Supply Chain Action Points
Read this first — the conclusion, and the moves to make:
- Lock trans Pacific north bound air capacity now at the 3.40 dollars per kg band before the next AI order wave reprices the lane.
- Split Northeast and Southeast Asia origin volume across carriers so a single lane peak does not roll your delivery.
- Track AI and semiconductor order flow weekly; if it slows, trigger a rate review because the 62 percent load factor has room to fall.
- Protect e commerce margin separately, since weak volume still pays a tech driven rate; raise prices or consolidate rather than absorb.
Xeneta's July report shows global air cargo demand rose 7% year-on-year in June, with spot rates at USD 3.40/kg, up 38% YoY. AI-related hardware and semiconductors offset weaker e-commerce; Northeast Asia–North America rates rose 41% and Southeast Asia–North America 42% versus late February. Dynamic load factor reached 62%.
The Analysis
AI and semiconductors carried June. Xeneta's July report shows global air cargo demand up 7 percent year on year in June, spot at 3.40 dollars per kg, up 38 percent year on year. AI hardware and chips offset weaker e commerce, and Northeast Asia to North America rates rose 41 percent and Southeast Asia to North America 42 percent versus late February. Dynamic load factor hit 62 percent.
For a shipper the story is not the average, it is the trans Pacific north bound lane, where the number is closer to 41 percent and the plane is full.
Five numbers are the fact slice. Global demand up 7 percent year on year in June. Spot at 3.40 dollars per kg, up 38 percent year on year. Northeast Asia to North America up 41 percent versus late February. Southeast Asia to North America up 42 percent. Dynamic load factor 62 percent. The global 38 percent hides the lane numbers near 41 and 42, which is where your boxes sit.
Cause chain is a mix shift, not a demand boom overall. AI hardware and semiconductors lifted the total while e commerce stayed weak, so the growth is concentrated in high value tech on the trans Pacific. The push comes from the AI build out, which fills north bound freighters first. E commerce softness is the offset, not the driver, so the rate rests on tech demand alone.
Layered impact splits by origin and cargo. Northeast and Southeast Asia to North America shippers pay the 41 and 42 percent lane premium and feel the tightest capacity. Other lanes track the 38 percent global average and hurt less. E commerce sellers are the odd ones out: their volume is soft yet they still pay a tech driven rate, so their margin is squeezed from both sides. Small shippers without block space get rolled most.
Transmission timing is already in the quote and likely holds through Q3. The 41 and 42 percent moves are versus late February, so the climb has been steady for months and shows no break. The lead time to protect cost is to lock capacity now, because a new AI order wave reprices the lane within weeks.
Overlooked point is the base. The 38 percent is year on year, not month on month, so the rate did not just jump, it has been high all year and the lane premium is the real story. The second miss is the offset risk: because e commerce is weak, if AI and semi demand eases, there is no second engine to hold the rate, and a 62 percent load factor has room to fall, so rates could drop faster than they rose. Shippers who read the 7 percent demand growth as safe miss that it is tech only and tech can pause.
Counterfactual flips on tech demand. If AI and semiconductor orders slow, the trans Pacific north bound rate falls toward the global 38 and then below, because the load factor has slack. If the build out accelerates, the 41 to 42 percent lane premium pushes higher. The conclusion holds only while AI and chip demand keeps outrunning e commerce.
Make it concrete with stated assumptions. Assume a 12 tonne Northeast Asia to North America shipment at 3.40 dollars per kg spot, that is 40,800 dollars. Versus late February assume 2.41 dollars per kg, that is 28,920 dollars. The 41 percent lane move is 11,880 dollars more on one shipment. At 30 shipments a month that is 356,400 dollars of extra freight versus February, on the same boxes, with e commerce too weak to share the load.
My own read is to lock. Hold trans Pacific north bound capacity now at the 3.40 band, watch the AI order flow weekly, and do not treat the 7 percent growth as a cushion, because it is tech only.
Another point the 41 and 42 percent lane numbers hide is the booking queue behind them. The rate is high because the space is gone, and a shipper without block space is not just paying more, they are waiting, which on a tech launch can mean missing the window entirely. The rate is the symptom, the queue is the disease. A second buried fact is the e commerce cross subsidy: because e commerce is weak, the lane is carried by tech alone, so if tech pauses the rate can fall faster than it rose, and a shipper who locked capacity at the top may sit on paid space the market no longer needs. The third angle is origin optionality: Northeast Asia and Southeast Asia price differently, and a shipper who can shift origin between the two gains a second quote on the same destination, which is the only real hedge when one lane runs hot. Map both origins before you commit the quarter.
For the capacity lead, the lesson of the 41 and 42 percent is to own space before the rate moves. Lock a block on the trans Pacific north bound lane now, even a partial one, because the queue behind the rate only grows as AI orders pile in. The second move is to map both Northeast and Southeast Asia origins so you can shift between them when one runs hot, which is the only real hedge on a concentrated lane. The third move is to protect the e commerce margin separately, because weak volume still pays a tech driven rate and there is no second engine to soften it. A shipper who pre books and keeps origin optionality enters the quarter with a plan instead of a panic.
↑ Back to the key points