If I could only own one semiconductor company for the next twenty years and never look at it, I would probably pick ASML. It is the closest thing the industry has to a true monopoly, it sits at the narrowest point of the entire supply chain, and its customers cannot advance without it. The catch, and it is a real one, is that the stock trades on order flow that swings wildly from quarter to quarter, so owning the best business in the sector still means sitting through drawdowns that would shake most people out. Reconciling those two facts, the quality and the timing, is the whole problem with the stock.
Why the Moat Is as Wide as It Looks
Making an advanced chip requires printing circuit patterns onto silicon at a scale measured in a few nanometers. The only machines in the world that can do this at the leading edge use extreme ultraviolet light, and ASML is the only company that makes them. Each system costs on the order of two hundred million dollars for the current generation and more for the newer High-NA version, contains hundreds of thousands of parts, and takes years and a fleet of aircraft to deliver and install. The technology took ASML and its partners decades and tens of billions of dollars to get working. No competitor is close, and the barriers to becoming one are close to prohibitive.
On top of selling the machines, ASML earns a large and growing stream of service and upgrade revenue from the installed base, which smooths the business over time and is far less cyclical than new system sales. That installed-base revenue is the part of the story that gets lost when the stock is falling on a weak bookings number.
The Lumpiness Is Structural
Here is the problem for anyone who wants a smooth ride. ASML’s customers are a handful of chipmakers, and each of them orders in large, irregular batches tied to when they build or upgrade a fab. A single customer delaying or pulling forward an order can swing ASML’s quarterly bookings by billions of euros. So the reported order number lurches from a record one quarter to a disappointment the next, and the stock, which many investors trade on that number, lurches with it. The ASML analyst page and its revision trend are more useful here than any single day’s price, precisely because they smooth over that noise. Over the past couple of years ASML has had quarters where bookings roughly halved sequentially and quarters where they nearly doubled, with no fundamental change in the multi-year demand picture.
Layered on top of that is China. For a stretch, Chinese chipmakers rushing to buy older-generation equipment before export controls tightened made up a very large share of ASML’s sales, and everyone knew that share would normalize down. As it has, the year-over-year comparisons have looked weak even while demand from leading-edge customers building for AI has strengthened, a pattern I traced through the supply chain in a piece on who actually profits from the AI infrastructure wave. The mix shift is healthy for the long run and ugly in any given quarter’s headline.
The Forecast Over the Right Time Frame
ASML has laid out a model for the end of the decade that points to meaningfully higher revenue and gross margins than today, driven by rising chip content across AI, data centers, and electrification, plus the ramp of High-NA systems. I do not take those targets as gospel, but the direction is well supported. The number of transistors the world wants to manufacture keeps rising, and every one of them at the leading edge passes through an ASML machine. That is also why I keep coming back to it in surveys of the chip stocks worth owning beyond Nvidia.
Over one to two years, the stock is nearly impossible to time, because it trades on the bookings sawtooth. Over five to ten years, the demand case is about as clear as anything in technology. The mistake is buying a big position right after a strong bookings quarter when the stock is up thirty percent and everyone is comfortable, and then selling it after a weak one when it is down thirty and the narrative has flipped. The business did not change in either direction.
| Time frame | What drives the stock | How predictable |
|---|---|---|
| Next quarter | The bookings headline and China commentary | Low |
| 1 to 2 years | Fab build timing, memory cycle, export policy | Low to moderate |
| 5 to 10 years | Total transistors manufactured at the leading edge | High |
How I Actually Own It
I own ASML and I add to it on weakness, specifically after the quarters where a soft bookings number sends the stock down and the coverage turns cautious. The forward multiple has usually been in the low thirties, which is not cheap, but for a monopoly on an input the entire digital economy depends on, I am willing to pay a premium as long as I am buying into pessimism rather than into a bookings-fueled rally. The ASML quote page has the live valuation, and the quant rating tool is most useful right after one of those weak prints, when Valuation and momentum move the most. If you cannot sit through a thirty to forty percent drawdown without selling, own this theme through a fund. If you can, ASML is the highest-quality single name in it, and the quarters everyone hates are the ones to buy.
Gavin Thorne writes on equity strategy and company-level research. This article reflects his personal research process and is intended for informational purposes only. It does not constitute investment advice.