rewrite this title Too Big to Fail, But Is ASML Too Big to Grow Revenue? – Nanalyze

We try to check in on stocks that we think might have a home in the Nanalyze Tech portfolio every year or so. This is to see if the company has made the necessary adjustments to satisfy our investment criteria or if the investment thesis has changed based on things like technology disruption or market fit. Our evaluations try to tune out the short-term noise while investigating the long-term prospects. Of course, certain trends tend to dominate at any given point in time. A recent comment by ASML (ASML) CEO Christophe Fouquet during the company’s Q1-2025 call captures the current zeitgeist for any company that develops technologies that support AI.

“Our conversations so far with customers support our expectation that 2025 and 2026 will be growth years. However, the recent tariff announcements have increased uncertainty in the macro environment and the situation will remain dynamic for a while. As previously shared, artificial intelligence continues to be the primary growth driver in our industry. It has created a shift in the market dynamics that benefits some customers more than others, contributing to both upside potential and downside risks as reflected in our 2025 revenue range.” – CEO Christophe Fouqet

In other words, Fouquet has no clear idea where this is all really headed amid ye olde macroeconomic headwinds. That doesn’t mean we can’t offer a clear-eyed analysis of where ASML is going based on recent results.

ASML Revenues Stall

ASML is a Dutch company that manufactures lithography machines, which are used to produce advanced computer chips at scale. In our last article on this pick-and-shovel semiconductor stock, we were greatly impressed with the revenue growth (30% in 2023) and gross margins (low 50%) for a decades-old hardware company. In addition, the technological moat is nearly uncrossable, helping turn ASML into Europe’s most valuable company.  On the downside, ASML relies on just a handful of customers for its $30 billion-plus in annual revenue. That’s not surprising, given the complexity and cost of its chip manufacturing equipment, especially its extreme ultraviolet (EUV) lithography machines, which boast an average selling price (ASP) of about $200 million. The highest-end systems retail closer to $400 million.

Revisiting ASML stock about 18 months down the road, we find that revenue growth has stalled spectacularly – just 2.6% in 2024. One only has to look at the revenue mix to understand a key factor. Logic System revenue fell 17%, reflecting the sluggishness in traditional chip applications in products like smartphones and PCs. Meanwhile, Memory revenue jumped 44%, driven by investments in high-bandwidth memory (HBM) and dynamic random access memory (DRAM), which are essential for AI and other high-performance computing (HPC) applications. EUV system sales, representing those high ASP machines, were down 9%. Service revenue (i.e., installed base management) climbed about 16%, but we’re mainly interested in the tech. 

Despite the high hopes in AI to drive growth, challenges are expected to persist in 2025, according to management. Indeed, total revenue dropped from nearly $11 billion in Q4-2024 to about $9 billion in Q1-2025. General uncertainty is reflected in their 2025 guidance, ranging from $35 billion (6% growth) to more than $41 billion (24% growth). That’s way off from projections in 2023, when the company projected annual revenue of between $33 billion and $44 billion by 2025. It’s hard to see ASML hitting the higher end of even the latest guidance range. (Note that ASML reports revenues in euros, so we’ve converted the number into USD for our U.S.-centric audience. The ongoing slide by the USD against the euro makes these numbers a little fluid.)

Near-term Uncertainty

There are a few reasons for pessimism. For instance, China contributed a record 41% in system sales in 2024, as the company worked to alleviate the backlog and respond to accelerated purchases ahead of anticipated export controls. Management expects the China market to normalize into the low-20% share in 2025, though Q1-2025 still saw a strong share of 27%. A surge in service revenue in Q4-2024, accounting for 24% of total revenue that quarter, also seems unsustainable, though the share in Q1-2025 actually ticked up to nearly 26%. Finally, the adoption of ASML’s most expensive and complex machines is a slow process, predicated on a number of factors, including customer capacity (more on that later).

In fact, it could be a while before revenue resumes double-digit growth. The company is devoting lots of R&D resources (15% of total sales) on its most advanced machines. However, customer adoption appears to be slow despite efficiency gains, especially given that production-scale deployment requires at least 12 months of testing and vetting (again, more on that later). The fact that China is prohibited from buying ASML’s most advanced machines – specifically, EUV lithography systems and the latest deep ultraviolet (DUV) tools – further limits the company’s near-term market prospects. For example, the number of lithography systems sold in Q4-2024 to Q1-2025 dropped from 119 to 73. 

Competition from China

The China ban is forcing the Chinese to develop their own EUV lithography technology. News reports claim that Chinese companies, led by Huawei and SMIC (the largest contract chip manufacturer in China), are advancing domestic EUV lithography systems using a technology called laser-induced discharge plasma (LDP). Trial production could happen as early as this quarter with mass production targeted for next year. The LDP method involves vaporizing tin between electrodes to generate EUV light, potentially offering cost and efficiency advantages over ASML’s laser-produced plasma (LPP) approach. Beyond changing up an acronym, Chinese companies are reportedly investing more than $40 billion in the effort.

Naturally, Fouquet characterized China’s EUV progress as “research news rather than product yields,” adding it would take “many, many years” to develop viable systems. It’s probably not just bluster. After all, ASML has spent decades building a global supply chain to integrate millions of components and sub-systems into a technology that’s about as sophisticated as the Death Star but with a much lower body count.

Competing Technologies

EUV lithography is not the only approach for working at the nanoscale required to produce advanced semiconductors for AI and HPC. Canon is developing something called nanoimprint lithography (NIL), which presses circuit patterns onto wafers like a stamp. This approach avoids  the complex, expensive, and power-hungry requirements associated with EUV technology. For instance, one Canon system has reportedly achieved promising results at 90% less power and potentially one-tenth the cost of ASML’s EUV tools for some applications. While often mentioned during debates of ASML rivals, Nikon is focused elsewhere.

Other attempts to extend Moore’s Law without relying solely on ever-smaller lithography revolve around new transistor architectures and advanced packaging. Without getting too technical (because we’re MBAs not aliens with advanced technologies that we’re slowly leaking to our human slaves), many of the concepts involve building up in three dimensions instead of just shrinking sideways. Stacking transistors vertically theoretically lets chipmakers pack more power into the same footprint. Companies like TSMC, Intel, and Samsung are leading the way on this front. However, these innovations are generally seen as complementary to advanced lithography, not as a replacement. 

Too Big to Fail or Flourish

In some ways, ASML is a victim of its own success. Thanks to its near-monopoly in EUV technology and the staggering expense of its most advanced systems, there are only a handful of companies that have the funds and expertise to play in this very elite sandbox – primarily TSMC, Samsung, and Intel. Some of these machines are the size of an elephant, reportedly requiring seven Boeing 747s to deliver each system to its destination.

As we alluded to earlier, it usually requires a minimum of a year for a customer to ramp up to full-scale production on a new ASML machine. Indeed, management emphasized in their Q1- 2025 call that the biggest barrier to rapid adoption of new systems isn’t price but tool maturity, referring to the laborious process of building up the tool’s reliability and performance before beginning full-scale production. While management insists that demand is robust, the ability of the semiconductor industry to adopt and adapt some of the world’s most advanced technologies requires vast amounts of resources. The ongoing tariff tiffs with Trump are probably adding another level of uncertainty and hesitation for these huge investments of time and money.

Another sign of general uncertainty in the industry is management’s decision to stop reporting quarterly bookings after 2025, shifting to annual backlog updates (more than $40 billion at the end of 2024). This decision apparently aims to smooth investor anxiety amid “lumpy order patterns.” For example, Q4-2024 bookings surged to more than $8 billion after customers ordered only about $3 billion in tools the previous quarter. While the backlog is healthy, the revenue growth appears to be sidelined for the near term.

Conclusion

While ASML continues to bend Moore’s Law, it is having a harder time breaking Adam Smith’s law of supply and demand given the inherent limitations of its business models – big, expensive machines that only a handful of customers can afford and operate. Based on annualized revenues (Q1-2025 revenue of $9 billion), ASML is currently projected to fall at the low end of guidance. That would give us a simple valuation ratio ($315 billion market cap/$36 billion annualized revenue) of 8.75. It’s certainly a reasonable valuation for a company unmatched and relatively unchallenged in its market niche. Factor in that the company also pays dividends (at a low 0.86% yield), with a dividend growth rate of nearly 21% over a five-year period, and ASML could be an intriguing add down the road for a dividend-growth investment strategy.