This piece explains how ASML went from a near-bankrupt Philips spin-off to the sole maker of the most advanced chip-making machines (EUV lithography). The author (Baillie Gifford fund) is bullish, arguing that as long as chips need to shrink, ASML is the only supplier. Key holdings: ASML itself (market cap €250B, monopoly on EUV machines), TSMC (~40% of ASML's revenue, its big customer), and Samsung (~30% of revenue, also a big customer).
ASML, originally a forgotten subsidiary of Philips, has become the world's most important semiconductor equipment company. In this episode, Baillie Gifford investment manager Tom Walsh explains how ASML rose from an unconventional path in the Netherlands to dominate extreme ultraviolet (EUV) lithogr
Tom Walsh (Investment Manager at Baillie Gifford) deconstructs how ASML evolved from Philips' "problem subsidiary" into the world's only manufacturer of extreme ultraviolet (EUV) lithography equipment. Core thesis: ASML's competitive moat is not another company, but Moore's Law itself—as long as chip manufacturing processes need to continue shrinking, ASML remains the sole supplier capable of providing the necessary tools.
Tom Walsh argues that ASML's rise was not due to inherent advantages, but rather a combination of technological windows, external luck, and sustained bets on "moonshot projects."
When spun off from Philips in 1984, ASML ranked 10th in the lithography industry (out of 10 companies), with no revenue, no commercial products, and no office—its first workspace was a wooden barracks on the Philips factory premises. Employees generally viewed it as a "delayed layoff." However, ASML possessed two leading technologies developed internally by Philips, along with a CEO capable of seizing the window of technological transition in the industry.
Early survival depended on weathering the brutal industry cycles of the 1980s, sustained by continuous capital injections from Philips and ASM. By the mid-1990s, ASML had entered the top three (Nikon, Canon, ASML). In 2002, it finally surpassed Nikon to become the industry leader—a goal set nearly 20 years after the spin-off.
But ASML did not stop there; instead, it accelerated its bet on extreme ultraviolet (EUV) lithography technology—a feat most believed impossible at the time. In 2019, the first EUV system entered high-volume production, granting ASML a 100% market share in the world's most advanced lithography equipment and over 90% share in next-generation tools.
> "The history of ASML is one of extraordinary ingenuity, tenacity, and innovation excellence. But in its early years, there were certain things that played out pretty well for them." — This implies that luck also played a role, such as the 1986 industry downturn eliminating competitors and the US transferring EUV technology to ASML rather than Japanese firms.
Tom Walsh believes that the R&D difficulty and scale of investment in EUV constitute the strongest technological moat he has ever seen.
The necessity of EUV stems from physical limits: when chip manufacturing processes enter the single-digit nanometer range (current leading node at 5nm), deep ultraviolet light (wavelength 193nm) is akin to "signing a name with a snow shovel"—the error rate is extremely high. In the 1990s, the industry consensus was that a next-generation technology was needed, but ASML was not initially the frontrunner in EUV.
Key turning point: The U.S. funded basic EUV research through the Department of Energy and DARPA, but was unwilling to hand the technology over to Japanese companies, and there were no domestic lithography equipment manufacturers left in the U.S. So the U.S. invited ASML to join the joint R&D consortium—a decisive moment in ASML's history.
The R&D difficulty of EUV is staggering:
ASML spent over 25 years and invested more than €10 billion in R&D. In 2013, it acquired the light source supplier Cymer (based in San Diego). Nikon and Canon successively abandoned EUV R&D.
> "I can't think of one that's more significant than ASML's. There's some barriers to entry, to say the least."
Reader's Note: This is a position-holder's perspective—Baillie Gifford is a long-term shareholder of ASML, and its judgment naturally carries the optimism of a stakeholder.
Tom Walsh notes that ASML’s financial profile is characterized by high unit prices, low sales volumes, extremely long equipment lifespans, and recurring service revenue.
| Metric | Data |
|---|---|
| 2022 Revenue | Approximately €21 billion |
| 2022 Operating Profit | Approximately €6.5 billion |
| Market Cap | Approximately €250 billion |
| 2022 Sales Volume | 345 lithography machines |
| Most Expensive Unit Price | Over €150 million |
| Share of New Machine Revenue | Approximately 75% |
| Service & Upgrade Revenue | Approximately 25% |
| Gross Margin | Approximately 50% (around 30% in the 2000s) |
| Operating Margin | Approximately 30% |
Equipment lifecycle is extremely long: Of all lithography machines sold in the past 30 years, 90% are still operating in fabs worldwide. Equipment can be upgraded on-site; machines from a decade ago can significantly boost capacity through upgrade modules—incentivizing customers to continue purchasing from ASML.
Shipping scale: Current EUV machines are roughly the size of a double-decker bus and require three Boeing 747s for transport. The next-generation High-NA EUV machines (expected to enter mass production in 2025-2026) will be even larger—because smaller patterns require larger lenses.
Tom Walsh believes that ASML's pricing strategy is key to its long-term success—despite holding a monopoly, it never abuses its pricing power.
In theory, ASML could double the price of its EUV equipment. However, the company chooses to "price based on the value created for customers," aiming for a fifty-fifty split with customers of the profit improvements driven by technological advances. This collaborative model stems from the industry's nature: a high degree of mutual dependence among ASML, TSMC, Samsung, and Intel.
Customer concentration is high: the top two customers accounted for nearly 60% of 2022 revenue, and the top three roughly 65-70%. Only a handful of companies worldwide can manufacture leading-edge chips.
ASML's manufacturing model is also distinctive: 80% of cost of goods sold comes from purchased components and materials, with only 20% from in-house labor. The company describes itself as an "architect and integrator"—a result of early capital constraints that prevented vertical integration like Nikon/Canon, which instead fostered deep partnerships with suppliers.
Tom Walsh outlines three primary risks facing ASML:
1. Supply Chain Capacity: ASML's suppliers represent both its greatest competitive advantage and its biggest vulnerability. Some suppliers are unable to keep pace with technological advancements, forcing ASML to acquire them (e.g., the acquisition of Cymer in 2013) or take equity stakes (e.g., in Zeiss). The capital expenditure required for next-generation equipment may exceed the budgets of independent suppliers.
2. Disruptive Technology: ASML's true competitor is not another company but Moore's Law itself. If ASML fails to deliver technological progress on schedule, the market will seek alternatives. A notable example: In 2012-13, when EUV technology was delayed in reaching mass production, the NAND flash memory industry pivoted to 3D NAND—increasing density through vertical stacking rather than planar shrinkage, significantly reducing reliance on leading-edge lithography equipment.
3. Geopolitics: In 2022, approximately 40% of ASML's revenue came from Taiwan, nearly 30% from South Korea, and about 15% from China. This customer structure inherently carries risk.
Tom Walsh believes ASML's investment appeal lies in structural growth combined with cash return capabilities.
The company allocates 15-16% of revenue to R&D annually while still paying dividends and repurchasing shares. Its free cash flow conversion rate is close to net profit. Currently, it is aggressively expanding production capacity: targeting an annual output of 90 EUV units (compared to approximately 40 units per year over the past two years), 600 deep ultraviolet (DUV) units, and about 20 High-NA EUV units (by 2027-28).
The measurement of technological progress differs from that in biopharma: EUV represents a generational leap occurring once every 20 years, but most improvements are modular and incremental—each module can be upgraded independently, and the risk of failure is diversified. Investors can track progress through testing metrics such as light source power.
| Position | Analyst Stance | Key Data |
|---|---|---|
| ASML | Bullish | 2022 revenue of €21 billion, operating profit of €6.5 billion, market cap of €250 billion; 100% market share in EUV; gross margin of 50%, operating margin of 30% |
| TSMC | Neutral (customer) | Accounts for ~40% of ASML's revenue (2022) |
| Samsung | Neutral (customer) | Accounts for ~30% of ASML's revenue (2022) |
| Intel | Neutral (customer) | Participated in the 2012 joint investment in ASML |
| Nikon | Risk alert (former competitor) | Surpassed by ASML in 2002; has abandoned EUV R&D |
| Canon | Risk alert (former competitor) | Has abandoned EUV R&D |
| Cymer | Acquired | Acquired by ASML in 2013; EUV light source supplier |
| Zeiss | Equity stake | ASML holds a stake; lens supplier |
1. Tom Walsh: ASML's competitive moat is not another company, but Moore's Law itself. As long as chip manufacturing processes need to continue shrinking, ASML is the sole supplier capable of providing the tools. Its true competitor is the physical limit.
2. Tom Walsh: ASML's monopoly position stems from U.S. technology transfer. The U.S. was unwilling to hand over EUV technology to Japanese companies, and with no domestic lithography equipment manufacturers left, it invited ASML to join the joint R&D effort—this was a decisive historical opportunity.
3. Tom Walsh: ASML's pricing philosophy is not to extract monopoly rents. Although it could theoretically double the price of EUV machines, the company chooses to split the profit gains from technological advancements with customers on a 50/50 basis—because abusing pricing power would incentivize customers to seek alternatives.
4. Tom Walsh: ASML's manufacturing model is that of an "architect and integrator," with 80% of costs coming from purchased components. This stems from early capital constraints, which paradoxically created a more flexible supply chain ecosystem than vertical integration.
5. Tom Walsh: 90% of all lithography machines sold in the past 30 years are still in operation. Equipment can be upgraded on-site, and machines from 10 years ago can significantly boost capacity through module upgrades—this generates recurring service revenue and customer stickiness.
6. Tom Walsh: The EUV light source requires bombarding tin droplets with a laser 50,000 times per second, reaching temperatures 40 times that of the sun's surface. This level of technical difficulty means that even if a competitor emerges, it would take over a decade to approach current EUV levels—by which time ASML would have advanced several generations.
7. Tom Walsh: When EUV mass production was delayed in 2012-13, the NAND industry turned to 3D NAND as an alternative path. This is ASML's biggest cautionary tale—if the company fails to deliver technological progress on time, the market will not wait.
8. Tom Walsh: When CTO Martin van den Brink joined ASML in 1984, veteran employees mocked him for joining a "department sent to die." He took over the company's make-or-break product development two years later and still leads the technology roadmap today—this is the best example of "human ingenuity" triumphing over skepticism.