Consensus is reading the wrong number entirely. China's July inflation data is not a statement about Chinese demand. It is a statement about crude oil, and the market has filed it in the wrong drawer.
The facts first. National Bureau of Statistics data released on Sunday showed consumer prices up 0.5 percent year on year in July, down from 1.0 percent in June and the softest reading since January. Producer prices rose 3.5 percent against 4.1 percent the month before, undershooting the roughly 3.8 percent economists had pencilled in. Bloomberg framed it as the oil shock from the Iran war beginning to ease, here.
Now the line items almost nobody quoted. The PPI for oil extraction fell 11.8 percent month on month. Refining fell 8.4 percent. Headline PPI dropped 0.7 percent sequentially after a 0.3 percent fall in June. Core CPI, stripping food and energy, rose 0.3 percent.
Two numbers are doing all the work in that release, and both of them are barrels. The core reading barely moved. This is not an economy that suddenly stopped buying. This is an economy that stopped paying a war premium for energy.
The mechanism is arithmetic, and it is only halfway through
Brent traded above $126 a barrel at the end of April. By early July it was back near $70, below where it sat the day before the war began, and more than a third off the peak. Front month contracts were trading under the following month, the classic signature of a market that has stopped worrying about the next cargo.
China is the largest crude importer on the planet. When the barrel price collapses, it does not enter Chinese inflation data as a slow, demand-driven grind. It enters at the top of the pipe, in extraction and refining, and it moves fast. That is exactly what the 11.8 percent and 8.4 percent sequential drops are. They are the war premium leaving the system.
From there the transmission is mechanical. Refined product costs fall into chemicals, plastics, packaging, freight and every industrial input downstream. Producer prices roll over sequentially. Two to three quarters later, that shows up in the price of what leaves Chinese ports. The old reflex, drilled into every macro desk since roughly 2004, is to treat falling Chinese PPI as a leading indicator of falling US core goods inflation.
The same arithmetic is loaded into the US data. American headline CPI already fell month on month in June on an energy swoon, with core at 2.6 percent. From late this year into the first half of 2027, every headline print in every major economy will be compared against a base built during a war spike. The negative energy contribution will be enormous. It will look like disinflation. It will not be disinflation. It will be a calendar.
Prices falling because a war premium expired is not the same event as prices falling because nobody wants the goods. Markets are about to price them as if they were identical.
The second-order effect: the pipeline is carrying different cargo
Here is the part that is not being discussed anywhere I can see, and it is the reason the old reflex is now dangerous.
China's export mix has changed beyond recognition in eighteen months. July exports rose 23.9 percent year on year to $397.85 billion. High-tech shipments were up 40.7 percent. Semiconductor exports have roughly doubled in value terms against last year. Vehicles, ships and rare earths are among the fastest growing categories. Toys, footwear, furniture and steel are the laggards.
So the goods China now sends the world in volume are the goods that are structurally short, priced into an AI infrastructure buildout that is not remotely satisfied. Chinese PPI can go straight back into negative territory as the energy base rolls off, and it will tell you almost nothing about the price of a shipment of accelerator components or the substrate that goes under them. Those prices are set by scarcity, not by the cost of diesel.
Layer on trade policy. Washington applied a 12.5 percent levy on Chinese goods in late July, replacing a 10 percent rate that had lapsed. That is a wedge sitting directly between the Chinese factory gate and the American shelf, and it widens exactly as the factory gate deflates.
The conclusion follows without much effort. The disinflation everyone is about to celebrate in headline prints is energy arithmetic. The channel that historically made goods disinflation durable, cheap Chinese manufactures flooding out at falling prices, is now dominated by categories where China has pricing power rather than surplus capacity. When the energy base effect flips in 2027, there is far less behind it than the last twenty years of experience would suggest.
There is a flow consequence too. A collapse from $126 to $70 while exports run at record levels is one of the largest terms of trade windfalls China has had in a decade. The July surplus came in at $112.5 billion. That purchasing power does not evaporate. It accumulates, and it gets recycled, and where it gets recycled is a rates and currency question that nobody is asking while they argue about a consumer price headline.
What would have to be true for me to be wrong
The bear case on my own argument is serious, and I want to state it properly.
Chinese domestic demand genuinely is weak. Core CPI at 0.3 percent is not a rounding error, it is a signal, and property remains a drag with no clean resolution. If the AI capex cycle decelerates even modestly, the semiconductor scarcity premium unwinds quickly, and China is left with enormous installed capacity in chips, EVs, batteries and solar looking for a buyer. In that world the old pipeline restarts with real force and exports disinflation to everyone, exactly as the reflex assumes. My composition argument depends on the tech cycle holding, and that is a genuine assumption rather than a fact.
The other way I am wrong is faster. The war is not settled, only quieter. Any renewed threat to Hormuz reprices crude in hours, the base effect inverts, and the entire disinflation narrative dies inside a single session.
What I am watching
Sequential Chinese PPI rather than the year on year figure, because the annual number is now mostly a story about April. Chinese export price deflators by category, separating tech from everything else. The energy contribution line inside US CPI, which is where the arithmetic will be most visible. Brent term structure, since the front month trading below the next is the market telling you supply anxiety has drained away. And the FOMC on 16 September, with the target range held at 3.50 to 3.75 percent since December and a committee that has explicitly attributed elevated inflation to supply shocks.
A delta-neutral volatility book does not express this as a view on direction. It expresses it as a view on the spread between instruments that key off headline inflation and instruments that key off core, because that spread is where the misreading actually lives. The repricing has not started, and the mechanism is not priced.
I am Komey Tetteh, and my question for you is simple. When the next three headline inflation prints come in soft, will you check whether anything actually got cheaper, or will you assume the calendar was telling you something about the economy?
General information and market commentary only. Nothing here is financial advice, a recommendation, or an offer to transact in any security.
