ASML

The company began in 1984 as a joint venture between two Dutch firms, the electronics conglomerate Philips and the equipment maker ASM International. It was set up as ASM Lithography to commercialize a wafer stepper that Philips had been developing in-house, and the early years were difficult. The first product was a commercial failure, the operation ran out of a shed on the Philips campus with a handful of employees, and the business consumed cash for most of its first decade. The turn came in the early 1990s with the PAS 5500 platform, which finally gave ASML a competitive and reliable product line. The company went public in 1995 with listings in Amsterdam and on the Nasdaq, and by the early 2000s it had passed its Japanese rivals to become the largest lithography supplier in the world. The defining strategic decision of its history was the multi-decade bet on extreme ultraviolet lithography, a research program ASML began studying in the late 1990s and funded through more than twenty years and billions of euros of cumulative investment before it generated meaningful revenue.

Lithography is the step in chip manufacturing where a pattern is projected through optics onto a light-sensitive coating on a silicon wafer, defining the transistors and wiring that make up an integrated circuit. ASML sells the machines that do this, and it organizes its business around two broad families of systems plus a large and growing service operation. The first family is deep ultraviolet, or DUV, which uses 193-nanometer light and remains the workhorse for the vast majority of chips made in the world, including memory and the mature-node logic used in cars, appliances, and industrial equipment. The second family is extreme ultraviolet, or EUV, which uses light at a wavelength of 13.5 nanometers to print the smallest features on the most advanced logic and memory. EUV is where ASML stands alone. No competitor sells an EUV system, and none is close. The newest evolution of that line, High-NA EUV sold under the TWINSCAN EXE name, uses larger optics to push resolution further and is intended for the nodes that come after current production. The third leg of the business is installed base management, the servicing, upgrading, and supplying of the thousands of machines already in the field, which generates recurring revenue and tends to be steadier than the sale of new systems.

The economic engine that makes ASML so difficult to dislodge is the EUV monopoly and the supply chain that sits beneath it. An EUV machine is one of the most complex objects ever mass-produced. It generates light by firing a high-power laser at droplets of molten tin tens of thousands of times per second, then steers that light using mirrors polished to near-atomic smoothness inside a vacuum. ASML does not build most of this itself. Roughly 85 percent of the components in an EUV system come from outside suppliers, and the two most important are deeply tied to ASML. The optics come from the German firm Zeiss, in which ASML took a stake of about a quarter and invested heavily to fund High-NA development. The light source comes from Cymer, a US company ASML acquired in 2012 for around 2.5 billion dollars precisely to control that critical technology. The result is a single integrated system that took decades and the coordinated effort of many specialized firms to perfect, which is the real reason a rival cannot simply copy it. Money alone does not reproduce twenty years of accumulated process knowledge and a supplier network that ASML largely orchestrates.

The relationship with the largest customers reinforces the moat. In 2012, facing the enormous cost of finishing EUV, ASML ran a Customer Co-Investment Program that sold roughly 23 percent of the company to its three biggest buyers, Intel, Taiwan Semiconductor Manufacturing Company, and Samsung. Those customers funded the research and development that they themselves depended on, binding the leading chipmakers to ASML's roadmap. Those three remain the anchor customers today. TSMC and Samsung are the dominant foundries for leading-edge logic, Intel is rebuilding its own manufacturing while also pursuing a foundry business, and the memory makers, including SK Hynix and Micron, are increasingly large EUV buyers as advanced memory adopts the technology. The adoption pattern for High-NA shows how the customer base actually behaves. Intel took the first production High-NA tool in 2025 and has been the most aggressive early adopter, Samsung and SK Hynix have moved to bring the tools in, and TSMC has signaled a more cautious timeline, planning to use High-NA later in the decade rather than rushing it onto its next node. The differing schedules reflect a genuine debate about cost and timing, not any doubt that ASML controls the only path forward.

Competition exists but has not closed the gap in any segment that matters most. In conventional lithography, the Japanese firms Nikon and Canon still sell DUV and older systems, but their combined share of the equipment market by revenue is small, in the low single digits to mid single digits depending on the year, and neither offers EUV at all. In the broader race, the most watched challenger is China, where the state-backed Shanghai Micro Electronics Equipment and related spinoffs are trying to build domestic lithography tools. Chinese makers have made progress at older nodes and have publicized EUV-related patents, but they remain many years behind ASML on advanced systems, and replicating the full EUV supply chain under export restrictions is a far harder problem than building any single component. For the foreseeable future, the practical reality is that anyone who wants to make a cutting-edge chip has to buy from ASML.

Leadership reflects the engineering-heavy culture of the company. Christophe Fouquet, a French physicist who spent more than a decade rising through ASML and previously worked at the equipment makers KLA-Tencor and Applied Materials, became chief executive in 2024 after leading the EUV business. Roger Dassen serves as chief financial officer, and Marco Pieters, a long-tenured insider, took the chief technology officer seat and a board position in the mid-2020s. The Netherlands company is run as a two-tier board structure common to Dutch firms, with a management board and a supervisory board, and the executive team is dominated by people who came up inside the technology rather than from outside the industry. That continuity matters for a business whose central asset is institutional knowledge built over decades.

The forward strategy rests on three connected bets. The first is High-NA EUV, the next generation of machines that ASML needs the industry to adopt to keep extending the roadmap and to justify the multi-hundred-million-dollar price of each system. The second is artificial intelligence demand, which has become the dominant driver of leading-edge chip investment and therefore of demand for ASML's most profitable tools. The company has framed AI as a structural tailwind that supports a higher long-term revenue trajectory. The third is the steady expansion of the installed base service business, which grows mechanically as more machines are deployed and which smooths the company's results against the swings of new-equipment orders. Underneath all three is a roadmap that extends past High-NA toward even higher-resolution optics in the 2030s, the kind of multi-decade planning horizon that few companies attempt.

The risks are real and specific. The most prominent is geopolitics. ASML cannot sell its EUV systems to China at all because of US and Dutch export controls, and the restrictions have steadily widened to cover more of the advanced DUV systems that Chinese chipmakers buy for mature-node production. China had been one of ASML's largest markets, accounting for roughly a third of system sales in 2025, and that share has been falling sharply toward the low twenties or below as controls tighten and as a pull-forward of orders unwinds. The company is replacing much of that lost Chinese DUV revenue with higher-margin EUV demand from customers in Taiwan, Korea, and the United States, but the trajectory of export policy is set by governments, not by ASML, and further restrictions remain a live possibility. The second major risk is cyclicality. The semiconductor capital equipment industry moves in pronounced boom-and-bust cycles tied to chip demand and customer capacity decisions, and ASML's new-system revenue can swing meaningfully from year to year even when its monopoly is intact. The third is customer concentration, since a small number of leading-edge chipmakers account for a large share of the most valuable orders, which means a single customer delaying a node transition or cutting capital spending can move results. There is also the long-horizon question of whether the physical roadmap eventually slows, since lithography is pushing against the limits of optics and physics, though no near-term ceiling is in sight.

Weighing these factors, ASML presents an unusual profile for an investor to consider. It owns a genuine monopoly on the single most important tool in advanced semiconductor manufacturing, protected not by a patent that expires but by an accumulated base of knowledge and supplier relationships that has resisted replication for two decades. That durability is rare. The open question is not whether ASML can be displaced in the near term, which appears unlikely, but how two forces outside its control will balance over time. On one side is the secular growth in computing and artificial intelligence that keeps demand for leading-edge chips rising, and with it demand for ASML's machines. On the other is the steady tightening of export policy that walls off a large market and the inherent cyclicality of an industry that buys billion-dollar tools in waves. How those two forces resolve, rather than any threat to the underlying monopoly, is the central thing to watch.