China rail freight holds at 83.0m t while highway trucks fall 8.4% to 55.3m

The weekly transport read for 21 to 27 September is a study in contrast, and I want to walk through every line before drawing conclusions, because the headline numbers lie if you only read the top one. Railway freight came in at 83.026 million tonnes, dipping only 0.39 percent from the prior week, which in plain terms means the rail network ran almost flat through the holiday build-up and the factory shutdowns that come with it. Highway truck trips, on the other hand, fell 8.43 percent to 55.325 million, a much sharper drop that tells you the flexible, on-call capacity left the system. Monitored port throughput slipped 1.36 percent to 278.781 million tonnes, but the container side actually rose 0.74 percent to 7.333 million TEU, so the box flow held even as bulk softened. Civil aviation logged 12,600 freight flights and 5,396 tonnes of cargo, up 2.93 percent week on week. Express pickups dropped 9.95 percent to 3.72 billion pieces. Put those together and you get a clear picture: the system did not seize up, certain modes did the heavy lifting while others stepped back for the holiday, and the split is exactly where your risk lives this week.

Let me be honest about what this means on the ground, because I have lived through enough Golden Weeks to know the pattern. The Mid-Autumn festival and the run-up to Golden Week are when a lot of factories in the Pearl River and Yangtze River deltas shut for a week or two, and when provincial transport bureaus tighten truck movement with extra inspections and route restrictions. That is exactly why road trips fell harder than rail. Rail runs on a fixed timetable and national allocation, so a factory pause barely dents it; the trains still run because they carry coal, grain, and the freight that cannot wait. Trucks are the flexible, last-minute mode, and when drivers head home and controls tighten, the count drops. As an importer or exporter, the practical read is that your rail bookings are the stable part of the plan this week, and your truck-dependent legs are the fragile part, and confusing the two is how you miss a sailing.

The container number is the one I would not gloss over, because it is the counter-intuitive bit. Port throughput overall dipped, yet containers still grew 0.74 percent to 7.333 million TEU. That tells me export box flow stayed resilient even while bulk commodities cooled, which is a decent signal that finished-goods shipping out of China did not stall the way the headline throughput drop might suggest. If you are an exporter with boxes booked, your slots are probably fine. If you are an importer waiting on inbound containers, the yard side is still moving. The place you feel pain is the road feed, where the 8.43 percent trip drop means fewer trucks available to dray boxes to and from the ports and inland hubs. So the ship gets there, the box is fine, but the truck that was supposed to collect it from the terminal might not show in the numbers you are used to.

Now let me put a number on it, with the assumption stated up front so nobody thinks I am quoting a rate sheet. Say you move 40 forty-foot containers, and a FEU is simply one forty-foot equivalent unit, the standard 40-foot box, a month from an inland city like Chengdu to the coastal export port. Your rail block-train allocation sits at 83 million tonnes of national rail freight holding nearly flat, so your train slots are dependable this week. But the drayage and the cross-province road feeder are where the 8.43 percent trip drop bites. Assume your normal road leg costs 800 dollars per container on the feeder, and assume the tightened truck supply plus holiday controls push that leg up by 15 percent during the window. That is 120 dollars extra per container, times 40 containers, or 4,800 dollars a month of avoidable cost if you stay on spot road instead of shifting volume to rail. None of those rates are published figures; they are illustrative assumptions so you can see the shape of the problem, not a quote, and your own lane will show a different absolute number but the same direction.

The aviation side is quietly the bright spot, and I would be doing you a disservice not to flag it. Freight flights rose 2.93 percent week on week to 12,600, carrying 5,396 tonnes. For high-value or time-critical shipments, air kept climbing even as road softened, which means if you have an urgent order that cannot wait for the post-holiday road recovery, the air option is there and actually a touch more available than last week. The express pickup drop of 9.95 percent to 3.72 billion pieces is the flip side: cross-border and domestic parcel volumes slowed, so if your model is small-parcel e-commerce that rides the express networks, build in a few extra days of transit buffer because the sortation and last-mile pace is lighter right now. I have seen teams panic about air when the real story was that road was thin and air was fine; do not make that mistake this week.

Begin by auditing where your own shipments sit on the rail-versus-road split this week, because you cannot manage what you have not mapped. I would pull the last four weeks of your lane data and mark which legs are rail-locked and which are truck-dependent, because the answer changes your risk profile completely. The rail legs are safe; the truck legs are where the 8.43 percent drop lives. Hand that map to whoever books your transport by Wednesday, because the longer you wait the more the spot truck market prices in the holiday tightness, and by the second week of October everyone is chasing the same recovered capacity and the rates snap back hard.

Next, shift what you can from road to rail for the next two to three weeks, and I mean the predictable volume, not the unpredictable emergencies. I am not saying abandon trucks, I am saying move the schedulable volume onto the block trains you already have allocations for, and reserve trucks for the genuine last-minute gaps. A freight forwarder I work with last year lost three days on a Shanghai-bound dray because he assumed trucks would be as plentiful in early October as in August; they were not, and the demurrage ate the savings. Lock your rail space this week for departures through 20 October, and confirm the truck feeder only for the volume you truly cannot rail, because every container you push to rail is one less container fighting for a thin truck.

After that, renegotiate or at least flag the road rate with your carriers, because silence is how you get surprised on the invoice. With trips down 8.43 percent, the carriers who stayed on the road are sitting on tighter capacity and will pass it along. Ask for a written rate hold through the Golden Week window, even if it is only a soft commitment, and get a fallback quote from a second trucker so you are not hostage to one. The 800-dollar-per-container example above is exactly the kind of uplift you want to pre-empt, not discover on the invoice when the shipment is already late and you have no leverage.

And watch your port free-time clock during the holiday like it is the only clock that matters, because in a thin-truck week it is. Containers still moved, up 0.74 percent, but terminals run lean crews over the break, so a box that arrives and sits because your truck could not collect it will start racking up demurrage and detention faster than usual. I would tell your destination team to pre-clear customs and line up the pickup truck before the vessel berths, not after, because the road side is the bottleneck, not the ship side, and a box cleared and waiting for a booked truck is a box that does not cost you money.

A useful alternative if your lane is purely road today is to open a rail option you have never used, and I would seriously look at it even if it sounds like a hassle. The 83.026 million tonne rail figure shows the network has slack to absorb more, and many inland-to-port corridors sell block-train slots that beat truck rates once you fill a train. The catch is lead time: rail booking usually needs three to five days of notice versus a truck's same-day, so you trade flexibility for cost and reliability. For steady monthly volume like the 40-FEU example, the trade is almost always worth it this week, and even if you only move half your volume to rail you have halved your exposure to the truck squeeze.

Another angle worth your attention is inventory positioning. Because road is thin and rail is steady, the smart play for many importers is to pull forward the stock you know you will need in the back half of October and let the rail network carry it while it is reliable, rather than waiting for the post-holiday road scramble. I would identify the top SKUs by steady demand, book them on rail now, and let the warehouse absorb the early arrival rather than risk a stockout when road recovers slowly. This is the kind of move that costs nothing extra if your rail slot is already there, and it buys you insurance against the second-week crunch.

Do not forget to talk to your customers about timing, because a missed delivery that was foreseen is a footnote and one that was not is a complaint. I would send a short note to any customer expecting China-origin goods in the first two weeks of October explaining that road capacity is unusually tight and offering a revised ETA, so the surprise lands on you gently rather than on them harshly. The factories are closed, the trucks are home, and a honest date now beats a missed date later, and your customer relationships are worth more than the awkwardness of a revised quote.

Set a monitoring cadence so this does not become a one-week fire and then forget. I would check the ministry weekly rail and road figures every Monday morning, track your own truck-availability from carriers, and re-run the lane map the moment the numbers move more than a couple of points. The 0.39 percent rail dip and the 8.43 percent road drop are this week's readings; next week they will shift, and the teams that watch the shift are the ones that book before the curve, not after it.

The pitfall I see most teams walk into is treating the whole week as one flat holiday lull, and that mistake is expensive. It is not. Rail is steady, road is thin, containers are up, express is down, air is up. If you plan all five the same way you will over-pay on road and under-use rail and air. Another trap is assuming the dip is demand. It is not, it is timing and controls, which means the volumes come roaring back in the second week of October and the road market snaps tight again, so booking ahead is the only way to avoid paying the snap-back rate.

Let me close with the worry that actually keeps me up, because it is the one you should carry into the week. The 8.43 percent road drop is a clean signal that the flexible capacity left the system, and the 0.39 percent rail dip says the backbone held. Going into Golden Week, the importers and exporters who planned around that split are fine; the ones who assumed trucks would be there on call are the ones eating demurrage and missing sailings. I would rather you pre-book rail today and have it as insurance than scramble for a truck on 8 October when everyone else is doing the same, because on that day the truck you need will be the truck everyone else is also calling for.

One more practical move is to get your carriers to commit to a weekly capacity number in writing rather than a per-shipment quote. When the road side is this thin, a trucker who owes you ten trucks a week on a standing commitment is worth more than a cheap spot rate you cannot actually get. I would ask each of your primary road carriers for a guaranteed weekly allocation through 20 October, even at a small premium, because a guaranteed ten is better than a quoted fifty you never see. The 8.43 percent drop means the aggregate supply is gone, so the fight is for a slice of a smaller pie, and a written slice beats a hopeful glance at the spot board.

For exporters specifically, the container number is your friend this week. With boxes up 0.74 percent to 7.333 million TEU, your export booking is unlikely to be the constraint; the road feed to the port is. I would stop worrying about whether your sailing exists and start worrying about whether your truck will make the cut-off, because a missed cut-off turns a fine booking into a rolled shipment, and a rolled shipment in a thin-truck week can cost you the whole sailing. The fix is boring but effective: book the truck before you book the optimism, and confirm the cut-off with the terminal in writing.

And for importers bringing finished goods in, the air option deserves a second look even if you normally ship ocean. Air freight flights rose 2.93 percent week on week, so capacity actually loosened on the premium lane while the cheap lane tightened, which is the rare moment when upgrading does not mean fighting for space. If a delayed ocean-to-road delivery would cost you a stockout or a penalty, the marginal cost of air on a few critical SKUs may be lower than the cost of the miss, and this week the air seats are there to take. I would identify the ten SKUs where a week of delay hurts most and pre-clear them for air if the road picture does not clear.

The bigger lesson under all of this is that China transport system is not one system, it is several, and they do not move together. Rail held, road fell, containers rose, express fell, air rose. Treating the country as a single logistics climate is how you get surprised; reading the mode split is how you stay ahead. I have built my own weekly habit of checking the ministry rail and road figures the moment they post, because the 0.39 and the 8.43 are the early warning that tells me which of my clients needs to shift before the rates move, and the rates always move after the volumes, never before.

If you want a single thing to do tomorrow, make it the lane map. Pull the last four weeks, split rail from road, flag the truck-dependent legs, and hand it to the person who books. Everything else in this note follows from that one sheet, and a sheet you build today is worth more than a perfect plan you start next week, because the trucks that left the road this week are not coming back until the factories reopen and the controls lift, and by then the rates will have moved past you.

I will say it once more because it is the part people forget: the dip is timing and controls, not demand. Volumes will return, road will snap tight, and the importers who booked rail and pre-arranged trucks this week will be the ones shipping normally while everyone else queues. The window to act is the quiet part of the curve, before the rebound, and that window is open right now.

There is also a supplier-angle most teams miss. If your Chinese factory shut for the holiday and your road feed is thin, the bottleneck is not the factory, it is the hand-off after it. I would call the factory this week and ask exactly when they resume and whether they can hold finished goods a few extra days so you can batch them onto a single rail departure rather than trickle them onto scarce trucks. A factory willing to stage is a factory that saves you the truck premium, and the ask is free if you make it before they restart and get busy.

• Audit your lane mix by Wednesday: separate rail-locked legs from truck-dependent legs using the last four weeks of data, owned by your transport booking lead.
• Shift predictable volume from road to rail and lock block-train space this week for departures through 20 October to dodge the 8.43 percent road-trip squeeze.
• Get a written road rate hold through the Golden Week window from your primary trucker plus a fallback quote from a second carrier to cap the ~15 percent feeder uplift.
• Pre-clear customs and line up the pickup truck before the vessel berths so containers do not sit and rack up demurrage while terminals run lean holiday crews.
• For time-critical or high-value orders use the air option, which rose 2.93 percent week on week to 12,600 flights, instead of waiting on post-holiday road recovery.
• Add three to four days of transit buffer to any express or small-parcel e-commerce flow because express pickups fell 9.95 percent to 3.72 billion pieces.
• Pull forward top-SKU stock onto rail now so it arrives before the second-week October road snap-back rather than risking a stockout.
• Send customers a revised early-October ETA note this week citing tight road capacity so a foreseen delay stays a footnote, not a complaint.