China's Register Is the Ceiling: 67.1 Per Cent Utilisation and a Government That Has Stopped Believing Its Own Capacity Number

There is a number in the Chinese cement industry that looks like a measurement and is actually a statement about paperwork. The 2025–2026 building-materials capacity plan requires producers to file a capacity-replacement scheme for tonnage above their project registration. Read that requirement carefully: it is the ministry conceding, in an administrative instrument, that a material share of Chinese clinker capacity is not on the register at all.

Which means the 67.1 per cent utilisation figure that the global cement market appears to support is a ratio to a document, not to a set of plants. It is a ceiling on what can be claimed. It is not a floor under what exists.

The arithmetic, and what it is a ratio to

Start with the register. 1,800 million tonnes of registered clinker capacity, divided by a 0.71 clinker factor, implies 2,535 million tonnes of cement-equivalent capacity. Apply the 67.1 per cent utilisation and you get roughly 1,701 million tonnes of output — about 44.8 per cent of the 3,800 million tonne world market.

Every step in that chain is defensible. The problem is what the chain is a chain of. Each stage inherits the register’s assumption. If unregistered capacity exists — and the replacement-scheme requirement exists precisely because it does — then the true denominator is larger, true utilisation is lower, and the industry is meaningfully more oversupplied than the published ratio suggests. The 1,701 million tonnes of output is not in doubt. The 67.1 per cent is.

For anyone modelling Chinese supply, that is the whole issue. A capacity figure that is administratively defined can be restated. A capacity figure that is physically defined cannot. This one has already been restated at least once, and the direction of the next restatement is not knowable from the outside.

Why this is a price question, not a volume question

Cement’s growth has stopped being about tonnage. Global output runs 3,900 to 3,800 million tonnes, with 2024 revised down from 4,000, and volume grows at 0.37 per cent. Value grows at 3.03 per cent, and on the report’s decomposition 86.3 per cent of the growth is price. The market is not expanding. It is repricing, on flat tonnage, and the repricing is the entire investment case.

A price-led market needs supply discipline to hold. Discipline in a fragmented industry with a large off-register capacity overhang has to come from somewhere other than the cost curve — because the cost curve is what produces the overhang in the first place. In China it comes from policy: capacity replacement ratios, output controls, and now a national emissions trading system whose quota allocation is tied to output.

That last instrument is the one to watch, because it links the register to the cost base. If quota volume is allocated against output, then output becomes both the denominator of utilisation and the numerator of a carbon liability. A plant that under-reports output pays less carbon and looks more efficient against capacity. A plant that reports accurately pays for the privilege. The incentive runs in one direction, and it runs against the accuracy of the very number the market is using to argue that the industry is disciplined.

The counter-argument, stated properly

It would be wrong to read this as an industry with no control over its own price. Three things argue the other way.

Consolidation is real. The largest producers have grown, and in a market where clinker is expensive to build and permits are hard to obtain, scale confers the ability to idle. Idling is how a fragmented industry behaves like a disciplined one without structural change.

Clinker substitution gives a second lever. A 0.71 clinker factor is not a fixed property of the industry; it is a choice per tonne. Raising the substitution rate reduces clinker demand without touching clinker capacity, which lets the industry manage clinker utilisation and cement output semi-independently — a flexibility that cement markets outside China have used for two decades.

The alternative investment has gone. Property construction has fallen far enough that the marginal new plant is not being financed. Capacity growth stops before capacity retires, and a market can hold price for a long time on a shrinking capex pipeline.

None of those resolves the measurement problem. All three explain why the industry can hold price while the measurement stays broken.

What the number is actually for

Three users, three different uses, and only one of them is legitimate.

An investor uses 67.1 per cent as a proxy for how much idling is needed to tighten the market. That is a reasonable use, provided it is treated as a scenario input with a wide band rather than a measurement. The band is the story.

A competitor uses it to infer the marginal cost position of a plant they cannot see. That is where the register’s weakness does the most damage — the plants that are off-register are, almost by definition, the ones that would set the marginal price.

A policymaker uses it to size the replacement task. That is the one use that works, because the policy is defined in terms of the register. The instrument and the measurement are consistent with each other even when neither is consistent with the physical world.

The error is to take a well-defined ratio to an administratively defined denominator and treat it as an observation about plants. It is not. It is an observation about what the ministry is willing to count, and the gap between the two is the most important number in the Chinese cement market — precisely because nobody has published it.

About the method

This analysis draws on the capacity, clinker-factor and utilisation series in the Global Cement Market Report, 2026–2033, weighted approximately 70 per cent secondary and 30 per cent primary research. The secondary layer is Global Cement and Concrete Association activity data, USGS National Minerals Information Center statistics, national statistical releases and producer disclosures. The primary layer is structured interviews and open-ended conversations with producers, traders and plant technologists on the C10+ summit circuit, addressing what “registered” means at plant level and how the replacement-scheme filings are being handled. The 1,701 million tonne output, the 2,535 million tonne cement-equivalent capacity and the 44.8 per cent world share are calculations from the register, the clinker factor and the stated utilisation — not published figures. No third-party market-research estimate is used anywhere in this piece.

Further reading

• Global Cement Market Report, 2026–2033 — capacity, clinker factor, product mix and the price decomposition.
• Global Ready-Mix Concrete Market, 2026–2033 — the downstream market where Chinese volume falls 30.5 per cent in four years.
• Carbon Stopped Being a Cost and Became a Price — why output is now a cost base as well as a capacity denominator.
• Twenty Bags Cost USD 3.77 Against a USD 90 Tonne of Cement — what a falling tonnage does to the packaging chain attached to it.