China factory gate shuts for a third time in four months
The official manufacturing gauge fell to 49.3 in October from 50.1, and the statistics bureau put the holiday calendar and the international environment in the account. Inside the average, construction, consumer goods and small firms are contracting while high technology and equipment expand, a split that does not fit the overcapacity charge signed in Atlanta the same weekend.
China’s official manufacturing gauge fell to 49.3 in October from 50.1 in September, the third month in four below the line that separates expansion from contraction, according to the National Bureau of Statistics. The non-manufacturing index fell to 49.7 and the composite to 49.7, from 50.7. Economists had expected the factory number to hold at 50.0, Reuters reported, and the median in Bloomberg’s survey was 49.9.
The bureau published its own explanation with the number. Huo Lihui, the chief statistician of the service that compiles the survey, wrote that the decline reflects demand pulled forward ahead of the October 1 national holiday and the complex international environment, and that the manufacturing index had therefore moved back below the threshold. The sentence matters more than the practice of explaining a single month: the office that produces the number put the calendar and the external picture into the account, and left domestic demand out of the first line.
What the average hides
The sub-indices carry the argument. Production fell to 49.8 and new orders to 49.4; consumer goods fell to 48.8 and construction dropped 2.3 points to 48.0. High technology stood at 51.3 and equipment at 50.6, both in expansion, and the figures were reproduced in English with the same reading. By size, large enterprises held at 50.2 while medium firms sat at 48.4 and small firms at 47.5. Input prices stayed in expansion at 56.8, down 4.0 points, and output prices at 51.7, down 2.3, so input and output prices were both higher than a month earlier, though rising more slowly than in September.
That split is the story the headline number cannot tell. The contraction is where credit is tightest and demand is closest to the household: small manufacturers, consumer goods, and construction, the sector that carries local government revenue and employment for workers who move between sites. The expansion is where the state has aimed industrial policy for a decade: high technology and equipment. China’s own official survey shows large firms and strategic sectors above 50 while the rest of the factory floor is below it.
Zhang Liqun, a researcher at the Development Research Center of the State Council, read the same table as evidence of insufficient demand, in comments carried by Reuters. The official note and the government economist’s reading sit next to each other without contradiction: demand weakened, orders fell, and the sectors the state has chosen to back kept expanding.
The weekend on both sides of the Pacific
The October table landed on October 31, 2026, the day the foreign ministers of the Quadrilateral Security Dialogue (Quad) signed a joint statement in Atlanta that records grave concern about economic coercion, arbitrary export restrictions and non-market policies that generate overcapacity. The two documents describe different things and are usually read as if they described the same one. A purchasing managers’ index measures the direction of activity reported by purchasing officers; it says nothing about installed capacity, nothing about export prices and nothing about whether a plant is running below its own designed output. The overcapacity charge needs utilisation data, which China publishes annually and with a delay, and it needs factory-gate prices moving down, which October did not show.
What the index does support is narrower and more useful. Chinese manufacturers reported weaker orders in a month when the holiday moved production forward, and the weakness concentrated in the smaller firms that employ most of the workers. An economy that is dumping surplus output on the world would show up differently: orders above the line, prices falling at the gate, and small firms riding the wave. October showed the opposite ordering in two of the three: orders below the line and small firms contracting.
The private Caixin manufacturing gauge for October had not been published as of November 1, 2026, and it usually diverges from the official one, because it samples smaller and more export-exposed firms. Its absence limits how far this reading can be pushed. So does the calendar: the bureau itself cited the holiday pull-forward, and a single month that loses working days can fall without the trend turning. The right test is November, when the holiday effect is gone and the export orders the table registered are tested against the winter.
The average fell below 50. Which half of the factory floor stayed above it is the number that matters.
The question the October survey leaves is what Beijing does with a contraction that sits in construction, consumer goods and small firms while its strategic sectors keep expanding. The bureau attributed the fall to the holiday calendar and the international environment; the government’s own researchers attributed it to insufficient demand; and the Quad attributed the problem to Chinese capacity the world has to absorb. Three explanations were published in the same 72 hours, and they cannot all be describing the same economy. Whoever is right will be visible in November, when the holiday is out of the data and the small firms either recover their orders or confirm that the weakness was never about the calendar.