ASML capacity is now the thing the company leads with. On Wednesday, after a record third quarter, CEO Christophe Fouquet said on the earnings call that “30% in 2027, 30% in 2028 will not allow us to meet our customer demand,” and that some customers are already discussing orders for 2029 and 2030. The company is looking at ways to go beyond the 30% it had planned for each year.

The quarter itself was big. Net sales were €11.66 billion, up about 55% from a year earlier, with a gross margin of 57.5% and net income near €5.0 billion. Revenue landed about €70 million under consensus and earnings per share beat by a wide margin, per Seeking Alpha. The fourth-quarter guide is €14.2 to €15.2 billion, and the full-year range moves to €44 to €45 billion. Net income includes a gain of roughly €500 million from revaluing ASML’s first tranche of its stake in Mistral, the company whose record €3 billion round Samsung led, so I would read operating income rather than the headline.

What I find more interesting than the sales is how demand shows up. Stock Titan reads the balance sheet and notes that ASML no longer publishes bookings; the line that moved is long-term contract liabilities, money customers have paid or been billed ahead of delivery, which went from about €3.6 billion to €9.3 billion in the quarter. That is Stock Titan’s reading of the filing, not ASML’s own framing, and I would not hang a thesis on it.

It is still a stranger and more convincing signal than a CEO quote.

On the technology side, Biggo’s write-up of the call has Samsung confirming a 2028 start of high-volume manufacturing with High-NA EUV, TSMC confirming a 2030 introduction, and TSMC asking ASML to lead a 12-inch reticle consortium. That is a long way from the position I covered when TSMC said it did not need High-NA and then signed up anyway. Stock Titan also calculates that China’s share of ASML’s machine sales fell from 42% to 23%, its own arithmetic, which squares with what the MATCH Act would do to servicing in China if it passed.

One caution, because this is finance and not my lane. Customer forecasts are not purchase orders, and added capacity has not been approved. Also, ASML’s tools are a bottleneck for chipmakers upstream of the packaging steps I looked at when Besi was downgraded over a hybrid bonder ASML had not announced. If ASML cannot ship enough, it is the first constraint in the chain, and everything priced on AI demand next year sits behind it.

Sources