ASML Holding N.V. (NASDAQ: ASML) — The Only Toll Booth to AI, at a Record-High Toll

Ticker: ASMLF (OTC, U.S. grey-market line for the Euronext Amsterdam ordinary shares) — economically identical 1:1 to NASDAQ/Euronext ASML Company: ASML Holding N.V. — Veldhoven, the Netherlands Sector / Industry: Technology — Semiconductor Capital Equipment (Wafer-Fab Equipment / Lithography) Report date: 2026-06-09 · Currency: Reports in EUR under U.S. GAAP; quoted price in USD Primary sources: FY2025 Form 20-F (filed 2026-02-25); Q1 2026 6-K (filed 2026-04-15); Q1 2026 earnings call; SEC EDGAR CIK 0000937966

⚡ Claude’s Take

This block is the author’s own independent opinion and general information only — not investment advice. The analytical body of this article (Sections 1–15) takes no position, sets no price target, and carries no recommendation; the single subjective view is fenced inside this block.

Verdict: HOLD the business, AVOID the entry — a genuinely great company at a price that already pays for greatness. Accumulate only on a 25–35% drawdown. Conviction: medium-high. Tag: “The only toll booth on the road to AI — but the toll on owning the toll booth is near record-high.”

ASML is one of the highest-quality businesses in the public market: the sole maker of EUV lithography, the single indispensable tool for every leading-edge logic and advanced-DRAM chip on earth. The moat is the strongest configuration in the Greenwald taxonomy — economies of scale plus extreme customer captivity plus genuinely proprietary, decade-and-multi-billion-euro technology (sole-source Zeiss optics it part-owns, its own Cymer light source). ROE ~50%, ~53% gross margins, €11B of free cash flow, net cash, and a customer base (TSMC, Samsung, SK Hynix, Intel, Micron) that pre-pays for capacity. The AI capex super-cycle runs straight through ASML’s order book; management says demand is outpacing supply and customers are “sold out.” There is no competitive threat to underwrite here. The whole debate is price and cycle, not quality.

And that is the rub. On its own ~10-year history the stock sits at the 96th percentile of valuation (P/E ~58x trailing, ~17–18x EV/sales, P/B 28x) and trades above the average sell-side target. You are buying a cyclical-growth monopoly at a trailing high-50s multiple, with ~29% of revenue still routed through China under an active export-control overhang and ~61% of revenue concentrated in four customers whose capex is itself cyclical. The market is underwriting the full 2030 Investor-Day dream (€44–60B sales, 56–60% gross margin) with little discount for an air-pocket. History says semicap demand is lumpy; a single guide-down or a China escalation re-rates a 58x multiple violently. What flips me bullish: a meaningful drawdown (toward a high-30s/low-40s forward P/E) or hard evidence the 2030 model is being pulled forward. What flips me bearish: a durable AI-capex digestion phase or export rules that strand the China service base. Great house; the street is just very, very expensive right now.

1. Executive Summary

ASML Holding N.V. is the world’s monopoly supplier of extreme-ultraviolet (EUV) lithography systems and the dominant supplier of deep-ultraviolet (DUV) immersion systems. Lithography — printing circuit patterns onto silicon — is the single most critical and most expensive step in semiconductor manufacturing, and ASML is the only company on Earth that can build the machines required to make the most advanced chips. No leading-edge logic node (3nm, 2nm and below) or advanced DRAM can be manufactured at volume without ASML’s EUV tools. This is as close to a true, sole-source monopoly on an enabling technology as exists in any major industry.

The financial signature of that position is unambiguous. In FY2025 ASML generated €32.7 billion of revenue (+15.6%), a 52.8% gross margin, €11.3 billion of operating income (34.6% margin), €9.6 billion of net income (29.4% margin), and ~€11.0 billion of free cash flow — on only €1.6 billion of capital expenditure. Return on equity runs near 50%. The balance sheet holds net cash. One-quarter of revenue (€8.2 billion) is high-margin, annuity-like Installed Base Management (service + upgrades) from a growing installed base — a recurring stream that compounds as more tools ship and that cushions the cyclicality of new-system sales.

The business is riding a powerful tailwind. Artificial-intelligence infrastructure investment is driving leading-edge logic and high-bandwidth memory demand to the point where, per management, “demand will continue to outpace supply” and memory customers are “sold out for the remainder of the year.” ASML raised its FY2026 revenue guidance to €36–40 billion at a 51–53% gross margin. High-NA EUV — the next monopoly-extending platform — shipped its first high-volume unit in April 2025 and ramps to production in 2027.

The debate, therefore, is not about business quality. It is about price and cyclicality. The stock trades near the 96th percentile of its own ten-year valuation history (trailing P/E ~58x, EV/sales ~17–18x, P/B ~28x), at a market capitalization of roughly $685 billion, above the average analyst price target. Embedded in that price is sustained double-digit growth toward the company’s 2030 model, durable monopoly margins, and no material disruption from the China export-control regime that still touches ~29% of revenue. Customer concentration is extreme (four customers = 61% of sales). The risk here is not that the moat erodes — it almost certainly will not on any relevant horizon — but that a cyclical demand air-pocket or geopolitical shock meets a very high multiple. This memo takes no position on the security; it lays out the monopoly, the economics, and the expectations the price now embeds.

2. Business Overview

ASML designs, manufactures, markets, and services photolithography systems for the semiconductor industry, plus the associated metrology/inspection, computational-lithography software, and lifecycle services. Lithography is the patterning step in chip fabrication: light is projected through a mask (reticle) onto a photoresist-coated silicon wafer to define the transistor and interconnect features of an integrated circuit. The smaller the feature ASML’s tools can resolve, the more advanced the chip a customer can build. ASML does not make chips; it makes — and is, for EUV, the only company that makes — the machines that make the chips.

Why this matters so disproportionately: chip advancement (Moore’s Law) is fundamentally a story of printing ever-smaller features, and feature size is governed by the wavelength of light used to print it. The industry spent two decades stretching 193nm DUV light to its physical limits (via immersion and multi-patterning) before the only way forward became EUV at 13.5nm — a fourteen-fold reduction in wavelength that required generating light by vaporizing tin droplets into plasma, reflecting it off mirrors polished to near-atomic flatness (no lens can transmit EUV), and doing it all in a vacuum with nanometer precision. ASML is the only company that solved this, after a development effort spanning more than two decades and well over €10 billion of cumulative investment, in partnership with Zeiss (optics) and its own Cymer subsidiary (light source). The result is that the entire global advance of computing — every leading-edge AI chip, every flagship smartphone processor, every advanced DRAM die — now passes through a tool that exactly one company on the planet can build. That is the business in one sentence.

Product lines (FACT, per FY2025 20-F):

• EUV lithography (TWINSCAN NXE / EXE platforms). Uses 13.5nm extreme-ultraviolet light to print the smallest features on leading-edge logic and advanced DRAM. ASML is the world’s only manufacturer of EUV systems — a literal monopoly. “Low-NA” EUV (NXE, 0.33 numerical aperture) is in high-volume manufacturing across all leading-edge fabs; “High-NA” EUV (EXE, 0.55 NA) is the next generation, the first HVM-ready EXE:5200B having shipped in April 2025, with high-volume manufacturing expected from 2027. Each low-NA EUV tool sells for roughly €180–220 million; High-NA tools for roughly €350–400 million.
• DUV lithography (TWINSCAN immersion and dry). Deep-ultraviolet (193nm) systems for mature and mid-critical layers. ASML is dominant (~80–90% of immersion) but here it competes with Nikon and Canon. DUV remains a large, profitable franchise and is the principal product sold into China.
• Metrology & inspection. YieldStar optical metrology and HMI e-beam systems that measure pattern quality and locate defects, increasingly bundled with the lithography roadmap (the “holistic lithography” strategy).
• Installed Base Management (IBM) — service & field options. Maintenance, spare parts, and performance/productivity upgrades sold across the multi-thousand-unit installed base. This is the recurring, high-margin annuity of the business.

Revenue model and segmentation (FACT, FY2025, EUR):

System sales are large, lumpy, and capital-cycle-sensitive (recognized on shipment/acceptance of individual machines). IBM is recurring and grows structurally with the installed base — it rose 26% in 2025 and is guided to “grow significantly” in 2026. In Q1 2026, system sales split roughly 49% Logic / 51% Memory, a reminder that ASML sells into both major end-markets and is diversified across the chip industry rather than levered to a single application.

Customers and end-markets. ASML’s customers are the world’s chipmakers: foundries (TSMC), integrated device manufacturers and logic players (Intel, Samsung), and memory makers (Samsung, SK Hynix, Micron). End-demand is downstream — AI accelerators, smartphones, PCs, data-center CPUs, automotive, and industrial chips. ASML therefore sells “picks and shovels” to the entire semiconductor gold rush rather than betting on any single chip architecture.

How the machine makes money — the unit economics. A single low-NA EUV scanner (NXE) carries an ASP of roughly €180–220 million and weighs ~180 tonnes, shipping in multiple freight aircraft and taking months to install and qualify. A High-NA EXE system roughly doubles that to ~€350–400 million. In FY2025 ASML recognized 48 EUV units (44 in 2024) — so a relatively small number of very large transactions drives system revenue, which is precisely why quarterly system sales are lumpy and why a handful of customer decisions swing a quarter. The offset is volume in DUV (hundreds of units a year at lower ASPs) and the IBM annuity. Critically, revenue is recognized largely on shipment/acceptance, while customers pay substantial down-payments in advance (reflected in the €3.6B+ of contract liabilities) — the customer funds ASML’s working capital while the machine is built, a structural sign of pricing power that few capital-goods businesses enjoy.

Holistic lithography — the bundling strategy. ASML does not sell a scanner in isolation; it sells a patterning system: the scanner plus YieldStar metrology (to measure pattern fidelity), HMI e-beam inspection (to find defects), and computational-lithography software (to optimize the mask and process). This bundle raises the value delivered per litho layer and deepens lock-in — the customer’s yield is co-engineered with ASML’s tools and software, making the relationship far stickier than a one-off equipment sale. It also extends ASML’s revenue per wafer beyond the scanner itself.

Logic vs. Memory balance. In Q1 2026 system sales split ~49% Logic / 51% Memory — a useful illustration that ASML is not a one-end-market story. Logic (TSMC, Intel, Samsung foundry) buys EUV for leading-edge nodes; Memory (Samsung, SK Hynix, Micron) buys EUV for advanced DRAM and, increasingly, for the high-bandwidth memory (HBM) that AI accelerators require. When one end-market digests, the other often offsets — a natural diversification within the semiconductor complex that smooths (without eliminating) the cycle.

Verdict: A focused, sole-source supplier of the single most critical tool in semiconductor manufacturing, with a growing recurring-service annuity layered on top of lumpy but extraordinarily profitable system sales. The business model is structurally advantaged: it monetizes the entire industry’s race to smaller nodes, not any one customer’s success — and it does so on customer-funded working capital, with a bundled software/metrology layer that compounds the lock-in.

3. Industry Dynamics

Structure. Semiconductor capital equipment (wafer-fab equipment, “WFE”) is an oligopoly of a handful of giants, each dominant in its own process step: Applied Materials and Lam Research (deposition/etch), KLA (process control), Tokyo Electron (coat/develop, etch), and ASML (lithography). Within lithography specifically, the structure is the most concentrated of all: ASML is a monopolist in EUV and the dominant player in DUV immersion, with Nikon and Canon relegated to older DUV and niche segments. Lithography is also the most valuable step — the EUV tool is the single most expensive machine in a fab — so ASML captures a disproportionate share of the WFE profit pool.

Market size and growth. WFE is a multi-tens-of-billions market growing structurally with three secular drivers, each of which is genuinely independent of the others:

1. AI infrastructure (the demand accelerant). AI accelerators (GPUs, custom ASICs) are built on leading-edge logic, and they require enormous quantities of HBM — stacks of advanced DRAM. Both are the most lithography-intensive products in the industry. The AI build-out is therefore a direct multiplier on EUV demand, and management’s language is unusually emphatic: demand “outpaces supply,” memory customers are “sold out for the remainder of the year,” and supply limitation is expected “beyond 2026.”
2. Node migration / rising litho intensity (the structural escalator). Each new logic node (5nm→3nm→2nm→below) and each DRAM generation adds EUV layers; High-NA then adds another tier. Litho spend per wafer rises with every node independent of unit growth — even a flat number of wafers started would consume more ASML content as those wafers move to advanced nodes. This is the quiet compounder beneath the cyclical headlines.
3. The installed-base annuity (the de-cyclicalizer). Every system shipped becomes a decade-plus stream of service and upgrade revenue. IBM grew 26% to €8.2B in 2025 and compounds mechanically as the EUV fleet grows. Over time this turns an increasing share of ASML’s revenue into a recurring, non-cyclical base.

Management’s framing (Q1 2026): the semiconductor industry growth outlook “continues to solidify, driven primarily by AI-related infrastructure investment,” with demand outpacing supply across logic and memory “beyond 2026.” ASML’s own 2030 model (2024 Investor Day) targets €44–60 billion of revenue at a 56–60% gross margin — implying roughly a doubling of revenue from the 2024 base and a structurally higher margin, driven by EUV/High-NA mix and the IBM annuity. That model is both the bull thesis and, as Section 10 shows, the figure the market has substantially already capitalized.

Barriers to entry (Greenwald). EUV is the textbook case of insurmountable barriers, combining all three genuine advantage types:

• Proprietary technology / supply advantage: EUV required ~three decades and well over €10 billion of cumulative R&D, plus a global supply chain (Zeiss optics, Cymer light source, tens of thousands of components) that no entrant can assemble. ASML owns Cymer and holds a 24.9% equity stake in Carl Zeiss SMT, the sole supplier of EUV optics — vertical control of the two hardest sub-systems.
• Economies of scale: ASML’s €4.7 billion annual R&D budget (14% of sales) exceeds the entire equipment R&D of any would-be competitor. Scale in R&D and in the installed base is self-reinforcing: more tools shipped → more service revenue and field data → faster roadmap → more tools shipped.
• Customer captivity: Once a fab is built around ASML tools, switching is impossible — the entire process is co-developed with ASML, and the installed base locks in a decade of service/upgrade dependence.

Regulation. The defining external factor is export control. EUV has been effectively banned for export to China since 2019 (via Dutch licensing aligned to Wassenaar/U.S. pressure); advanced DUV immersion restrictions to China tightened from 2024. This is a double-edged sword: it constrains ASML’s China TAM but also reinforces the strategic, quasi-national-asset character of the franchise. Management explicitly flags that 2026 guidance bandwidth “accommodates potential outcomes of ongoing discussions around export control.”

Capital-cycle read (Marathon). The industry is in a capital-cycle boom — customers are “aggressively adding capacity,” raising capex, and pre-committing via long-term agreements. Normally a boom invites supply, erodes returns, and mean-reverts. The critical distinction: ASML is the toll-collector, not a commodity capacity-adder. It does not face new entrants; its monopoly suspends the mean reversion that capital inflows would otherwise cause. The capital-cycle risk for ASML is therefore not competitive erosion but its customers’ over-building — a downstream capex digestion that would show up as a lumpy, cyclical air-pocket in ASML’s system orders. That is the cyclicality embedded in this stock.

Profit-pool concentration. Within a fab’s equipment budget, lithography is not merely one line item among many — it is the most expensive single category and, because ASML is sole-source in EUV, the one with the least price elasticity. As nodes advance, litho’s share of WFE spend rises (more EUV layers, then High-NA layers), so ASML captures a growing slice of a growing pie. This is the structural reason ASML’s margins expand with scale while a deposition or etch peer, facing two strong competitors, must share its profit pool. The “agency”-like dynamic Marathon describes also applies: the chipmaker’s customers (Apple, Nvidia, the hyperscalers) effectively demand leading-edge nodes, and leading-edge nodes require ASML — so the buyer cannot substitute away even if it wanted to. Pricing power flows from indispensability, not negotiation.

Regulatory timeline (the China overhang in detail). EUV exports to China have been effectively blocked since 2019 (Dutch export-licensing aligned with Wassenaar and U.S. pressure); from 2024, the most advanced DUV immersion tools also require licenses for China. The practical effect: China bought heavily in 2023–2024 (DUV pull-forward ahead of and around the restrictions, lifting China to 36% of 2024 sales), then normalized to 29% in 2025 as that backlog shipped. The forward risk is asymmetric and two-sided — tighter rules could curtail remaining DUV-to-China and, in a severe case, the China service base; retaliation (China restricting rare-earth/gallium inputs, or favoring domestic SMEE litho) is a tail. Management’s decision to size the 2026 guidance band (€36–40B, a €4B spread) explicitly around “ongoing discussions around export control” tells you how live this variable is.

Verdict: Structurally excellent — among the best industry positions in global equities. Lithography is the most concentrated, highest-value, highest-barrier node of an oligopolistic, secularly-growing industry, and ASML owns its monopoly tier outright, capturing a rising share of the profit pool. The only structural caveat is that ASML’s fortunes ride on its customers’ capex cycle, which is genuinely cyclical even if the long-run trend is up and to the right.

4. Competitive Position

The moat, named. ASML’s competitive advantage is the strongest configuration in the Greenwald framework — economies of scale combined with customer captivity, reinforced by genuinely proprietary technology — applied to a monopoly product. In EUV there is no second source. This is not “differentiation” (which Greenwald shows does not protect profits); it is a structural barrier that competitors cannot scale, because the technology, the supply chain, the patents, the learning curve, and the installed-base data all compound in ASML’s favor and cannot be replicated on any commercial timeframe.

Mechanism, sub-system by sub-system:

• Optics: EUV mirrors must be polished to atomic-scale flatness; Carl Zeiss SMT is the sole supplier, and ASML owns 24.9% of it — a deliberate lock on the single hardest input.
• Light source: EUV light is generated by vaporizing tin droplets with a high-power laser ~50,000 times per second. ASML acquired Cymer (2013) to own this. The roadmap (1,000-watt source demonstrated; path to 330 wafers/hour at the start of the next decade) is itself a moat-deepening exercise.
• Systems integration: An EUV tool has ~100,000 parts and a supply chain of thousands of specialized vendors orchestrated only by ASML.
• Installed-base lock-in: Each tool ties the customer to a decade of ASML service and upgrades; the “holistic lithography” bundle (metrology + computational litho + scanner) raises switching costs further.

The market-share and ROIC tests (Greenwald). Market share in EUV is 100% and has been since the product existed — the most stable share imaginable (Greenwald’s test: <2-point change over 5–8 years = formidable barriers; ASML has zero share loss). Sustained after-tax returns on capital run far above the 15–25% threshold that signals real advantage — ROE ~50%, ROIC in the 30%+ range. Both tests are passed decisively, and the source is identifiable and durable, which is exactly the condition under which Greenwald says high returns are not merely temporary or good luck.

Where the moat is narrower. In DUV, ASML is dominant but not sole — Nikon and Canon compete in dry and older immersion. DUV is a very good business but not a monopoly, and it is the segment most exposed to China and to commoditization at mature nodes. The monopoly economics are concentrated in EUV (and, prospectively, High-NA).

Direct comparison vs. competitors. Nikon and Canon abandoned the EUV race a decade ago; neither can credibly re-enter. Their reasons are instructive about the moat’s depth: EUV required betting billions over twenty-plus years with no certainty of a working product, mastering tin-plasma light sources and atomic-flatness optics, and orchestrating a supply chain that did not exist. Nikon and Canon — both profitable, competent optics companies — concluded the investment was uneconomic, and were proven right that only one firm could amortize that spend across the whole industry’s demand. That is the essence of an economies-of-scale-plus-captivity moat: the market supports exactly one supplier, and ASML is it. Within WFE, peers like Applied Materials, Lam, KLA, and Tokyo Electron are excellent businesses but operate in their own steps and do not compete with ASML in lithography. There is, simply, no competitor to ASML’s core product. The “competition” that matters is downstream: whether ASML’s customers keep investing.

China’s domestic-litho effort (SMEE) as the only entry attempt. The one place an entrant is being willed into existence is China, where SMEE and state-backed programs are attempting domestic DUV (and, aspirationally, EUV) lithography under sanctions pressure. This is the Marathon “state-capitalism breaks the capital cycle” scenario — capital deployed irrespective of returns. Realistically, SMEE is years behind in DUV and effectively nowhere on EUV (which needs the very Zeiss optics and Cymer sources that are export-controlled). It is a long-dated geopolitical risk to ASML’s China DUV revenue specifically, not a threat to the EUV monopoly on any foreseeable horizon. But it is the reason to watch China policy as a strategic variable, not just a near-term revenue one.

Verdict: A durable, identifiable, sole-source monopoly — one of the widest moats in the public markets. The advantage is not theoretical; it shows up in 100% EUV share, ~50% ROE, ~53% gross margins, and customers who pre-pay. The honest caveat is that the depth of the moat (EUV) is narrower in scope than the breadth of the revenue (DUV + service), and the franchise’s growth is hostage to its customers’ capex decisions, not to any competitive threat.

5. Growth History and Forward Opportunities

Historical growth (FACT, EUR): Revenue grew from €27.6B (2023) to €28.3B (2024, +2.6%) to €32.7B (2025, +15.6%). The 2024 near-plateau reflected a digestion year (post-2023 China DUV pull-forward and a memory downturn); 2025 reaccelerated on AI-driven leading-edge logic and recovering memory. The longer arc is steep: ASML has compounded revenue at a low-teens rate across the past decade as EUV moved from introduction to the backbone of leading-edge manufacturing.

Composition of 2025 growth (FACT, per 20-F MD&A): Of the €4.4B revenue increase, Logic net sales rose ~€2.9B, “primarily driven by leading-edge foundry growth in support of strong AI demand”; Memory was roughly flat-to-slightly-down; and IBM (service + upgrades) grew 26% to €8.2B. Growth is overwhelmingly organic — ASML does not buy revenue; its few acquisitions (Cymer, HMI, Berliner Glas) were capability deals, not growth deals.

Forward opportunities:

1. AI capacity build-out (near term). Management’s clearest signal: demand outpacing supply, customers “sold out,” capex being raised and accelerated, “supply limitation across advanced nodes beyond 2026.” FY2026 revenue guided up to €36–40B (a ~16% jump at the midpoint over 2025), H2-weighted.
2. EUV layer intensity (structural). Each new logic node (2nm and below) and each advanced DRAM generation adds EUV layers, increasing lithography spend per wafer independent of unit growth.
3. High-NA EUV (the next monopoly). The EXE platform extends the EUV monopoly into the 2027+ era at ~2x the ASP of low-NA tools, with HVM from 2027 and a roadmap to the next decade.
4. Installed Base Management (the compounding annuity). As the EUV fleet grows, service and upgrade revenue compounds at high margin — guided to “grow significantly” in 2026 and structurally the steadiest growth vector.
5. The 2030 model. ASML’s 2024 Investor Day targets €44–60B revenue and 56–60% gross margin by 2030 — the explicit basis for the market’s long-run expectations.

Decomposing the 2030 bridge. The €44–60B 2030 target (from ~€28B in 2024) implies a roughly 8–16% revenue CAGR — a wide band that itself signals how much hinges on AI-capex durability. The bridge has three identifiable planks: (1) EUV unit growth and ASP — low-NA EUV output scaling from ~48 units (2025) toward the 80+ capacity (2027) and beyond, layered with High-NA at ~2x ASP; (2) DUV resilience — guided to grow in 2026 (previously expected flat) as customers add immersion capacity for trailing-and-mid nodes; and (3) IBM — guided to grow “significantly,” compounding off the installed base. The gross-margin path to 56–60% rests on EUV/High-NA maturation (lower cost-to-serve as platforms standardize on the common NXE/EXE architecture) and the high-margin upgrade business. None of these planks requires share gains — ASML already has 100% of EUV; they require the industry to keep investing, which is the cyclical bet.

EUV unit trajectory as the leading indicator. EUV units recognized rose 44 (2024) → 48 (2025), with the 2026–2027 output plan stepping to ≥60 then ≥80 low-NA systems. Because each unit is ~€180–220M (more for High-NA) and carries a decade of attached service, the EUV unit count is the cleanest forward gauge of both system revenue and the future IBM annuity. Watching units shipped and the order-to-ship cadence is more informative than any single headline revenue print.

Quality of growth. This is high-quality growth: organic, funded by internal cash flow (not dilution or debt), at expanding margins, and underpinned by a monopoly rather than by share gains that could reverse. The principal qualification is cyclicality — the path is lumpy, and a customer capex pause (as in 2024) can stall a year even within a secular uptrend. Reported order intake is “very strong,” but ASML’s bookings are notoriously volatile quarter-to-quarter and should be read as a trend, not a metronome — a single strong or weak bookings quarter routinely swings sentiment far more than it should.

Verdict: High-quality, monopoly-underpinned, organically-funded growth — but cyclical in cadence. The secular drivers (AI, node migration, High-NA, service annuity) are real and largely supply-constrained at the customer level. The risk to the growth thesis is timing and digestion, not durability.

6. Financial Quality

ASML’s financial statements are among the cleanest and most powerful in large-cap technology. Every figure below is FACT from the FY2025 20-F or Q1 2026 6-K (EUR, U.S. GAAP).

Profitability and margins.

Margins are not only high but expanding with scale — the single most important test of a real moat (Greenwald: economics must improve with scale). Gross margin rose ~150bps in 2025 on mix (more EUV, more high-margin IBM/upgrades), and operating leverage carried it further as R&D and SG&A grew slower than revenue. Q1 2026 gross margin hit 53% specifically because of “very high-margin components within our installed base business” — i.e., the annuity is the margin engine, not just the volume engine.

Cash generation. Free cash flow (company definition) was €9.1B (2024) and €11.0B (2025) — a ~95% conversion of net income, on capex of only €1.6B (capex/sales ~5%). This is a capital-light monopoly: it spends little to produce enormous cash. The caveat is quarterly lumpiness driven by customer down-payments: Q4 2025 operating cash flow was €11.4B (a huge prepayment inflow), while Q1 2026 was negative €2.2B (FCF −€2.6B) as those down-payment timings reversed. Annual FCF is the right lens; quarterly FCF is noise.

Working capital and the prepayment signal. Inventories are large (€11.7B) — the cost of building €350M machines with year-plus cycle times — but the offsetting feature is contract liabilities (customer prepayments): €3.6B non-current plus a large current portion. Customers pre-fund ASML’s working capital. A business whose customers pay in advance for the privilege of a place in the queue is exhibiting textbook pricing power.

Balance sheet (Mar 29, 2026). Total assets €48.1B; total equity €20.8B; long-term debt €2.7B against €8.4B cash + short-term investments → net cash of ~€5.7B. (Net cash was ~€10.6B at Dec-2025; the Q1 decline is the down-payment-timing swing plus €1.6B of dividends/buybacks, not deterioration.) Goodwill is €4.6B (legacy Cymer/HMI/Berliner Glas) — modest relative to equity, so book value is well-supported by tangible and financial assets. Leverage is negligible; liquidity is fortress-grade.

Returns on capital. ROE ~50% (NI €9.6B / ~€19–20B average equity). ROIC is similarly extraordinary given the small net-debt and high EBIT (€11.3B). These are franchise returns, sustained over a decade — the Greenwald signature of genuine, identifiable competitive advantage rather than transient good fortune.

Quality-of-earnings checks. Net income tracks cash flow closely on an annual basis (no divergence flag). Share-based compensation is small (~€0.2B/yr, ~0.6% of sales) — negligible dilution; the share count is falling via buybacks, not rising via SBC. Equity-method profit (€217M in 2025, mostly the Zeiss SMT stake) is real economic income, not a manufactured gain. Tax rate is a clean ~17%. There are no obvious one-time distortions inflating the run-rate (unlike, e.g., bank one-offs); the only normalization needed is to smooth quarterly cash flow for down-payment timing.

Multi-year trend read. Three patterns stand out across 2023→2025. First, operating leverage is real: revenue grew 18.5% cumulatively while operating income grew far faster (op margin 31.9%→34.6% in the last year alone), because R&D (+9.2%) and SG&A (+7.9%) grew well below revenue (+15.6%). Second, mix is steadily enriching margins — IBM/service revenue (the highest-margin line) grew 26% and now funds the gross-margin step-up; within systems, EUV (higher value-add) is displacing DUV. Third, the recurring layer is de-risking the model over time: IBM rose from €6.5B (2024) to €8.2B (2025) and is guided “significantly” higher in 2026; as the installed base of EUV tools compounds, this annuity grows structurally regardless of the new-system cycle, meaning a larger and larger share of revenue is non-cyclical.

ROIC, computed. With FY2025 EBIT of €11.3B, a ~17% tax rate (NOPAT ~€9.4B), and invested capital of roughly €15B (equity €20.8B less net cash €5.7B, plus modest debt), after-tax ROIC sits comfortably above 30% — far beyond the 15–25% threshold Greenwald uses to confirm genuine competitive advantage, and sustained for a decade. This is not a business earning a market return on capital; it earns a monopoly return.

Working-capital deep-dive. The €11.7B inventory balance looks heavy until set against the build cycle: ASML carries long-lead components and work-in-process for machines that take a year-plus to assemble and that sell for hundreds of millions each. The crucial offset is negative working-capital funding from customers — contract liabilities (down-payments) of €3.6B non-current plus a large current portion mean buyers finance the build. The net effect is that ASML’s enormous inventory is substantially pre-sold and pre-funded; the inventory is a sign of a full order book, not of stranded product. The risk to watch is the reverse: if orders were cancelled or pushed out, that inventory and those prepayment dynamics would unwind unfavorably — which is exactly why order momentum is the key monitorable (Section 13).

Peer cross-read. Against the WFE oligopoly and the broader semiconductor complex, ASML’s economics stand out specifically because of the monopoly. Best-in-class WFE peers (Applied Materials, Lam, KLA, Tokyo Electron) earn excellent but competed returns — gross margins in the high-40s to high-50s and operating margins generally in the high-20s to low-30s, sharing their profit pools with one or two strong rivals. The pure foundry that ASML enables, TSMC, earns extraordinary margins (~50%+ gross) but must sink enormous, recurring capex (tens of billions per year, capex/sales ~40%+) to defend its lead — a capital-intensive moat. ASML, by contrast, pairs comparable margins with capex of only ~5% of sales and a literal monopoly, so its cash economics (FCF/sales ~34%, ROIC 30%+) are arguably the best in the entire semiconductor value chain. Memory peers (Micron, SK Hynix, Samsung) sit at the cyclical, commoditized end — capital-hungry, price-takers, with returns that swing from superb to negative across the cycle — which is precisely the volatility ASML’s toll-collector model is insulated from on the competitive axis (though not the demand axis). The cross-read confirms the thesis: ASML is not just a good semiconductor company; it is the structurally best-positioned node in the chain.

Verdict: Exceptional financial quality — economics improve with scale, cash conversion is ~95%, ROIC exceeds 30%, the balance sheet is net cash, and earnings quality is high. This is, on the numbers alone, one of the best businesses in the index. The single watch-item is the working-capital/cash-flow lumpiness from prepayments, which is a presentation artifact, not an economic weakness — provided the order book holds.

7. Capital Allocation

Philosophy. ASML runs a disciplined, shareholder-friendly capital-allocation policy: fund the R&D roadmap that protects the monopoly, invest modestly in capacity, and return the bulk of free cash flow to shareholders via a growing dividend and large buybacks. It carries net cash and uses no financial leverage to flatter returns. On the Marathon asset-growth test, ASML is firmly on the right side — it does not empire-build; it shrinks the share count.

R&D — the moat-maintenance budget. €4.7B in 2025 (14% of sales), self-funded. This is the most important “capital allocation” decision the company makes, and it is well-spent: it funds High-NA, source-power gains, and the holistic-lithography roadmap that keep the monopoly intact. Unlike many tech companies, ASML’s R&D is a genuine competitive-advantage investment, not a treadmill to stand still.

Capital returns (FACT). In FY2025 ASML returned ~€8.5B — €2.55B in dividends and €5.95B in buybacks. The 2025 dividend was raised 17% to €7.50/share. A new 2026–2028 buyback program is active (€1.1B repurchased in Q1 2026 alone). The diluted share count fell from ~392M (Q1 2025) to ~385M (Q1 2026). Payout ratio is modest (~26–30%), leaving ample room for the dividend to keep growing.

M&A. ASML’s acquisition history is small, strategic, and capability-driven — Cymer (2013, light source), HMI (2016, e-beam inspection), Berliner Glas (2020, optical components) — plus the 24.9% Zeiss SMT stake. These were vertical-integration plays to secure the supply chain, exactly the kind of cost-based, control-securing M&A that Greenwald endorses (and the opposite of the revenue-synergy deals that destroy value). There is no record of large, dilutive, ego-driven acquisitions.

Incentive alignment. As a Dutch foreign private issuer, executive remuneration is disclosed in the 20-F and approved by the AGM under a Remuneration Policy. Pay is tied to financial and strategic KPIs rather than to size/growth alone — though, as with most large caps, insider ownership is small (~0.8% of shares), so alignment runs through compensation design more than founder-style equity stakes. (Open question: the precise LTI metrics and their rigor merit a dedicated proxy read.)

Capacity investment — the right kind of “growth capex.” ASML’s modest capex (€1.6B) is partly capacity expansion — building the cleanrooms and supply-chain capacity to lift low-NA EUV output to “at least 60” systems in 2026 and “at least 80” in 2027. This is demand-pulled, pre-sold capacity (customers are queueing), not speculative capacity-adding of the kind Marathon warns destroys returns. ASML is the bottleneck the industry waits on, so its capacity additions are de-risked by a visible order book — the opposite of a commodity producer building into hoped-for demand. This is capital allocation in service of the moat: every unit of added EUV capacity is a unit of monopoly revenue plus a decade of attached service.

The one critique — procyclical buybacks. ASML is repurchasing stock at the 96th percentile of its own valuation history. Buying back shares at ~58x trailing earnings is value-neutral-to-destructive relative to buying them in a drawdown; a more contrarian allocator would let cash build at these multiples and repurchase aggressively in the next demand shock (as in early 2025, when the stock traded near €600/$683). Returning cash via a steady buyback is defensible for a company with this cash generation and no better internal use for it — but the timing is the textbook large-cap procyclicality Marathon warns about. It is a quibble against an otherwise exemplary record, but it is the right quibble: the best capital allocators buy their own stock cheap, and ASML is buying it dear. (To be fair, the dividend — modest payout, +17% growth, ample headroom — is unambiguously well-judged.)

Verdict: Excellent capital allocation — disciplined, shareholder-aligned, net-cash, no value-destructive M&A — with one fair criticism: buybacks are running hot into a record valuation. Management has allocated capital intelligently over a full cycle; the only thing one would change is buyback timing.

8. Changes and Headwinds — Last Two Years

Strategic / operational developments:

• AI demand inflection (2024→2026). The single biggest change: AI infrastructure investment turned a 2024 digestion year into a 2025–2026 reacceleration, with management now describing demand as outpacing supply and customers “sold out.” FY2026 guidance was raised to €36–40B in Q1 2026.
• High-NA EUV commercialization. First HVM-ready EXE:5200B shipped April 2025 (175 wafers/hour); the platform moves toward high-volume manufacturing in 2027 — a genuine product-cycle milestone that extends the monopoly.
• Roadmap progress. 1,000-watt EUV source demonstrated; low-NA throughput roadmap to ≥330 wafers/hour at the start of the next decade; near-term throughput upgrades (NXE:3800E/F). These deepen the moat and the upgrade-revenue annuity.
• Capacity expansion. ASML is executing a 2026 output plan of “at least 60 low-NA EUV systems,” raising capacity “to at least 80 systems” in 2027 — i.e., ASML itself is the supply bottleneck the industry is waiting on.
• Leadership. Christophe Fouquet became CEO in 2024 (succeeding Peter Wennink); Roger Dassen remains CFO. The transition has been smooth and strategy is continuous.
• Capital returns reset. New 2026–2028 buyback program; 17% dividend increase.

Headwinds and overhangs:

• Export controls / China. The dominant headwind. EUV is banned to China; advanced DUV restricted since 2024. China revenue normalized from 36.1% of sales (2024) to 29.1% (2025) as the pre-restriction DUV backlog shipped out; further restrictions or Chinese retaliation are live risks. Management explicitly sized 2026 guidance bandwidth to “accommodate potential outcomes of ongoing discussions around export control.” The nuance worth holding: China demand was always DUV-weighted (mature/mid nodes, since EUV was never permitted), so the China revenue at risk is mostly the lower-ASP, more-commoditized end of ASML’s range — and the highest-value EUV franchise has essentially zero China exposure by construction. A China shock would dent revenue and sentiment but would not touch the monopoly’s profit core. The slower-burning concern is the China service base: if new-tool sales are curtailed, the attached multi-year service/upgrade annuity on China’s existing installed base could erode over time, and that is higher-margin revenue.
• Customer concentration intensifying. The largest customer rose to 23.9% of sales (from 16.6%); four customers now = 61% of revenue. AI is concentrating demand into a few leading-edge buyers.
• Cyclicality / digestion risk. 2024 demonstrated that a customer capex pause can flatten a year. A future AI-capex digestion phase is the principal cyclical threat.
• Tariffs / trade environment. Flagged in the FY2025 risk factors as a new source of uncertainty.

Verdict: On balance these changes strengthen the thesis — the AI inflection, High-NA, and roadmap progress are large positives that outweigh the (real, persistent) China/export overhang. The headwinds are concentrated in geopolitics and cyclicality, not in competitive position.

9. Risk Analysis (Risk Matrix)

The two risks that interact dangerously. The matrix’s individual rows understate the real danger, which is correlation. The valuation risk and the demand/China risks are not independent — a 96th-percentile multiple is only sustainable while the AI-demand narrative holds. The bear scenario is not “one risk fires” but “a demand or China shock fires and the multiple compresses simultaneously,” because the same event that cuts the earnings estimate also breaks the narrative that justified the multiple. That is why the illustrative bear case (Section 10) compounds a ~25% earnings decline with a multiple cut into a >50% drawdown: the two move together. Investors who think of “valuation risk” and “fundamental risk” as separate, diversifiable line items mis-size the downside here.

The risks that are genuinely remote. Conversely, the risks that would actually impair the franchise — technology disruption of EUV, loss of the monopoly, a credible competitor — are very-low-probability on any relevant horizon. There is no alternative patterning technology approaching EUV’s capability at scale; nanoimprint and directed self-assembly remain niche/experimental. This asymmetry — high cyclical/price risk, near-zero competitive risk — is the defining feature of ASML as an investment versus ASML as a business.

Reading of the matrix: The high-impact risks (digestion, valuation, Taiwan tail) are cyclical, geopolitical, and price risks — not competitive-position risks. ASML’s business is exceptionally safe; ASML’s stock at this price is not low-risk. That distinction is the whole investment debate.

10. Valuation Discussion (Embedded Expectations)

No price target and no recommendation here — this section frames what the current price implies. The single position in this memo is fenced in Claude’s Take above.

Where the stock trades (FACT). ASML (≈ASMLF) at roughly $1,778 (2026-06-09), a 52-week range of $683–$1,831 (the stock has ~2.6x’d off its low). Market capitalization ≈ $685B on ~385M shares; net cash (~€5.7B) makes enterprise value ≈ market cap. Headline multiples:

The single most important valuation fact: the stock is at the ~96th percentile of its own ten-year valuation history on every metric, and it trades above the average sell-side target (~$1,663). Short interest is negligible (0.19% of float) — there is essentially no bearish positioning; the market is uniformly long and uniformly paying up.

Forward multiple. On FY2026 guidance (€36–40B sales, ~52% GM, ~30% net margin → net income ~€11.0–11.5B, EPS ~€28.5–30), the forward P/E is ~50–55x in EUR terms, easing toward the high-30s/low-40s on 2027 Street consensus if the AI cycle delivers another up-year. This is a franchise multiple that requires the growth to show up.

Embedded-expectations analysis (the right lens). At ~$685B EV, the market is underwriting, roughly: (1) continued double-digit revenue growth toward the 2030 model (€44–60B sales, 56–60% gross margin); (2) durable monopoly margins with no competitive erosion (a safe assumption); (3) no China cliff — i.e., the export-control regime stays roughly where it is rather than worsening materially; and (4) no extended capex digestion of the kind that flattened 2024. Put differently, the price embeds the bull case and most of the dream case with little margin of safety for the cyclical or geopolitical air-pockets that history says are likely at some point.

Scenario framing (illustrative, not a target). The table below sketches FY2028 outcomes to show how much of the return depends on the multiple rather than the business. It holds ~385M shares and converts at EUR/USD ~1.15; figures are deliberately round.

The point is not the precision (none is claimed) but the shape: even a bull operating outcome delivers only a modest price return because the starting multiple is so high, while a bear outcome compounds an earnings decline with a multiple de-rating into a >50% drawdown — which is exactly the round-trip the stock made inside the last 52 weeks ($1,100→$683→$1,800). The asymmetry is unfavorable at this price, despite the business being excellent in every scenario.

Reverse-DCF intuition. To justify ~$685B EV at a ~9% cost of capital with ~€11B of current FCF, the market must believe FCF roughly doubles over the next 5–6 years and then keeps growing — i.e., the 2030 model (€44–60B revenue at 56–60% GM) is delivered and extended. That is a plausible path for a monopoly riding AI, but it is the bull path priced as the base case, leaving no discount for the cyclical interruptions that the company’s own history (2019, 2023–24) shows are recurring features, not bugs.

What the market is pricing correctly vs. incorrectly. Correctly: the monopoly, the margin structure, the AI tailwind, the inevitability of EUV/High-NA. Debatably: the price it is paying for those certainties, and the low probability it appears to assign to a cyclical or geopolitical interruption. The variant perception is not “is this a great business” (it is) but “is a 96th-percentile multiple the right price for a cyclical monopoly with 29% China exposure.”

Greenwald cross-check (EPV vs. asset value). Asset/reproduction value is a small fraction of the market cap — €20.8B of equity vs. $685B market cap — so essentially all of the value is franchise/growth value, justified only by the monopoly and its growth. That is appropriate for a business with this moat (Greenwald explicitly allows growth value for firms with identifiable, defensible advantages), but it also means there is no asset-value floor: in a de-rating, the stock falls to whatever multiple the market will pay for the earnings, with little tangible support beneath it.

Verdict (framing only): A monopoly business priced for monopoly perfection. The valuation is not absurd given the quality and growth, but it is historically extreme on the company’s own terms and offers little protection against the very cyclicality and geopolitics that define the franchise’s risks.

11. Variant Perception

Consensus belief. ASML is a generational compounder and the indispensable enabler of AI — “you cannot make an advanced chip without it” — and therefore deserves a premium multiple. The buy-side is near-universally long (negligible short interest), and the debate is mostly about how much to pay, not whether to own it.

Strongest bull case. The EUV/High-NA monopoly is unbreakable; AI is a decade-long secular driver that is supply-constrained at the customer level (customers are “sold out”); litho intensity rises every node; the IBM annuity compounds at high margin; and ASML’s own 2030 model (€44–60B, 56–60% GM) may prove conservative if AI capex keeps accelerating. At mid-teens compounding with expanding margins, today’s multiple is “expensive but justified,” and the stock works through earnings growth even if the multiple drifts down.

Strongest bear case. This is a cyclical monopoly trading at a 96th-percentile own-history multiple with 29% China exposure under an active export-control regime and 61% of revenue in four customers. Semicap demand is lumpy; 2024 already showed a digestion year is possible. A single guide-down — from AI-capex digestion, a China escalation, or a customer capex pause — meets a ~58x multiple with no asset-value floor and no short base to cover, and the de-rating is violent (the stock round-tripped from ~$1,100 to $683 and back inside one year). You are paying a perfection price for a business whose principal risks (cyclicality, geopolitics) are precisely the ones a high multiple cannot absorb.

The 3–5 assumptions that matter most:

1. AI capex durability — does leading-edge/HBM demand sustain through 2027–2030, or does it digest? (Bull vs. bear hinge.)
2. China / export controls — does the regime hold steady, worsen, or trigger retaliation that strands the China DUV + service base?
3. Margin trajectory — does mix (EUV + IBM + High-NA) push gross margin toward the 56–60% 2030 target?
4. High-NA adoption — do customers adopt EXE at scale on schedule (2027 HVM), validating the next growth leg?
5. Multiple — does the market keep paying ~55–58x, or does it normalize toward ~35–40x?

What would falsify each side. Falsify the bull: two-plus consecutive quarters of order/guidance deterioration, or a definitive China/DUV restriction that cuts a high-single-digit % of revenue. Falsify the bear: sustained order intake and a 2026 result at/above the top of the €36–40B range with gross margin tracking toward 53%+, demonstrating the cycle is being pulled forward rather than digested.

Why the variant perception is unusually narrow — and why that itself is a signal. Most variant-perception debates pit “great business” bulls against “deteriorating business” bears. Here, both sides agree the business is exceptional — there is no bear who thinks the EUV monopoly is at risk. That unusual consensus on quality is precisely what produces the crowded positioning (negligible short interest, price above the average target, 96th-percentile multiple): when everyone agrees a business is the best in its industry, the disagreement migrates entirely to price, and the price tends to overshoot. The contrarian observation is not “ASML is a bad business” (it is a superb one) but “a near-universally-loved cyclical monopoly with no bears is, by construction, priced for the bull case.” The edge, if there is one, is in patience and entry discipline — owning this franchise at a sane multiple after the next inevitable demand scare — rather than in any differentiated view of the company itself.

The information the market may be under-weighting. Two things. First, the cyclicality is structurally underestimated when AI is the narrative — the market extrapolates the AI capex wave linearly (Marathon’s “extrapolation in a cyclical world”), when ASML’s own recent history (the 2024 flat year, the 2019 air-pocket) shows demand digests periodically. Second, the China revenue quality — 29% of sales, normalizing down, under an active and tightening export regime — is a slow-moving impairment risk to one of the larger revenue legs that a momentum tape tends to discount. Neither changes the monopoly; both are reasons the embedded “no interruption” assumption may prove optimistic.

12. Fact vs. Interpretation Table

13. Open Questions

1. Net bookings/order trend. ASML stopped headlining a single backlog number; the precise order intake and book-to-bill by quarter (and its volatility) need ongoing tracking — management’s “very strong” is a hypothesis until the shipments arrive.
2. China DUV cliff timing. How much of the 29% China revenue is at risk under plausible 2026–2027 export-control tightening, and how durable is the China service base if new-tool sales are curtailed?
3. High-NA unit economics. Confirmed ASP, gross margin, and customer adoption curve for EXE — does it hit the 2027 HVM and the throughput/margin targets?
4. Executive incentive metrics. The exact LTI performance metrics and rigor (revenue/EPS/ROIC/relative TSR) from the Remuneration Report warrant a dedicated proxy read for alignment quality.
5. 2030 model credibility. What customer roadmaps and litho-intensity assumptions underpin the €44–60B / 56–60% GM 2030 targets, and how sensitive are they to an AI-capex slowdown?
6. Carl Zeiss SMT dependency. Extent of the single-source optics risk and the terms/control conferred by the 24.9% stake.

14. What Must Be True (Bull and Bear, with Falsification Tests)

For the BULL case to be right, the following must be true:

1. AI-driven leading-edge and HBM demand sustains double-digit WFE/litho growth through the 2027–2030 window (not a 2-year AI capex pulse). Falsification test: two or more consecutive quarters of declining order intake or a cut to full-year revenue guidance.
2. The China/export regime does not materially worsen — DUV-to-China and the China service base hold up roughly as today. Falsification test: a new restriction (or Chinese retaliation) that strands a high-single-digit % or more of revenue.
3. Margins march toward the 2030 model (gross margin to 56–60%) on EUV/High-NA/IBM mix. Falsification test: gross margin stalls below ~53% for a full year despite revenue growth.
4. High-NA reaches HVM on schedule (2027) at acceptable economics, validating the next growth leg. Falsification test: repeated EXE ramp/adoption delays beyond 2027.

For the BEAR case to be right, the following must be true:

1. A cyclical AI-capex digestion arrives (customer over-build → order pause), producing a flat-to-down ASML revenue year. Falsification test: 2026 lands at/above the top of the €36–40B range with sustained order momentum into 2027.
2. The 96th-percentile multiple normalizes toward ~35–40x as growth or sentiment cools, capping or reversing total return regardless of earnings. Falsification test: the multiple holds in the mid-50s through a full year of in-line results (the market keeps paying up).
3. A China/geopolitical shock (export escalation or Taiwan tail) impairs revenue or the customer base. Falsification test: China revenue stabilizes and export discussions resolve without material new curbs.

Synthesis: The bull and bear cases agree entirely on the business (a durable monopoly) and disagree only on cycle timing, geopolitics, and the price paid. The falsification tests are therefore mostly about order momentum, China policy, and the multiple — not about whether the moat survives. Watch the bookings, watch the export headlines, and watch the multiple.

15. Source Appendix

(See the dedicated Source Appendix — Appendix B in the combined report — for the full citation list. Primary sources: ASML FY2025 Form 20-F filed 2026-02-25, SEC CIK 0000937966; ASML Q1 2026 6-K filed 2026-04-15 (EX-99.1 press release, EX-99.3 U.S.-GAAP statements); ASML Q1 2026 earnings call transcript, 2026-04-15; ASML 2024 Investor Day 2030 model; third-party valuation-history data (own-history percentiles, 2026-06-08); market-data provider quote/stats for USD price/market-cap context. Third-party aggregator data are flagged as signals, not evidence, and reconciled to the filings.)

APPENDIX A — Standard Diligence Questionnaire

ASML Holding N.V. (ASMLF) — Standard Diligence Questionnaire Appendix

Supplemental to the analysis above. Grounded in primary filings (FY2025 20-F; Q1 2026 6-K; Q1 2026 earnings call). Fact / Interpretation / Assumption labeled where it matters.

General

What thoughtful questions have other investors asked about this company? The recurring, high-quality questions are: (1) Is the AI capex driving ASML’s order book a durable secular wave or a digestible pulse? (2) How much of the 29% China revenue is structurally at risk under export controls, and how sticky is the China service base? (3) Can gross margin actually reach the 56–60% 2030 target, or is mid-50s the ceiling? (4) Is High-NA EUV economically adopted at scale, or does it slip? (5) At a 96th-percentile valuation, what is the downside if a single quarter disappoints? Note (Interpretation): the debate among serious investors is almost never about the moat — it is about cycle, geopolitics, and price.

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? (Interpretation) Closer to a cyclical high within a secular uptrend — 2025 reaccelerated (+15.6%) on AI after a flat 2024 digestion year, and 2026 is guided up again (€36–40B). Margins (52.8% GM, 34.6% op) are near record. The risk is that current earnings reflect peak-ish AI capex intensity.

Driven by the external environment or internal actions? Both. Externally: the AI super-cycle and node migration drive volume. Internally: ASML controls its own supply (it is the bottleneck — “at least 60 EUV systems” output plan) and its margin via mix (EUV + high-margin IBM upgrades).

How stable are revenues? (Fact/Interpretation) System sales are lumpy and cyclical (recognized per-machine, sensitive to customer capex); IBM/service (25% of revenue) is recurring and stable, growing with the installed base (+26% in 2025). The blended revenue is more cyclical than a software company but cushioned by the annuity.

Outlook for products/services? Strong near term: demand outpacing supply, customers “sold out,” guidance raised. EUV intensity rises every node; High-NA is the next leg; IBM compounds.

How big will this market be? Growing. WFE is multi-tens-of-billions and growing structurally with AI/node migration; ASML’s own 2030 model targets €44–60B company revenue (from €32.7B in 2025). International by nature — sales are ~25% Taiwan, 25% Korea, 29% China, 12.5% US.

Business Quality & Competitive Moat

Is the industry getting more or less competitive? (Interpretation) In EUV, not at all — ASML has been a 100% monopolist since the product existed, with no credible entrant. In DUV, stably competitive (Nikon/Canon). Overall WFE is a stable oligopoly.

How profitable is the business (ROIC, ROE)? (Fact) ROE ~50%; ROIC in the 30%+ range (net cash, €11.3B EBIT). Among the highest sustained returns in large-cap tech.

How profitable is the industry — competitors, barriers? Highly profitable oligopoly; lithography captures a disproportionate share of the WFE profit pool. Barriers in EUV are effectively insurmountable (proprietary tech + sole-source Zeiss optics + Cymer source + €4.7B R&D scale + installed-base captivity).

Can the business be easily understood? Yes at the level of “monopoly toll-collector on advanced chipmaking,” though the underlying physics/engineering is extraordinarily complex (which is itself the moat).

Can it be undermined by foreign low-cost labor? No. The product is the most sophisticated machine ever mass-produced; cost of labor is irrelevant to the barrier. The threat vector is state-sponsored technology development (China), not low-cost labor — and EUV has resisted that for a decade-plus.

Do brands matter? Nature of competition? Not “brand” in a consumer sense; the competition is technological capability and there is no alternative. Customers co-develop nodes with ASML.

Customers’ switching costs? Extreme. A fab is architected around ASML tools; switching is effectively impossible, and the installed base locks in a decade of service/upgrade dependence.

Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet? (Interpretation) Yes — the EUV technology platform, IP, and the customer relationships/installed base are worth vastly more than their carrying value. Book equity (€20.8B) is a small fraction of the franchise’s economic value.

Off-balance-sheet liabilities? None material flagged. Standard lease (right-of-use €0.3B) and purchase commitments to the supply chain. Contract liabilities (customer prepayments, €3.6B+ non-current) are a favorable item — customers pre-fund working capital.

How conservative is the accounting? (Interpretation) High quality. US GAAP; net income tracks annual cash flow; SBC is small (~0.6% of sales); tax rate clean (~17%); no evident one-time distortions. The only nuance is quarterly cash-flow lumpiness from down-payment timing (a presentation artifact).

How CapEx-hungry is the business? Remarkably capital-light for its output: capex ~€1.6B (5% of sales) vs. €11B FCF. The heavy “investment” is R&D (€4.7B), expensed not capitalized.

Capital Allocation & Management

How much FCF, and how is it used? ~€11B FCF in 2025. Policy: fund R&D, modest capex, return the rest. FY2025 returned ~€8.5B (€2.55B dividends + €5.95B buybacks); dividend +17% to €7.50; new 2026–2028 buyback program (€1.1B in Q1 2026).

Significant acquisitions recently? No recent large M&A. Historic deals were strategic supply-chain control: Cymer (2013), HMI (2016), Berliner Glas (2020); 24.9% Carl Zeiss SMT stake. (Greenwald-approved: cost/control-based, not revenue-synergy.)

Buying back shares? Yes, materially — share count down ~392M→385M over the past year. (Critique: repurchasing at a 96th-percentile valuation is procyclical.)

Issuing large amounts of new shares to insiders? No — SBC is small (~€0.2B/yr); net dilution is negative (buybacks dominate).

Compensation policy / motivations of management? Disclosed in 20-F, AGM-approved Remuneration Policy tied to financial/strategic KPIs. Insider ownership small (~0.8%); alignment runs through pay design. (Open question: precise LTI metrics merit a dedicated proxy read.) New CEO Christophe Fouquet (2024); CFO Roger Dassen — smooth transition, continuous strategy.

Valuation & Market Data

ADR, MLP, or K-1 issuer? Neither MLP nor K-1. ASMLF is the OTC line for the Euronext Amsterdam ordinary shares; ASML (NASDAQ) is the same ordinary share listed in the US (not a sponsored-ADR-with-ratio structure — 1 share = 1 ordinary share). Dutch dividend withholding tax (15%) applies to US holders; reclaim mechanics under the US-NL treaty.

Dividend policy? Growing dividend (€7.50 for 2025, +17%), paid in interim instalments (€1.60 × 3 + €2.70 final proposed); modest ~26–30% payout, ~0.5% yield. Policy is “growing dividends + buybacks.”

How profitable is the business? Exceptionally — 52.8% GM, 34.6% op margin, 29.4% net margin, ~50% ROE.

Is net income diverging from cash from operations? Not on an annual basis (FCF ~95% of NI). Quarterly divergence is large and routine due to customer down-payment timing (Q4’25 op cash €11.4B; Q1’26 op cash −€2.2B) — analyze annually, not quarterly.

Risks & Downside

What factors would cause the stock to decline? (1) An AI-capex digestion / order air-pocket; (2) China export-control escalation or retaliation; (3) multiple compression from the 96th-percentile valuation; (4) a major customer capex cut (concentration: top-4 = 61%); (5) a Taiwan geopolitical shock; (6) High-NA ramp disappointment.

Risk of a catastrophic loss? (Interpretation) Low at the business level (monopoly, net cash, no leverage). The one catastrophic tail is a Taiwan conflict, which would devastate the entire semiconductor supply chain, not just ASML.

Chance of a total loss? Negligible barring a systemic geopolitical catastrophe. The realistic adverse case is a sharp drawdown (30–45%) from a demand/geopolitical shock meeting a high multiple — not a permanent impairment of the franchise.

Recent News & Events

Has the business environment changed recently? Yes — materially positively on demand: the AI inflection turned a flat 2024 into a reaccelerating 2025–2026, with FY2026 guidance raised to €36–40B and management describing demand as outpacing supply (“sold out” memory customers). (The recent-events read here is built from primary filings and the latest earnings call rather than from headline news flow.)

Significant acquisitions? None recent.

Change in accounting policies? None flagged; continues US GAAP.

Recent changes — markets, facilities, management? New CEO (Fouquet, 2024); High-NA EXE:5200B first HVM shipment (April 2025); capacity expansion to ≥60 EUV systems (2026) and ≥80 (2027); new 2026–2028 buyback; China revenue normalizing from 36% to 29% of sales under tightening export controls.

APPENDIX B — Source Appendix

ASML Holding N.V. (ASMLF) — Source Appendix

All figures in the memo trace to the sources below. Primary (filings) prioritized over secondary; third-party aggregator/AI signals labeled as such and reconciled to filings. Accessed 2026-06-09 unless noted.

Primary — SEC Filings (SEC EDGAR, CIK 0000937966)

Primary — Management Commentary (treated as hypothesis, validated vs. filings)

Secondary / Quantitative Helpers (aggregated data — not primary; reconciled to filings)

Analytical Frameworks

Notes on Currency & Listing

• ASML reports in EUR under US GAAP; price quoted in USD. EUR/USD ≈ 1.15–1.16 over the period (used to reconcile USD multiples to EUR fundamentals).
• ASMLF (OTC) and ASML (NASDAQ) are the same ordinary share (1:1), distinct trading lines for the Euronext Amsterdam security. No ADR ratio.