Asia-Pacific-US Air Rates Run 41% Above Last Year as Europe Volumes Drop 16%

The Analysis

Every corridor has a story and every corridor has a pothole, and this week the story is about two corridors draining into each other. Flexport's weekly air update, drawing on WorldACD market data, puts Asia Pacific to United States rates roughly forty-one percent above a year ago, with chargeable weight up ten percent week on week. Turn the map around and Europe is bleeding: Asia Pacific to Europe volumes ran sixteen percent below last year as the European Union's low-value parcel measures squeezed flows out of the air network, Hong Kong to Europe tonnage thirty-two percent below a year ago, and Hong Kong to United States tonnage up twenty-seven percent week on week. Korea to Europe tonnage managed to rise thirty-four percent week on week. Then comes the line every operator I know reads ahead of the rest of the page: on the United States lanes out of Taiwan, Vietnam and Cambodia, which are running full, book five to ten days ahead.

Those are not four separate stories. Read them as one machine with two outlets. An aircraft does not retire when its usual cargo stops booking; it is redeployed to wherever the freight pays. Tonnage that used to be Chinese low-value e-commerce into Liege and Amsterdam has been pulled off those lanes, and it is looking for anything with a higher yield, and the obvious place for it to go is the transpacific eastbound market that is already running hot. Anyone reading only their own lane this month is blind to half the mechanism driving their own rate.

I have walked most of this map, from the freight village at Incheon to the docks behind Ho Chi Minh and the cargo handlers in the old Hong Kong sheds. Where each parcel ended up and what it did to the promise the shopper was given is her map, and Nadia has that one drawn properly. My job here is the lane itself, so let me answer the three questions I get asked whenever a number like forty-one percent lands: which node is going to jam, what does it actually cost to move around it, and is this road still worth walking.

Start with the coupling, because it is the thing nobody put in a headline. Europe down and the United States up look like opposite stories and they are closer to the same story. Freighter capacity is mobile in a way that shipping lines envy and belly capacity is not mobile at all, and that difference explains most of what we are seeing. Belly slots follow passenger schedules, which are filed months in advance and do not much care that your e-commerce volume collapsed. Freighters follow yield. So when European e-commerce tonnage fell off, the part of the network that could move did move, and it moved toward the lane where the rate is forty-one percent higher than last year. That is why a shipper who only ever looks at their own lane is reading half a map. Your Hong Kong to Los Angeles rate is partly being set by a customs measure taken in Brussels.

A word of discipline before anyone acts on three-week-all-decisions numbers. Week-on-week readings taken across a holiday period are treacherous. The ten percent jump in chargeable weight and the twenty-seven percent jump out of Hong Kong to the United States both sit near a mainland holiday, and a jump measured against a week when half the factories were shut flatters itself badly. I do not throw those numbers out, but I insist on three readings before I believe a trend. Ask whoever sends you the weekly for the same week last year and for the four-week moving average on the same lane, and only then decide whether you are looking at a surge or at a calendar artefact. Anyone who has tried to book a freighter the week after a holiday knows how fast a paper trend evaporates once normal production resumes.

Now the map, node by node, because the node is where you lose money. The originating nodes that hurt most right now are Taipei, Ho Chi Minh and Phnom Penh. Those are the lanes Flexport flagged as running full, and there is a dull reason and an interesting reason. The dull reason is that transpacific eastbound capacity into the United States is genuinely tight against demand, and when a lane is tight the last ten percent of available space goes to whoever asked first. The interesting reason is that Vietnam and Cambodia have become the release valve for every shipper trying to be somewhere other than mainland China, which means everyone who diversified out of China in the last two years is now queuing at the same three airports with the same fortnight of cargo. Diversification did not remove congestion from your supply chain. It relocated it to airports with fewer runways, fewer ground handlers and worse road access, and it added the thumbprint costs of smaller consolidation batches.

Hong Kong is the valve on the whole system, and it is doing something strange. Tonnage to Europe down thirty-two percent year on year while tonnage to the United States rises twenty-seven percent week on week is not two cities. It is the same ramps, the same trucks queuing at the same cargo terminal gate, the same handlers building the same pallets, and somebody upstairs deciding what the building is going to run this week. For a shipper that means something specific: even though Europe has emptied out, you should not expect the terminal to be noticeably faster at handling your United States freight, because the capacity freed by Europe has been refilled. Free hidden capacity is a myth in this business. Space that shows up anywhere gets sold within a week.

Korea is the one that deserves a paragraph and a warning. Tonnage to Europe up thirty-four percent week on week while Hong Kong to Europe is down thirty-two percent year on year could mean two very different things. The first reading is that this is ordinary rerouting: Chinese-origin cargo trucked or barged across to Incheon because Incheon has the freighter frequencies into Europe that have gone thin elsewhere, plus a genuine electronics and automotive-parts season out of Korea itself. The second reading is less comfortable: that some of this is origin management, goods leaving through a Korean gateway because the paperwork that leaves through that gateway reads differently. I will not pretend to know the split from a weekly report.

But there is a cheap way to check, and it takes your own documents rather than somebody's opinion. Pull the place of receipt on the master air waybill and line it up against the actual shipper address on the commercial invoice. Pull the commodity lines and ask whether an economy of that shape plausibly produces them. If the receipts say Incheon and every invoice says Guangdong, the cargo is being relabelled rather than produced, and the lane you are relying on will be closed by an enforcement action rather than by market forces. Run that check before you build a plan on Korea.

Which node jams next, and when does it hit your ledger. In a tightening air market the sequence is boringly consistent, and it starts with time, not price. Booking lead time is the first thing to stretch, and it has stretched from a comfortable two or three days to the five to ten days Flexport is quoting. The price moving second is a mirage, because by the time you notice the price you have already lost the space. The third thing to go is rate validity: airlines shorten validity windows, quotes that used to hold for two weeks hold for seventy-two hours, and the habit of booking today and arguing about the number later stops working.

The fourth thing to go is the discount: the negotiated off-tariff you were quietly enjoying gets quietly withdrawn before anyone sends you a letter about it. Fifth comes roll protection: your booking gets pushed to the next departure without anybody offering compensation, and you learn about it from the tracking page rather than from your forwarder. Last comes the emergency charter, and if you have reached that node you have already paid for everyone else's poor planning. Skip harangues and count days: from today's booking decision to a cost consequence is roughly ten to fourteen days, which is one booking cycle plus one transit, and that is the shortest warning you will ever get.

Then the switching arithmetic, which is what my whole job amounts to. Assume for the rest of this section that these are my assumptions and not published figures: current all-in air from Taiwan or Vietnam to the United States of four dollars fifty a kilo,Incheon transshipment adding forty cents a kilo on top for inland haulage and transfer handling, and two extra days elapsed. Take a twenty tonne shipment with a value density of four hundred dollars a kilo, which is eight million dollars of cargo. Each day of inventory carrying at twenty percent annual is eight million times nought point two divided by three hundred and sixty-five, a little under four thousand four hundred dollars, which works out at about twenty-two cents per kilo per day.

If staying put costs you another five days waiting for space, that is roughly a dollar ten per kilo burned, plus whatever a missed promotion costs you. Routing through Korea costs forty cents plus two extra days of carrying, call it forty-four cents, so about eighty-four cents a kilo, sixteen thousand eight hundred dollars all in. On those assumptions moving pays, and it pays because value density is high enough that time hurts. Now run it again at sixty dollars a kilo, which is a lot of ordinary consumer goods. Carrying per kilo per day drops to about three cents. Waiting five days costs sixteen cents a kilo.

Korea costs forty cents plus six cents. Waiting wins, comfortably, and the clever move is to sit tight and let somebody with high-value cargo take the expensive routing. There is your rule, and it is a rule rather than an opinion: value per kilo times your financing rate sets the price of a day, and a day is the only unit these lanes actually sell you.

Let me put real monthly money on it, because a percentage is a costume that a number wears. Assume sixty tonnes a month out of Vietnam and Taiwan to the United States and an all-in air cost of four dollars fifty a kilo, which is my assumption rather than a published rate. Monthly spend is sixty thousand kilos times four fifty, two hundred and seventy thousand dollars. At forty-one percent above last year, last year's equivalent rate was four fifty divided by one point four one, three dollars nineteen, so the same sixty tonnes cost about one hundred and ninety-one thousand five hundred dollars then.

The gap is roughly seventy-eight thousand five hundred dollars a month, call it nine hundred and forty thousand dollars on an annual run rate. That is a real number to take into a pricing meeting, and it is the number your finance director will ask for. Then ask what it buys. Air gets you roughly twelve days door to door on those lanes on my assumption; an expedited ocean or sea-air routing gets something nearer twenty-six days at around a dollar twenty a kilo, my assumption again. The premium for air is three dollars thirty a kilo for fourteen days of the calendar.

Purely on inventory carrying that premium only justifies itself above roughly five hundred dollars of value per kilo, which almost nobody ships. The honest reason people still fly is not carrying cost, it is selling windows: a seasonal item that misses its six week window loses something like a quarter of its value, and at a hundred dollars a kilo that twenty-five dollars a kilo of markdown risk dwarfs a three dollar thirty freight premium. Know which of the two arguments you are making before you sign.

Europe needs its own paragraph and a warning rather than congratulations. Sixteen percent below last year and Hong Kong to Europe down thirty-two percent will bring brokers to you with cheap offers, and the offer is real. The question is what the schedule looks like behind it. Tonnage falling on a lane does not usually show up first as a lower rate. It shows up as frequency, and frequency is what actually wrecks your plan. Lane running seven departures a week gives you seven sets of closing times to aim at; cut it to three and your worst-case dwell before uplift roughly doubles, while your consolidation opportunity window closes on the days you need it.

Say your lane drops from daily to three a week and your average dwell grows by about one and a half days on ten tonnes at a hundred and fifty dollars a kilo: that is one and a half million dollars of cargo and roughly one thousand one hundred dollars a shipment of extra carrying, sixty thousand dollars a year if you ship weekly. Small change compared with a rate jump, but the real damage is not the carrying cost. The real damage is that your quoted booking window of three days is now fictional, and you discover it on the day a production line is waiting for the part.

Where might I be wrong, and what would flip this whole reading. If the forty-one percent is mostly holiday distortion and front-loading rather than genuine demand, this unwinds, and it unwinds fast, because air cargo has no shortage of idle belly capacity looking for work in the first quarter. In that world a shipper who panics and signs a twelve month commitment at the peak spends the following spring explaining a rate that is twenty to thirty percent above the market. If it is genuine demand, it does not unwind and the correct behaviour is to secure space rather than to chase price.

Three tests, all cheap. Watch whether the twenty-seven percent Hong Kong to United States jump survives a second week once factories are back at full output. Watch whether the Europe year-on-year gap narrows, because narrowing means capacity drifting back across the Pacific rim and thinner US space. And watch the booking guidance itself: when the recommended lead on those Taiwan and Vietnam lanes goes past ten days, you are no longer in a rate cycle, you are in an allocation cycle, and the response changes completely. Nobody rings a bell at that boundary, so watch the number rather than waiting for the call.

So what does a sane importer do with a Friday afternoon and this sheet of numbers. Take the five to ten day instruction literally today rather than filing it as background: it is not advice for someone else, it is the new minimum on the busiest lanes you use, and the cost of ignoring it is paid in days rather than in dollars. Then decide which of your shipments is allowed to wait and which cannot, using the value per kilo test above rather than anybody's intuition, because that test is the difference between paying forty cents extra and paying five days of delay. Open one second gateway before you need it, and pay for one genuine trial through Korea so that you have a measurable alternative, not a theoretical one.

On Europe, take the cheap rate and simultaneously demand the schedule in writing, with a maximum dwell commitment, because cheap capacity without predictable departures is an expensive product wearing a discount. And keep the two halves of this map in the same picture. The readiest mistake this month is treating the weak European lane and the strong American lane as unrelated news items. They are two ends of the same pipe, and the day Europe fills back up is the day your five day booking window on the transpacific becomes a ten day booking window.

One more thing I learned the expensive way on these lanes, and it costs nothing to ask. Demand the load factor on the specific flights you are booking rather than accepting a lane-level answer, and ask which of those flights is belly and which is freighter. A lane average hides everything that matters: the departure you need may be ninety percent sold while the lane as a whole sits at sixty, and belly space disappears on a passenger schedule change that nobody warns the cargo desk about in advance. If your forwarder cannot produce a per-flight answer within twenty-four hours, that itself is the answer. You are buying space from somebody three steps removed from the metal, and on lanes asking five to ten days of lead time, distance from the metal is the most expensive distance there is.

Every corridor has a story and every corridor has a pothole. Right now the Pacific lane is charging forty-one percent more than last year and selling days rather than kilos, the European lane is cheap because it is shorter of cargo than of discipline, and Hong Kong is standing in the middle of both like a ferry terminal deciding which bank it prefers this week. Decide your own lane policy before somebody else's decides it for you.

↑ Back to the key points

— By Ömer Kaya