Key takeaways
- China's Q1 2026 exports rose 11.9% to 6.85 trillion yuan; total goods trade rose 15% to 11.84 trillion yuan.
- Electric vehicle exports rose 77.5% year on year, lithium batteries 50.4%, and wind turbines and parts 45.2%.
- The US import price index for goods from China fell 1.41% year on year in March 2026, while the all-commodities import price index rose 2.26%.
- US core goods CPI rose just 1.07% in the year to May 2026, and has decelerated for three consecutive months.
- The China import price index bottomed at 96.9 in July 2025, 3.6% below its June 2024 level of 100.5, and has since recovered to 98.8.
Tariffs are levied on prices at the border. Disinflation arrives inside them.
The argument that a tariff wall stops imported disinflation assumes the disinflation arrives as a finished consumer good with a country-of-origin label on it. Much of it does not. It arrives as an intermediate input — a battery cell, an inverter, a motor, a cell module — bought by a manufacturer in a third country, assembled there, and shipped onward. The tariff schedule sees a Vietnamese or Mexican or Hungarian product. The cost curve sees a Chinese one.
China's Q1 2026 customs data show what is moving. Total goods trade rose 15% year on year to 11.84 trillion yuan, with exports up 11.9% to 6.85 trillion yuan and imports up 19.6% to 4.99 trillion yuan. Within that, the green-goods complex grew at multiples of the aggregate: electric vehicle exports up 77.5%, lithium battery exports up 50.4%, wind turbines and components up 45.2%.
The proof is in the US import price index, and it is measurable
The BLS publishes an import price index specifically for goods from China. It is the cleanest available reading of what Chinese supply is doing to US prices, because it strips out the exchange rate story and measures what US buyers actually pay.
That index stood at 100.5 in June 2024. It fell to 96.9 by July 2025 — a decline of 3.6% while overall US CPI rose. In the year to March 2026 it was still down 1.41%. Over the same month, the all-commodities import price index was up 2.26%, and by May it was up 6.74% on the year as the energy shock passed through.
The chart plots the year-on-year change of both series for March 2026, in percent: goods from China at -1.41%, all imported commodities at +2.26%. Two lines, same month, opposite signs. That gap is the disinflation channel, and it is 3.7 percentage points wide.
Where it lands in US inflation: core goods at 1.07%
US core goods CPI — commodities less food and energy — rose 1.07% in the year to May 2026, having decelerated from 1.20% in March and 1.14% in April. In index terms it fell outright between April (167.767) and May (167.575).
Set that against headline CPI of 4.17% and CPI energy of 22.97% over the same 12 months. The entire inflation problem is in energy. The goods complex — the part of the basket most exposed to Chinese supply — is running near 1% and falling, through a tariff regime designed to prevent exactly that.
The channel is narrowing, which is the part nobody says
Here is where the thesis needs discipline. The China import price index has stopped falling. It bottomed at 96.9 in July 2025 and has risen every month since: 97.2 in December, 97.5 in February 2026, 97.9 in March and April, 98.8 in May. Year on year it was -1.41% in March, -0.61% in April, and +1.13% in May.
The disinflation impulse from Chinese goods flipped positive in May 2026. Volume is still surging — 77.5% growth in EV exports is not a slowdown — but price is no longer falling. The channel that suppressed US core goods inflation through 2025 is, on the most recent print, closing.
What this is worth to a portfolio
The practical test is the goods-services spread inside CPI. Core goods at 1.07% against core PCE at 3.41% means services are carrying the entire inflation burden. A portfolio positioned for goods-led disinflation to keep suppressing headline CPI has, as of the May data, lost its engine.
The measurable threshold: if the China import price index sustains a positive year-on-year rate for three consecutive months, core goods CPI loses its principal downward force, and the 1.07% figure becomes a floor rather than a trend. One month is not three. May 2026 is the first.
What would falsify this and what the data cannot show
The import price index measures prices paid by US importers, which include tariff pass-through effects and exchange-rate moves. A rising index does not prove Chinese factory-gate prices are rising; it may reflect tariffs being absorbed differently, or a weaker dollar. The BLS series cannot distinguish those.
Nor does the customs data prove the transshipment argument. It shows Chinese export volume growing fast in exactly the categories that are intermediate inputs, and it shows US import prices from China falling while overall import prices rose. That is consistent with the mechanism, and it does not establish it. Bilateral trade data are required to trace a battery cell from Shenzhen through Hanoi to a US warehouse, and no single dataset publishes that path.
The series worth a standing alert
Not China's export headline — that will keep printing double-digit growth and will keep being read as a trade story. The number that carries the inflation signal is a BLS index that almost nobody quotes: import prices, all commodities, from China. It went positive in May 2026 for the first time since December 2022. If it stays positive, the cheapest input into the US goods basket has stopped getting cheaper, and the disinflation that tariffs failed to stop will have ended without them. and .
This content is for informational purposes only and does not constitute financial advice. Always do your own research or consult a qualified financial adviser before making investment decisions.