the cut off for a holiday week shipment is actually a week earlier than the flight date suggests. Shippers who miss that find out that a flight they secured is irrelevant because nothing can reach it. I would map my holiday week back from the flight date and work out when the last truck and the last declaration have to happen, and then decide whether the holiday week is actually available to me.

And there is a commercial use for a week like this that I would not overlook. When capacity is loose and rates are only nudging up 2 percent, your forwarder has room, and a forwarder with room is a forwarder you can ask for things you normally cannot get. Better free storage at origin, a longer acceptance window, a firm commitment on the uplift rather than a best effort, and a rate review clause that works both ways. None of these show up on a rate sheet and all of them are worth real money over a year. I would use this week to ask for them, and I would ask for them in the booking note rather than in a phone call, because the answer you can point to later is the answer that matters. In a market where volume is flat and capacity is long, the forwarder needs your tonnes more than you need their space, and that is the week to test it.

Let me also say something about what I would not do, because the temptation this week is to lock everything. I would not sign a peak season surcharge agreement on air cargo for October on the strength of a 2 percent weekly move. The two week move was 3 percent with flat capacity, and September capacity was up 10 percent. Those are not the ingredients of a rate spike. The ingredients of a rate spike are falling capacity and rising volume, and we have the opposite of both. The rate move this week is a pre-holiday composition effect, and composition effects reverse. If you lock a surcharge for three months based on one week, you are locking the reversal in with it.

I would also be careful about the difference between the index rate and your rate. The 2.37 dollars per kilo is an average across a global mix of lanes, commodities and service levels. Your rate depends on your lane, your density, your volume, your contract terms and how much handling your cargo needs. A 2 percent move in the index tells you which direction the wind is blowing, it does not tell you what your rate should be. I have seen shippers quote the index at a contract meeting and be told, politely, that they are not buying the index, they are buying their lane. Use the index to time the conversation, not to set the number.

On the operational side, and this is the boring part that saves the most money, I would treat the Golden Week window as a documentation and cut off problem rather than a rate problem. Air cargo cut offs tighten before a holiday because warehouse and handling staff are thinner, and the flights you booked two weeks ago may have a different cut off than the one in your system. Confirm with your forwarder the actual cut off for each of your bookings, in writing, and build in a buffer of a day. A shipment that misses the last pre-holiday uplift does not just arrive late, it arrives after the holiday, and for time sensitive cargo that is the difference between a sale and a write off.

And I would watch the fuel side quietly, because it is the one input that can turn a soft market hard. Jet fuel is the big cost line under air freight, and when fuel moves sharply the carriers pass it through regardless of how loose capacity is. I do not have a view on where fuel goes from here, and I would not bet a rate strategy on it. What I would do is make sure my contract has a fuel mechanism I can actually read, with a defined index and a defined adjustment frequency, so that a fuel move is a small known cost rather than an argument after the invoice arrives.

On the cargo side, I would also use a week like this to look at density, because it is the single biggest lever a shipper controls over an air rate. Air freight is priced on chargeable weight, which is the greater of actual weight and volumetric weight, so a low density cargo pays for air it is not using. If you can consolidate loads or change packaging so a pallet carries more kilos in the same footprint, you can cut your chargeable weight without touching the rate at all. Assume your cargo has a volumetric ratio where 40 percent of your chargeable weight is volumetric rather than actual, and you shift your packaging to cut that by a quarter. On 20 tonnes of monthly chargeable weight at 2.37 dollars, that is around 4,740 dollars a month saved, roughly 57,000 dollars a year, and it does not require a single negotiation. Most shippers never look at density because it is not in the forwarder's interest to raise it, and it is the highest return hour you can spend in a soft rate market.

There is also a service level question worth revisiting, because in a week where rates are only moving 2 percent, the money is not in the rate, it is in what you buy with it. If you have been paying for a guaranteed uplift on cargo that has two days of slack in it, this is a good week to step down a service tier and keep the saving. Conversely, if you have cargo where a one day delay genuinely costs you a customer, that is the cargo to keep on the guaranteed product regardless of the rate. I have seen shippers buy the same service level for everything, and the ones who split their cargo into must fly and can wait categories consistently spend less for the same reliability on the part that matters.

One more thing on the capacity number, because 10 percent capacity growth with a 2 percent volume decline is a wide gap and it has to go somewhere. It goes into load factors, and when load factors fall the carriers start cancelling freighter frequencies or parking aircraft on the soft lanes. That is the mechanism by which a soft market turns, and it usually starts on the routes where the capacity came from, which in this cycle is Asia Pacific outbound. So I would keep an eye on whether the Asia Pacific freighter frequencies start getting trimmed in October and November, because that is the early warning that the loose capacity underpinning the 2.37 dollar rate is going away. If frequencies hold, the soft market holds and I keep buying on the market. If they get cut, I move to contracts earlier than I otherwise would have.

So here is the plan I would run. This week, book my must fly cargo early and book it across two alternative dates rather than paying a premium for one fixed uplift, and accept that I am buying flexibility rather than certainty. I would look at my lane mix and see whether any volume can move onto the lanes that have not risen, keeping the assumption in front of me that this only works if the origin can actually change. I would pull my air contracts out and check the fuel clause and any peak surcharge language, and I would refuse to lock a spike premium that the capacity data does not support. I would confirm every cut off for the next two weeks in writing and add a day of buffer. And I would put a review date in the diary for the second week of October, after the holiday bulge has cleared, because that is when I will know whether this was a calendar blip or the start of something firmer. If by then the rate is back below 2.37 and volume has recovered, I keep buying on the market. If the rate is holding above 2.37 with volume coming back, that is a different market and I will change my posture then, with data rather than with a feeling.

• Book must fly cargo this week across two alternative uplift dates rather than paying a premium for one fixed date, because capacity is flat and the rate move is only 2 percent.
• Check your lane mix against the data: Middle East and South Asia to Asia Pacific rose 12 percent and Asia Pacific to Europe 7 percent while North America to Asia Pacific fell 4 percent, so shift volume only where the origin can genuinely change.
• Refuse to lock a peak season surcharge or a fixed rate premium for October; capacity was up 10 percent in September with volume down only 2 percent, which is not the setup for a spike.
• Confirm the actual cut off for every booking in the next two weeks in writing and add one day of buffer, because a missed pre-holiday uplift means arrival after the holiday.
• Read your air contract fuel clause and require a defined index and adjustment frequency so a fuel move is a known cost rather than a dispute.
• Set a review date in the second week of October and change posture only if the rate holds above 2.37 dollars per kilo while volume recovers.