Decoding China's Four Monthly PMIs for AUD Traders

Traders focused on the Australian dollar should note that China's PMI data comprises four monthly releases, each with distinct significance.

Unexpected manufacturing readings, particularly from the official NBS survey, usually trigger the most immediate and pronounced market moves because they closely reflect demand in heavy industry and construction—key drivers of Australian commodity exports. In contrast, services and non-manufacturing data tend to have a more gradual impact, yet they remain important for assessing domestic demand, especially the construction sub-index inside the NBS Non-Manufacturing PMI, which offers a direct gauge of Chinese property activity and its spillover into steel and iron ore use.

The private Caixin/RatingDog services survey, with its focus on consumption and small enterprises, can at times provide better insight into Chinese household spending than either of the official measures. When both manufacturing and services figures from the two sources point the same way, confidence in an AUD position strengthens quickly; but when they differ, especially between factory and services readings, markets usually wait for additional evidence before taking a stance.

---

China releases four PMIs each month from two different sources covering two sectors, and AUD traders who focus solely on the manufacturing headline are neglecting a significant portion of the data.

Main points

• China's PMI data set includes four monthly components: the official NBS Manufacturing PMI, the NBS Non-Manufacturing PMI (services and construction), the private Caixin/RatingDog Manufacturing PMI, and the private Caixin/RatingDog Services PMI.
• Each provider also issues a composite output index that combines their manufacturing and services readings into one measure of total economic activity.
• The NBS surveys rely on large panels biased toward larger, state-owned and state-linked companies, whereas the private surveys use smaller panels that favor small and medium-sized, often privately owned, export-oriented or consumer-focused businesses.
• All indices use the same underlying diffusion index methodology, constructed from weighted sub-components, with 50 marking the boundary between expansion and contraction, although the specific questions and weightings vary somewhat by sector and provider.
• The construction sub-index within the NBS Non-Manufacturing PMI offers an especially clear view of Chinese property and infrastructure activity, which in turn influences demand for Australian iron ore, coal, and other commodities.
• The private services survey is more oriented toward consumer-facing activity, so it can provide a distinct, and occasionally earlier, signal on Chinese household demand compared to the state-weighted NBS services index.
• Recent data exemplifies this: the NBS Manufacturing PMI for August increased to 49.8 from 49.2, while the NBS Non-Manufacturing PMI remained unchanged at 49.0. The private manufacturing survey, expected later today, is forecast to rise further from July's 50.9.

China stands out among major economies by releasing two complete sets of Purchasing Managers' Index surveys every month—one official and one private—each covering manufacturing and services. Grasping how the four resulting data points interrelate, along with their composite indices, clarifies why the Australian dollar can respond strongly to one release while barely budging on another.

The official data comes from China's National Bureau of Statistics, in partnership with the China Federation of Logistics and Purchasing. It produces a Manufacturing PMI and a separate Non-Manufacturing PMI, which covers services and construction, and combines them into a Composite PMI Output Index. The survey panels are weighted toward larger companies, including a substantial proportion of state-owned and state-linked enterprises. This bias makes the construction sub-index within the NBS Non-Manufacturing PMI particularly revealing, as it provides a relatively direct window into Chinese property and infrastructure activity—a sector that significantly influences demand for Australian iron ore, coal, and other bulk goods.

The private side, long known as the Caixin PMI and now branded RatingDog in collaboration with S&P Global, follows the same structure: a Manufacturing PMI, a Services PMI, and a composite index. Its panels are smaller, typically around 500 companies, and are more weighted toward small and medium-sized private businesses. On the manufacturing side, this bias favors export-oriented producers; on the services side, it captures more consumer-facing activities such as retail, logistics, travel, and tech-related services. Consequently, the private services index serves as a useful, and sometimes early, cross-check on Chinese household demand—an area where the state-weighted NBS panel is less adept at providing a clean reading.

All four headline indices are constructed as diffusion indices (explained further below), combining weighted sub-components like output, new orders, employment, and prices, with readings above 50 indicating expansion and below 50 indicating contraction. Because they use the same methodology, the manufacturing pair and the services pair generally move in tandem over time, as they measure the same economy. However, divergences frequently occur and can be informative—for example, when infrastructure spending boosts state-linked manufacturers more quickly than it reaches smaller private firms, or when export orders rise while domestic consumption stays weak.

For traders using the Australian dollar as a liquid proxy for Chinese economic exposure, all four data points carry weight, but not equally. Manufacturing surprises, especially from the NBS survey, typically prompt the fastest and most pronounced reactions because they are closely tied to the commodity-intensive, capital-heavy side of the Chinese economy. Non-manufacturing and services releases tend to have a slower impact on AUD, but the construction sub-index within the NBS Non-Manufacturing PMI warrants special attention due to its link to the property sector, while a strong or weak private services reading can alter the overall assessment of Chinese consumption. This week provided a real-time example of this interplay: the NBS Manufacturing PMI rose to 49.8 in August from 49.2, while the NBS Non-Manufacturing PMI stayed flat at 49.0, showing that factory improvement has not yet been matched by services and construction strength. Today's private manufacturing survey, expected to edge up from July's 50.9, will add more context, and traders should consider it—along with its services counterpart when released—alongside the official data rather than in isolation.

---

Understanding Diffusion Indices

A diffusion index converts qualitative survey responses—"did conditions improve, worsen, or stay the same?"—into a single numerical value that can be monitored over time. PMIs are the most familiar instance, but the method originated much earlier, in the 1940s and 1950s, when it was created to summarize business cycle surveys before timely production or sales data existed.

Construction of a Diffusion Index

Each month, purchasing managers report whether specific indicators—such as output, new orders, and employment—have increased, decreased, or remained unchanged from the previous month. The index is then computed as follows:

Diffusion index = (percentage reporting an increase) + (0.5 × percentage reporting no change)

Respondents who report no change receive half weight because they do not contribute to either direction, thus pulling the index toward the midpoint instead of counting as a full vote.

The Significance of 50

If all respondents reported an increase, the index would reach 100. If all reported a decrease, it would fall to 0. If responses are evenly divided among increases, decreases, and no change, or if nothing changes, the index settles at 50. Thus 50 serves as the threshold between expansion and contraction: above it, more firms are improving than deteriorating; below it, the opposite holds.

What a Diffusion Index Measures and What It Doesn't

A key point is that a diffusion index measures the breadth of change, not its magnitude. A PMI of 55 means more firms are experiencing growth than not, but it does not indicate the actual amount of output growth—a small increase reported by 60% of companies lifts the index the same as a large increase reported by 60% of companies. This is why PMIs are best interpreted as sentiment and momentum indicators rather than direct proxies for GDP growth, and they are valued for their early availability each month, before harder production or trade data emerges, despite being based on opinions rather than actual output.

Relevance for Comparing NBS and Caixin/RatingDog Data

Because composite PMIs are weighted combinations of several component diffusion indexes—such as output, new orders, employment, delivery times, and inventories—a change in the headline number can stem from very different underlying factors. Two PMIs might show the same headline figure for entirely different reasons: one because more firms reported slightly stronger orders, another because fewer firms reported a sharp drop in output. This underscores why the sub-indices, not just the headline, deserve attention when NBS and Caixin/RatingDog readings diverge.