Is ASML Overvalued? Pricing a Monopoly With a Finite Runway

ASML designs and manufactures the photolithography systems that chipmakers use to etch circuit patterns onto silicon wafers. The company operates in two technology bands. Extreme ultraviolet machines, known as EUV, use 13.5nm wavelength light to print features at the most advanced nodes, currently below 5nm. Deep ultraviolet machines, or DUV, serve a broader market of legacy and mature-node chipmakers who do not need the latest process technology.

The EUV monopoly is the part that matters for the valuation discussion. TSMC, Samsung, and Intel cannot make leading-edge chips without ASML machines. There is no alternative supplier, no substitute technology ready to scale, and no credible timeline for a competitor to close the gap. The R&D investment required to reach EUV capability has been estimated in the tens of billions of euros over decades.

DUV is a different story. ASML has dominant share here too, but Chinese domestic equipment makers have been investing aggressively in DUV alternatives, partly as a result of export controls cutting off ASML's China business. This matters because China represented a meaningful portion of ASML's revenue mix, and that revenue is now subject to political risk that did not exist five years ago.

The semiconductor equipment industry is also cyclical. Chipmakers spend heavily on capacity during expansion cycles and sharply reduce equipment orders during corrections. ASML's backlog provides some insulation, but the underlying demand for machines still tracks foundry capex, which tracks semiconductor demand, which tracks the broader economy and the AI infrastructure build cycle.

ASML's five-year financial progression is clean and impressive. Revenue grew from $18.6 billion in 2021 to $32.7 billion in 2025, a compound annual growth rate of approximately 15%. Gross margins have been remarkably stable, hovering in the 50.5% to 52.8% range throughout the period. Operating margins moved between 30.7% and 36.3%, reflecting the operating leverage in ASML's model as revenue scales.

Net income grew from $5.9 billion in 2021 to $9.6 billion in 2025. That is not explosive growth, but it is consistent and high-quality earnings from a business that requires minimal ongoing capital relative to its revenue.

Free cash flow tells a nuanced story. In 2023, FCF dropped to $3.2 billion from $7.2 billion the prior year, driven by a capex increase and changes in working capital as the company prepared for higher production volumes. It recovered to $9.5 billion in 2024 and $10.6 billion in 2025. The recovery matters because it confirms the earnings are real and converting to cash.

The balance sheet is conservative. As of year-end 2025, ASML held $12.9 billion in cash against $2.7 billion in long-term debt. Net cash of over $10 billion provides meaningful cushion and eliminates financing risk. For a company with $9.6 billion in net income, this is a pristine balance sheet.

Capital returns have been active. ASML returned $5.7 billion via buybacks in 2025, the largest single-year return in the five-year window, and paid $2.4 billion in dividends. Total shareholder return of $8.1 billion represents approximately 84% of net income. Share count has declined modestly from 0.42 billion in 2020 to 0.39 billion in 2025, a reduction of roughly 7%.

The EUV moat is not a marketing claim. It is an engineering reality that took 20 years to build. EUV machines produce 13.5nm wavelength light by firing a tin droplet 50,000 times per second with a high-powered laser, converting it to plasma that emits extreme ultraviolet radiation. The system then focuses that light through 11 mirrors polished to angstrom-level tolerances.

The entire machine is roughly the size of a bus, costs over $150 million per unit, and requires components from hundreds of suppliers across Europe, Japan, and the United States. No single company other than ASML has assembled this supply chain or built the internal expertise to produce EUV at yield.

TSMC has reported taking 10 or more years of collaboration with ASML to achieve reliable EUV production. That timeline is not available to a new entrant today. The result is that ASML's largest customers have no credible threat to walk away. When TSMC builds a new fab, it orders ASML machines. When Samsung wants to compete at leading-edge nodes, it orders ASML machines. Intel does the same. There is no workaround.

High-NA EUV, the next generation of the technology, extends this moat further. ASML is the only company shipping High-NA systems, which are required for gate-all-around transistor architectures expected to dominate in the second half of this decade. Each High-NA machine carries a price tag in excess of $350 million. ASML has already delivered the first units to Intel and TSMC for development work. The transition to High-NA is not a risk to ASML's position. It is an extension and a price step-up.

At 45.7x trailing earnings, ASML is priced for sustained above-average growth. The question is what growth rate is required to justify that multiple and whether the addressable market can deliver it.

Working backwards from the current price: a $511 billion market cap against $9.6 billion in net income implies the market expects net income to grow to at least $20 to $25 billion within seven to ten years, assuming any reversion toward a more normalized 20 to 25x earnings multiple. That requires sustained double-digit earnings growth compounding for close to a decade.

ASML's own guidance for 2030 calls for revenue in the range of 44 to 60 billion euros, which at current exchange rates is approximately $48 to $66 billion. If gross margins hold at 52% and operating margins reach the high end of their historical range, that implies operating income of roughly $17 to $23 billion and net income in the $14 to $19 billion range. The midpoint of that range supports the valuation. The low end does not.

The price-to-sales ratio of 15.7x compounds the concern. For context, a 15x price-to-sales multiple on a technology company typically implies either very high margins or very high growth, ideally both. ASML has the margins but not the growth rate. Revenue has grown at roughly 15% annually over five years. Sustaining that against a larger base is harder, particularly with China revenue impaired by export controls.

The EV/EBITDA of 33.9x is a similar story. ASML generated $12.6 billion in EBITDA in 2025. At a normalized 20x multiple, that is $252 billion in enterprise value. At 33.9x, the market is paying $427 billion. The gap represents embedded expectations for EBITDA growth toward $20 billion or more.

ASML's most important forward indicator is its order backlog. The company does not disclose the backlog in granular detail, but management commentary and industry analysis suggest it remains above 30 billion euros as of early 2026. That implies more than one year of revenue at current run rates is already contracted.

The backlog is predominantly driven by AI-related semiconductor demand. TSMC, the primary beneficiary of AI chip orders from Nvidia, Apple, and AMD, has announced aggressive capacity expansion plans through 2027 and 2028. Those fabs require EUV machines, and the lead times for EUV delivery run 12 to 18 months. Orders placed today show up in revenue 18 months from now.

High-NA EUV is an additional driver that is not yet material to revenue but will become so. The per-unit price is more than double that of standard EUV. If adoption follows the pattern of standard EUV, where uptake was slow initially before becoming mandatory for leading-edge production, High-NA could add several billion euros to annual revenue by 2028 or 2029.

DUV revenue is more complicated. China historically represented 25 to 30% of ASML's DUV revenue. Export controls have curtailed new system sales to China, though spare parts and service revenue remain permitted. The loss of incremental China DUV business reduces growth potential in the near term. Any further escalation in export restrictions could accelerate that headwind.

Replacement demand from other regions, including Japan, Southeast Asia, and Europe through CHIPS Act-funded expansions, can partially offset what China represented. Whether the offset is full or partial depends on the scale of domestic chipmaking investment in each region, which is still early-stage in Europe and Southeast Asia.

ASML's customers are not passive. TSMC, Samsung, and Intel are three of the most sophisticated engineering organizations on the planet. They do not accept monopoly pricing passively. Each of them has at some point invested in alternative lithography research, and each ultimately concluded that EUV from ASML was the only viable path.

That pattern is likely to repeat. TSMC and Samsung will push back on pricing at contract negotiation. They will demand service level commitments and uptime guarantees. They will use their combined buying power to extract favorable terms. But they will keep buying. The alternative is falling behind on process node, which means losing competitive position in the chips that matter most.

The greater competitive pressure is at the DUV level, where Chinese alternatives are developing capability. SMEE, the Shanghai-based lithography company, has been working toward 28nm DUV capability for several years. It has not shipped at volume, and 28nm is not where the leading-edge economics are won. But if SMEE achieves yield at mature nodes over the next three to five years, it could displace ASML in the segment of the market that China's domestic chipmakers serve.

For ASML's overall revenue, the DUV China risk is meaningful but contained. The EUV business has never been available to Chinese chipmakers and is not affected by incremental SMEE progress. TSMC concentration is a different issue. TSMC accounts for a majority of EUV revenue. If TSMC's investment cycle moderates sharply, ASML's order flow moderates with it. The two companies are bound together in a way that creates mutual dependency but also mutual exposure to the same cyclical forces.