China Adds 15 International Air Cargo Routes in September, 130 So Far This Year

The Analysis

Zhengzhou at four in the morning has a particular kind of quiet. The apron lights are on, the nose of a freighter is open, and a line of box trucks that rolled out of the industrial parks around Xinzheng an hour earlier is waiting to be called forward. I have stood on that apron more times than I can count, and every single time my first question is not what the freight costs. My first question is where the truck came from and how long it sat at the gate. Freight in my line of work is a road before it is a rate, and the road is made of kilometres, gate queues and the mood of a customs officer on a night shift.

That is the lens I am carrying into the number that landed on 11 October. The China Federation of Logistics and Purchasing said China opened 15 new international air cargo routes in September, worth 42 weekly round trips. Nine of those routes were intra-Asia, four went to Europe, two went to North America. Nine airlines are operating the new capacity and eight of them are Chinese carriers. The cargo filling the new lanes is cross-border e-commerce, fresh and cold-chain produce and agricultural goods, electromechanical and electronic products, high-end manufacturing goods, and daily consumer items. As of 30 September, 2026 had produced 130 new routes and 334 weekly round trips.

Every write-up I read that morning counted the 130. Nobody drew the map. So let me draw it, because 15 routes on a press release and 15 routes on a map are two completely different animals. On the map, the first thing you see is not how many. It is where, and where tells you what kind of year this has been for anyone moving cargo out of China.

Start with the plain slice of fact, the part nobody argues about. Fifteen routes opened in September. Forty-two weekly round trips came with them. Nine stayed inside Asia, four crossed to Europe, two crossed to North America. Nine carriers are flying them, eight of them Chinese. The cargo mix is cross-border e-commerce, fresh and cold-chain agricultural produce, electromechanical and electronic goods, high-end manufacturing output and ordinary consumer goods. By 30 September the year had banked 130 routes and 334 weekly round trips. That is the filing. Everything after this is my reading of it, and I will keep telling you which is which.

Now put those fifteen lines on a map and the shape changes. Six of them leave the Asian landmass. Nine do not. Whatever the year looks like in the totals, September was a month of regional stitching, not of transoceanic expansion. A route from Zhengzhou to Frankfurt is a different business from a route from Chengdu to Kuala Lumpur, and the difference is not prestige. It is what the lane is for. The intra-Asia nine are feeding a manufacturing and consumption loop that already runs at high frequency: components out, finished goods partway out, e-commerce parcels straight to a sorting centre. The six long ones are a bet on peak season, and a thin one at that.

Forty-two weekly round trips spread across fifteen routes averages out at under three per route. That average is the most useful number in the whole filing and it appeared in none of the summaries. Three round trips a week means you do not have a daily departure. You have departures on certain days, and if your cargo is ready on the wrong day it waits. Waiting is not free. It shows up as inventory, as a missed promotion window, as a cold-chain pallet sitting somewhere it should not be sitting. When a lane runs daily, a planner stops thinking about the calendar. When a lane runs three times a week, the calendar becomes the whole job.

The intra-Asia concentration makes sense once you look at what is actually in the belly. Cross-border e-commerce is a parcel business, and parcels want short legs and dense frequency because the customer on the other end is counting days on a phone screen. Fresh and cold-chain produce wants the same thing for a harder reason: a longer leg is a longer temperature risk, and temperature risk does not degrade gracefully. Electromechanical and high-end manufacturing goods move inside Asia because they are going into another factory, not onto a retail shelf, and the next factory is four hours of flying away, not fourteen. Put those three cargo types together and you get a network that thickens inside the region and only nibbles at the long-haul edges.

Europe got four and North America got two. If you are an exporter whose book is weighted to the Atlantic and to Western Europe, that is the line worth circling. Four lanes is not a flood, but four lanes in a single month, on top of whatever has been added since January, is a real change in the options available to you in a peak season that has already started. The catch is the one I keep coming back to: frequency. Two new North America routes do not give you a daily pipe out of China. They give you two more doors, and doors are only useful if you know which days they open and if your truck can reach them before they close.

Eight of the nine carriers are Chinese. That line deserves more attention than it got. When new capacity is mostly home-carrier capacity, the people you negotiate with change, and the way capacity is committed changes with them. Chinese carriers expanding all-cargo operations tend to build around their own hubs and around the industrial policy of the provinces those hubs sit in. A route out of Ezhou is not an accident of scheduling. A route out of Zhengzhou is not an accident either. These are airports that were planned as cargo airports, with land for sorting facilities and road access designed for trucks rather than for private cars. When capacity lands at a purpose-built cargo airport, the ground side usually works better than it does at a passenger airport where freight is the tenant that gets whatever space is left.

Which brings me to the thing the coverage skipped, and I am going to say it plainly so there is no mistaking it for a footnote: the new routes are overwhelmingly second-tier point-to-point lanes, and what decides whether they actually work is not the aircraft at all. It is the trucking in the hinterland behind the departure airport and the customs clearance at the port where the goods land. A route being announced is not the same thing as cargo flowing smoothly, and the gap between those two states is measured in kilometres of inland haulage and hours of queue at a gate. Headlines counted 130 routes. Nobody asked how many of those 130 have a bonded trucking fleet, a cold room and a night-shift customs desk sitting behind them on day one. That is the part we are adding here, and for an exporter it is worth more than the count.

I have a rule from years of walking these aprons. Do not underestimate a port. Not the famous one, the small one. The small one is where your cargo goes to lose a day. I remember a consignment of electronics out of a second-tier airport in the centre of the country where everything about the flight was perfect: the pallets were built on time, the aircraft left on time, the arrival was on time, and the shipment still missed its delivery window, because the receiving country's inland clearance point had one X-ray machine and a backlog of three hundred entries. The air leg was eleven hours. The ground leg ate two days. Nobody on the invoice called that a failure. Every planner on the receiving end did.

That story scales, and it scales in a specific direction. When capacity moves to a second-tier airport, the air leg often gets shorter and cleaner, because there is less congestion and the freighter is the airport's reason for existing. The ground legs on both ends get longer and messier, because the hinterland road network and the customs establishment were sized for the old volume, not the new one. An exporter in Zhengzhou gains maybe a thousand kilometres of trucking avoided by not sending goods overland to Shanghai. A buyer in Central Europe gains or loses depending on where the aircraft lands and how good the road from that airport is. Both ends have to be looked at, and looking at only one end is how people conclude that a new route "did not work" when what actually happened is that one end of the pair was never built for it.

Layer the impact and it splits cleanly by cargo type. For a cross-border e-commerce seller, a new regional lane is close to pure upside, because parcels are light, the volume is predictable week to week, and the alternative is often a hub that is already saturated. Sellers who can commit volume early will get space and will get it at a commitment rather than at a spot rate. For fresh and cold-chain shippers, the calculus is harsher. A new lane is only real if the cold chain is unbroken at both ends, and cold-chain breakage happens at transfer points, not in the air. A shorter air leg with an extra handling step is worse than a longer direct one. For electromechanical and high-end manufacturing shippers, the question is loading plan and damage rate. New stations, new ground handlers, new pallet-build standards. The first six weeks on any new lane are when your packaging gets tested by people who have not handled your cargo before.

For freight forwarders the new capacity is an opportunity with teeth. Blocked-space agreements on brand-new lanes are how forwarders make margin in peak season, and they are also how forwarders lose money, because a blocked-space commitment is a commitment whether or not your customers fill it. Eight domestic carriers chasing utilisation on fifteen new lanes is a buyers' market for the first two months and a sellers' market the moment peak season tightens. If you are going to sign anything, sign it while the lanes are young and sign it with an exit. For factories sitting in the hinterland of a second-tier airport the meaning is simpler and bigger: someone just moved the departure point several hundred kilometres closer to your loading dock, and the value of that is not in the freight rate, it is in the trucking invoice and in the hours your finished goods spend on a road instead of in a warehouse.

Timing is where this filing turns into a decision. A route announcement is not capacity you can book today in most cases. What follows an announcement is a negotiation phase, and in my experience that phase runs somewhere between one and four weeks before the first meaningful commercial commitment exists. After first flight, it takes roughly another four to eight weeks before a lane has settled into a rhythm the ground side can trust. Hinterland trucking and customs capacity, which is the slowest piece, typically needs two to three months to catch up, because you cannot hire night-shift inspectors or build a cold room on the cycle time of a press release.

This filing covers September and was published on 11 October. Peak season is already running. If you count from the middle of October to the middle of December, you have roughly eight or nine weeks of usable peak window left. Anything that needs three months to mature will not help you this year. That is not a reason to ignore it. It is a reason to treat this year as reconnaissance and next year as the payoff, and to commit this year only to the lanes that are already flying regularly.

Let me put numbers on it, because a claim about hours and money that never shows its arithmetic is just an opinion wearing a tie. Take one shipment, 800 kilograms of electromechanical goods, built on a pallet, ready at a factory outside Zhengzhou on a Tuesday morning, going to Frankfurt. Compare two paths. Path one is a new direct freighter lane out of Zhengzhou. Path two is the traditional move: truck the pallet to Shanghai Pudong and fly from there.

Assumptions, all of them mine, all of them stated so you can replace them with your own: the Zhengzhou factory is 60 kilometres from the airport and the truck takes 2 hours including gate time. Receiving, screening and building at Zhengzhou takes 6 hours. The direct lane departs three times a week, so the average wait for the next departure is 28 hours, and if you miss one you wait another 48 hours; call it a 20-hour average schedule penalty on a bad week. Trucking 60 kilometres inside the province costs 600 yuan. Local handling in Zhengzhou is 1.2 yuan per kilogram, so 960 yuan. Air leg and customs clearance at origin are treated as identical in both paths, because for this comparison they are: I am isolating the ground side and the frequency penalty, which is exactly the part a new route changes. I am deliberately not putting a rate on the air leg, because that is not what this piece is about.

Path two: trucking Zhengzhou to Shanghai Pudong is 950 kilometres and, with driver rest, roughly 13 hours. Pudong receiving, screening, waiting for the consolidated flight and the usual congestion at a saturated hub costs 18 hours. Waiting for a connecting long-haul slot at a hub that has many departures but also many customers is, call it, 10 hours. Transshipment handling at Pudong is 2.5 yuan per kilogram, so 2,000 yuan. Storage and dwell while waiting is 400 yuan. The truck is 5,800 yuan.

Path one ground cost, using my assumptions: 600 plus 960 equals 1,560 yuan, plus a schedule penalty that I will value at 800 yuan a day in working capital and missed-window cost, so a 20-hour wait is roughly 700 yuan. Call it 2,260 yuan all in on the ground side. Path two ground cost: 5,800 plus 2,000 plus 400 equals 8,200 yuan, and the hub path has no meaningful frequency penalty because Pudong flies every day, so call it 8,200 yuan. The difference is about 5,940 yuan per shipment and about 28 hours of elapsed time, which is a day and change.

Run four shipments a month and you are near 23,800 yuan a month, a bit under 285,000 yuan across a year, on one lane, one commodity, one customer. Those are my assumptions, not market quotes, and the reason I show them is to show the shape of the calculation, not the answer. The shape is this: the money is in the trucking and in the dwell, not in the air.

Now the part that turns the arithmetic into a decision, and it is the thing nobody outside our trade says out loud. Those savings assume the direct lane departs when it says it will and that the ground side in Zhengzhou and in Frankfurt can absorb your pallet without a two-day queue. On a lane that is six weeks old, neither assumption is safe. So the money is real but it is conditional, and the condition is the hinterland. Which is why the exporters who will actually eat from this in peak season are not the ones who read the press release and wait for their forwarder to offer the new lane. They are the ones who locked space on the new lanes before the lanes became fashionable, and who asked, at the same time, what the bonded trucking fleet looks like at the departure airport and how many inspectors are on the night shift at the arrival point.

There is a contrary reading and it deserves a hearing. A colleague of mine argues that none of this matters for a serious exporter, because the mature hubs have density, and density beats novelty. On her reading, Pudong or Hong Kong will still win for anything that has to arrive on a specific day, because a daily departure with five fallbacks beats a three-times-weekly departure with one, and because the ground handlers at a mature hub have seen your cargo before. She is right more often than I like to admit. If your cargo is high value, time-definite and unforgiving, a new lane is a risk you take with a small share of your volume, not with the whole book. If your cargo is regular, price-sensitive and forecastable, the new lane is where the money is.

There is also the possibility that this whole wave thins out. New lanes announced in a strong peak season are sometimes quietly suspended in the first quarter that follows, when e-commerce volumes normalise and the utilisation maths stops working. Fifteen routes in one month is a fast pace, and fast-paced additions have a habit of including a few that were announced for reasons other than sustained demand. If you are going to build a plan on a new lane, ask how long the carrier has committed to it and what happens to your blocked space if the lane pauses. Get the answer in writing before peak season, not after.

So does the road hold. Yes, but not the way the headline suggests. The 130 figure is not a promise. It is a map of where capacity has been added, and on that map the thick new growth is regional and the thin new growth is long-haul. What you do with it depends entirely on which end of that sentence your business sits on. If you ship parcels and components inside Asia, this is the best capacity news of the year and you should be having conversations this month. If you ship to Europe and North America, you have gained two more doors and four more, and the correct response is not celebration. It is a calendar, a trucking quote and a question about who is on the night shift at the other end.

This line I have walked, in one form or another, for most of my working life, and the lesson it keeps teaching me is boring and reliable. Aircraft are easy. Roads and gates are hard. Anyone can add a route. Not everyone can add a road behind it. The exporters who understand that difference will take the 15 routes and turn them into something that shows up on a delivery report. The ones who only read the count will spend the peak season wondering why a flight that left on time still delivered late. Put simply, it is a road problem.

↑ Back to the key points

— By Ömer Kaya