Supply Chain Action Points
Read this first — the conclusion, and the moves to make:
- By July 31, convert every booking for the next four weeks into a named flight with an aircraft type and a ULD type, and treat any booking that cannot name the flight as unconfirmed.
- Ask for an all-in rate per kilo for those four weeks; if the quote comes back as a base rate plus surcharges, budget at $6.00, because at eight tonnes a week the annual gap between the two ends of the band is about $416,000, which is our assumed volume and not the article figure.
- Before August 15, move every SKU that can wait three weeks to the water, and keep air only for the ones where a stockout costs more than the $1.00 per kilo gap.
- For the air you keep, tender firm daily tonnage ten days before cutoff so you sit inside the buildup plan instead of on the waitlist.
- Put a capacity review in the last week of October, when carriers re-cut schedules, and if direct capacity is still down double digits, carry $6.00 into the first-quarter plan.
Shanghai (PVG) to Australia spot air freight rates reached $5.00-$6.00/kg in mid-July 2026, touching $6.00/kg with steady upward momentum. PVG-Melbourne direct capacity dropped 19.1% in early July, driving a roughly 20% price surge; Qantas' new converted freighter launching July 26 is expected to bring only minor short-term relief.
The Analysis
Six dollars a kilo is not a price. It is a statement about how many pallet positions exist between Shanghai and Melbourne on a given Tuesday.
In mid-July the Shanghai to Australia spot sat at $5.00 to $6.00 a kilo and touched the top of that band with upward momentum. The cause is not mysterious: direct PVG to Melbourne capacity fell 19.1% in early July, and the price came up roughly 20% behind it. Qantas puts a newly converted freighter on the lane on July 26, and the read across the market is minor short-term relief.
Read that last clause twice, because it is where most planning goes wrong. Relief and capacity are not the same word, and the gap between them is the whole risk on this lane.
Start with the ratio. Nineteen point one percent of the direct capacity goes away, and about twenty percent arrives on the price. That is close to one for one, and a lane that moves one for one has no buffer in it anywhere. When there is slack, losing a fifth of the uplift costs you a couple of points, because the market absorbs it somewhere you never see. When there is none, the loss passes straight through to the rate. So the useful reading is not that air freight got expensive. It is that this lane was already running with nothing in reserve, and one schedule decision was enough to show it.
The other number worth staring at is the band itself, five to six dollars. That is a twenty percent spread inside a single week's quote, which means the price depends less on the cargo than on the hour you called. A market that quotes a range that wide is telling you it does not know where the next pallet position is coming from.
Now the physical part, because this is where the relief story falls apart. A nineteen percent cut in direct capacity on a passenger-dominated lane is a belly-hold event. Those positions live in the lower deck of widebody passengers, they are booked against a specific flight, a specific aircraft type and a specific cutoff, and they are measured in ULD positions, not in tonnes of goodwill. A converted freighter adds main-deck pallets. Different ULD footprint, different loader, different buildup window, and a per-flight tonnage that is a fraction of what a widebody passenger schedule contributes across a week. You cannot subtract one from the other and call it a wash. The July 26 aircraft adds positions on its own rotation. It does not put back the specific uplift that disappeared when the passenger schedule changed, and it does not arrive on the flights your cargo is already booked against.
The cargo does not vanish either, which is the part that costs money quietly. When direct uplift thins, freight does not stop moving; it reroutes through a hub, and every reroute adds a handover. A handover means a second buildup, a second chance to miss a connection, a second set of labels, and a recovery problem that lands on whoever owns the delivery promise. On the invoice that shows up as a day, not as a dollar, and a day is what breaks a retail launch. The rate is the visible cost. The handover is the one that shows up in a customer complaint three weeks later.
Who this lands on. A parcel consolidator with volume and a weekly block will pay the six dollars and pass it through, because its customers are buying speed and have already priced it. Perishables and seafood will pay anything, because the alternative is a write-off, which is why they never really leave the lane. Industrial spares and machine parts can wait a week, and a chunk of them will quietly slide to the water this month, which is exactly why the price has not gone further than twenty percent. A forwarder holding blocked space is protected for the moment and exposed at renewal. A small shipper buying spot on a Tuesday is the one who pays the top of the band, and the one whose booking gets rolled when a bigger account calls.
On timing, July 26 is a date on a calendar, not a fix. It will add a rotation and it will take a little heat out of the spot print for a fortnight, and then the fourth quarter arrives, which is when this lane gets tested for real. Watch the direct capacity number weekly rather than the rate, because the rate is the symptom and capacity is the cause, and capacity moves about two to three weeks before the price does. If direct capacity is still down double digits when the carriers re-cut schedules at the end of October, then six dollars is not a spike, it is the new floor for the first quarter, and everything you promised your sales team was built on a number that no longer exists.
Run a small account with my assumptions on the table. You fly eight tonnes a week from PVG to Australia, which is a modest program, and you budgeted at five dollars because that is roughly what June cost. The difference between the two ends of July's band is one dollar a kilo, so eight thousand kilos a week is eight thousand dollars a week, or about four hundred and sixteen thousand dollars across a year. That is a program-level number, not a rounding error, and it is the number your finance team will ask about in September. The eight tonnes is mine, not the article's; the dollar spread is not.
The traps here are all about mistaking a schedule for a supply. Booking a rate without a flight number is booking an intention. Budgeting at the bottom of the band because the top looks like a bad week is how a good quarter turns into a miss. And moving volume to the water without counting the inventory it takes to bridge the gap is trading a visible freight cost for an invisible working capital cost, which is usually the worse of the two trades.
So do one thing today. Ask your forwarder for the flight number, the aircraft type and the ULD type on every booking you hold for the next four weeks. If they cannot name the flight, you do not have space, you have a rate. On a lane running one for one, that distinction is the entire difference between a plan and a hope.
↑ Back to the key points