SANTA CLARA, 15 AUG 2026 — Applied Materials reported record quarterly revenue on Thursday, beat its own guidance, forecast a bigger quarter still, and watched its shares fall about five per cent.
The disconnect is explained by one line in the results. China is now roughly 28 per cent of sales, down from about 35 per cent a year earlier, and the company expects Chinese spending on chipmaking equipment to fall further in 2026.
The quarter
Revenue beat the consensus by about US$130 million. The fourth-quarter guide beat it by nearly US$700 million, which is not a rounding difference on a business this size. On any ordinary reading this was a strong set of results.
Why the market looked past it
Applied Materials sells the machines that make chips. That position makes it the clearest available instrument for reading the industry, and it also makes it the place where export controls are felt as revenue rather than as policy.
Chief executive Gary Dickerson put a number on the effect. Trade rule changes reduced the company's addressable market in China by an amount equivalent to roughly ten per cent of that market in fiscal 2024, and more than double that in fiscal 2025. The China share of sales has fallen by about seven percentage points in a year, and the company expects Chinese equipment spending to decline again in 2026.
Investors see a business pulled in two directions. Demand from the AI build-out is accelerating, while its large China market is shrinking by design. The record quarter speaks to the AI demand. The forward-looking guidance is about the weakness in China.
The other side of a story we ran this morning
This connects directly to a story we ran this morning on SMIC.
We wrote earlier today that SMIC is raising prices because it cannot expand capacity fast enough — its monthly capacity grew 1.7 per cent while shipments rose 14 per cent. The reason it cannot expand is that the tools are hard to buy.
Applied Materials is one of the companies that makes those tools. The seven percentage points of China revenue Applied Materials has lost is the flip side of the production capacity SMIC cannot add. One firm reports it as a shrinking addressable market; the other experiences it as a constraint that hands it pricing power.
This symmetry offers the clearest evidence yet of what the export controls are actually doing. They are not stopping Chinese production. They are slowing its growth, and the slowing is showing up as higher prices for Chinese customers and lower revenue for American suppliers.
Why this company is read as a forecast
Equipment orders lead chip output, which is why a tool vendor's guidance is treated as information about everybody else.
A fab commits to machines long before it produces wafers on them. Tools are ordered, built, shipped, installed and qualified over a period measured in quarters, so what Applied Materials books today describes capacity that arrives well after. When it guides to US$10.3 billion, it is describing decisions its customers have already taken about capacity for a period beyond this one.
That is why the China line carries more weight than its current revenue share suggests. A seven-point drop in China's share represents more than revenue already lost. It represents Chinese chip capacity that will not exist in two years, tightening the future supply of mature-node parts well into the decade. The same logic runs the other way on the AI side, where the guidance implies customers have already committed to capacity that has not yet been built.
Who is paying for the policy
Set the three results side by side and the distribution becomes visible.
SMIC is more profitable, because scarcity lets it charge more. Applied Materials is growing strongly and has lost a chunk of one market. Intel has just raised US$20 billion, five times oversubscribed, against capacity that includes the packaging step everyone is short of.
Nobody in that list is losing. The net effect is a redistribution. Demand that would have funded Chinese expansion is now funding AI-related capacity elsewhere, and the equipment makers have, for now, traded one for the other.
This raises the question of what happens if AI demand slows while the China restrictions remain. The equipment sector would then have a smaller total market and no offsetting engine, and the guidance conversation on this quarter's call is a mild preview of it.
What it means for this region
This has two consequences for the region.
First, mature-node parts stay tight. If Chinese equipment spending falls again in 2026, slower growth in Chinese capacity for power management chips, display drivers and image sensors means higher input costs for the electronics manufacturers of Malaysia, Thailand, Vietnam and the Philippines, arriving with a lag.
The second is opportunity, and it is the same one South Korea just put five trillion won behind. Equipment and materials are where the constraint sits, and a supplier tier that is not exposed to Chinese demand is now more valuable than one that is. Regional suppliers with customers spread across Taiwan, Korea, Japan and the United States are in a structurally better position than they were two years ago, and that is a market position rather than a subsidy.
The caution attached to both is that this is a policy-made market. The seven points of China revenue did not move because of anything Applied Materials did, and they could move back the same way.
What we could not establish
How much of the China decline is export controls and how much is a Chinese customer base deliberately buying domestic. Both are happening, the company's disclosure does not separate them, and they have very different implications: one reverses with policy, the other does not.
Also unestablished: the split of the fourth-quarter guidance between memory and logic, which specific tool categories lost the most China revenue, whether any of the lost demand has been recovered elsewhere, and what the company assumes about Chinese spending beyond 2026. The share price reaction is a market judgement rather than a fact about the business, and single-day moves routinely reverse.
What to watch
Whether the China share keeps falling at this rate. Another seven points in a year would take it near 20 per cent, and at that level it stops being a swing factor in the guidance.
Whether the fourth quarter delivers the US$10.3 billion. Guidance that far above consensus sets a bar the company then has to clear, and missing it would be read as the China effect arriving faster than the AI offset.
And whether Chinese domestic equipment makers start appearing in these numbers as competitors rather than as an absence. That is the outcome the controls make more likely over time, and it is the one that would not reverse if the policy did.
View author profile → · Editorial policy
