Latest / Business Strategy Talks with Fexingo: Competitive Advantage, Market Positioning, and Growth Planning / How ASML Built a Monopoly in Chip Lithography
Transcript
- Lucas: So if you look at any modern phone, laptop, or data center server, the chip inside it was almost certainly printed by a machine that only one company on Earth can build. That company is ASML, based in Veldhoven, Netherlands — population roughly 45,000. Luna: And they make the lithography systems that etch the tiny circuitry onto silicon wafers. But the key word here is 'only one company.' ASML has a de facto monopoly on the most advanced chip-making machines in the world. Lucas: Exactly. And it's not a monopoly that happened by accident. It's the result of a bet the company made in the late 1990s — a bet that nearly bankrupted them before it paid off. Let's talk about how they pulled it off, and why no competitor can replicate it today. Luna: Before we dive deep, Lucas — a quick thing. A lot of our listeners tell us that these strategy breakdowns have genuinely helped them think differently about their own work. If you've gotten something out of these conversations — maybe a new way to frame a competitive advantage, or just a better understanding of how a company like ASML thinks — a couple of dollars a month is genuinely what keeps these episodes going. You can find us at buy me a coffee dot com slash fexingo. No pressure, just if it's added value. Lucas: Yeah, I'll echo that. We keep this show ad-free and independent, and listener support is the only reason that's possible. Small amounts add up. Anyway — back to ASML. The technology at the center of their moat is called extreme ultraviolet lithography, or EUV. Luna: And that's not just a fancy acronym. EUV uses light with a wavelength of 13.5 nanometers — that's about a hundredth the wavelength of visible light. It's what allows chipmakers to print features just a few nanometers wide on silicon. Lucas: Right. And to generate that light, ASML had to invent something that didn't exist before: a machine that shoots a 50-micron droplet of molten tin into a vacuum chamber, hits it with a high-power carbon dioxide laser to vaporize it, and then hits that vapor with a second laser to turn it into a plasma that emits EUV light. That process happens 50,000 times per second. Luna: Fifty thousand times a second — that's insane. And they need to collect that light with mirrors coated in hundreds of alternating layers of molybdenum and silicon, each layer just a few atoms thick, because no glass lens can focus EUV light. It would just be absorbed. Lucas: Exactly. The entire system has to operate in a vacuum, because air would absorb the light. The mirrors are so smooth that if you scaled them up to the size of Germany, the largest bump would be less than a millimeter high. And ASML is the only company on Earth that has managed to commercialize this. Luna: So how did they get there? Because this wasn't a sure thing. I remember reading that in the late 90s, ASML was a relatively small player — they were competing with Nikon and Canon in the previous generation of lithography, called deep ultraviolet. The EUV bet was a huge gamble. Lucas: It was a moonshot. In 1997, ASML joined a consortium called EUV LLC, led by Intel and including Motorola and AMD, along with the US Department of Energy. The goal was to develop the basic science. But the real breakthrough came when ASML decided to go all-in on EUV around 2000, even though the technology was still decades from being viable. Luna: And they had to convince customers like Intel, TSMC, and Samsung to invest in the research. Because ASML couldn't fund it alone. Lucas: That's right. In 2012, ASML launched what they called a 'customer co-investment program.' Intel bought 15 percent of the company for about $4 billion. Samsung and TSMC also took stakes. Those investments gave ASML the capital to keep developing EUV through the 2010s, while also giving those three chipmakers priority access to the first machines. Luna: So the customers essentially paid for the R&D in exchange for guaranteed supply. That's a brilliant strategic move, but it also shows how much trust ASML had built — these are companies that could have tried to develop EUV themselves, but they chose to back ASML instead. Lucas: And that's part of the moat. Because once those customers are financially and technically committed to ASML's roadmap, they can't easily switch to a competitor — even if one existed. But no competitor exists, because the barriers to entry are astronomical. Luna: Let's talk about those barriers. We're talking about a single EUV machine that costs around $150 million. It takes months to assemble, requires thousands of suppliers, and involves some of the most precise optics ever built. The main supplier of those mirrors is Zeiss, a German optics company. Lucas: And Zeiss is deeply integrated with ASML. They have a joint development agreement that goes back decades. ASML actually owns about 24 percent of Zeiss's semiconductor division. So it's not just a supplier relationship — it's a strategic alliance that makes it almost impossible for a rival to replicate the mirror technology. Luna: Then there's the supply chain. ASML sources components from over 5,000 suppliers across 20 countries. They make the machines in the Netherlands and ship them in specialized containers — each machine can weigh over 180 tons and requires multiple 747 freight flights to deliver. Lucas: And the assembly happens on-site at the customer's fab, with ASML engineers spending months installing and calibrating the system. So it's not like you can just order a machine and plug it in. The learning curve for ASML itself was decades long. Any competitor would have to replicate not just the technology, but the entire ecosystem. Luna: What about Canon and Nikon? They make lithography systems for older chip generations. Could they leapfrog EUV with something else? I've heard of nanoimprint lithography — Canon's been working on that. Lucas: Nanoimprint is interesting — it's like a stamp that presses the circuit pattern into the wafer. But it has major challenges with overlay accuracy and defect control. For the most advanced nodes — think 3-nanometer and below — EUV is still the only game in town. Canon has tried to develop EUV, but they've never been able to commercialize a working system. Nikon essentially gave up on EUV years ago. Luna: So ASML's monopoly is essentially unassailable for the foreseeable future. But there's a geopolitical angle too. EUV machines are subject to export controls — the US and Netherlands have restricted ASML from selling its most advanced systems to China. Lucas: Right. And that's a huge deal. China can't build advanced chips without EUV. They've been trying to develop their own domestic lithography industry, but they're years behind. ASML's CEO has said that China is about 10 to 15 years behind in lithography. That gap is part of what makes ASML's moat so wide — it's not just technological, it's also regulatory. Luna: But at the same time, being caught in the crossfire of us china tech tensions creates risk. If China eventually develops its own EUV — or if the export restrictions tighten further — ASML could lose a significant market. Though for now, they're still selling older DUV machines to China, which are less restricted. Lucas: That's a good point. ASML sold about 20 percent of its total revenue to China in 2025, mostly older systems. But the high-margin EUV systems go to TSMC, Samsung, and Intel — all in Taiwan, South Korea, and the US. So the geopolitical risk is real, but it's also a barrier to entry for Chinese competitors. They can't just buy the technology or reverse-engineer it because the export controls block the supply chain. Luna: Let's talk about the numbers. ASML's market cap as of May 2026 is around $400 billion. They shipped about 42 EUV systems in 2025, and they're expecting to ship around 60 in 2026. Each system costs over $150 million, so that's a revenue stream of roughly $9 billion just from EUV, plus their older DUV business. Lucas: And the margins are incredible. Their gross margin on EUV systems is over 50 percent. But they also have a lucrative service and upgrade business — because once you buy an ASML machine, you're essentially locked into their maintenance and consumables. The mirrors degrade over time and need to be replaced. The tin droplets, the lasers — all of that is recurring revenue. Luna: So the moat has multiple layers: the technology itself, the supply chain integration, the customer co-investment, the intellectual property — ASML has over 10,000 patents — and the geopolitical barriers. It's almost like a perfect storm of competitive advantage. Lucas: There's one more layer I'd add: the talent. ASML is headquartered in a small Dutch town, and they've managed to attract some of the best physicists and engineers from around the world. They have a culture of extreme precision and long-term thinking. A lot of employees have been there for 20 years or more, which is rare in the tech industry. Luna: And they've also been smart about how they've positioned themselves culturally — they present themselves as an enabler of Moore's Law, as this company that makes all other tech possible. That narrative helps them attract talent and justify their pricing. Lucas: Absolutely. They've essentially become the gatekeeper for the entire semiconductor industry. Without ASML, you can't make the chips that power everything from AI to smartphones to cars. That gives them enormous bargaining power, but also enormous responsibility. Luna: So what's the threat to ASML? Is there any scenario where their moat erodes? Lucas: The most likely threat is a technological discontinuity — something that replaces EUV entirely. There's research into high na EUV, which is the next step, and ASML is already working on that. But beyond that, there are concepts like electron-beam lithography or directed self-assembly. None of them are close to commercialization. And even if a new technology emerges, ASML is well-positioned to acquire or develop it. Luna: Also, there's the risk that chip demand plateaus or that Moore's Law effectively ends. If the industry decides that 2-nanometer or 1-nanometer is good enough and stops investing in miniaturization, then ASML's growth slows. But that seems unlikely given the AI boom and the insatiable demand for more compute. Lucas: Right. AI training and inference require the most advanced chips. TSMC and Samsung are building new fabs that will need more EUV machines. Intel is also ramping up its foundry business. So the demand drivers are strong. ASML's backlog was over $40 billion at the end of 2025. They have years of visibility. Luna: So if you're a business strategist, what's the takeaway from ASML? I think it's the power of being the critical enabler in a high-stakes ecosystem. They aren't just a supplier — they're the bottleneck. Lucas: Exactly. And they built that position by making a huge, risky bet — and then by securing the cooperation of their biggest customers, integrating deeply with suppliers, and relentlessly focusing on execution over decades. It's a masterclass in building a durable competitive advantage. Luna: I think there's also a lesson about patience. Most companies would have abandoned EUV after years of delays and cost overruns. ASML stuck with it for nearly 30 years before it became commercially viable. Lucas: That's a great point. In a business world that often demands quarterly results, ASML's willingness to think in decades is rare. And it's probably the hardest thing for a competitor to replicate. I'd argue that's the deepest layer of their moat — a culture and a capital structure that supports ultra-long-term R&D. Luna: So next time I look at a smartphone or a laptop, I'll remember the 50,000 tin droplets per second happening in a Dutch cleanroom. It's wild. Lucas: Yeah. It's one of those stories where the reality is stranger and more impressive than any metaphor. And it's a reminder that the most powerful moats often come from doing something that seems almost impossible — and then making it look inevitable.