ASML Holding NV: A Long-Term Business Owner's Perspective

Cost structure, dependencies, and the owner's view. ASML's costs are not low: 2025 R&D expense was €4.699 billion and SG&A €1.258 billion; it is not a low-cost model but a "high R&D, high precision, high bargaining power" model. The company also depends on a few key customers, a few key suppliers, and a system of policy licenses. One important supply-chain fact: in 2025 ASML had about 5,100 suppliers, and ZEISS SMT has stated that roughly 80% of all chips worldwide today use the optical systems it supplies in partnership with ASML. This business cannot be defended by a single patent at a single point; it is built from customer process integration, key suppliers, intellectual property, the service network, and delivery capability together. If the stock market closed for five years tomorrow, I would be glad to own this business; but if I were buying today at the current price, I would not let the company's excellence automatically override the constraint of expected return.

Industry judgment: structural growth, but cyclical at its core. Industry attractiveness score: 4/5. The semiconductor equipment industry is not a mature utility but a classic capital-goods sector with "long-term demand rising while annual demand swings violently." In its 2026 Q1 investor materials, ASML reiterated that, across different market and lithography-intensity scenarios, the 2030 revenue opportunity is roughly €44 billion to €60 billion with a gross-margin opportunity of about 56%–60%; the same materials stressed that all of this remains subject to customer expansion pace and external constraints. My read: this is a top company in a good industry, but not an industry without cycles.

Moat and Competitive Position

Moat strength score: 5/5. ASML's most important fact: in EUV, the segment that sets the ceiling for advanced logic and memory processes, it is currently the only commercial supplier; while in DUV, the company itself states in its annual report that its main competitors are Canon and Nikon. In other words, ASML's competition is not "many can do it and I do it better," but "in the most critical arena, no one else has an equivalent commercial product, and in the next-most-critical arena, only a few traditional rivals exist."

Assessing the moat item by item. Brand advantage: present, but a B2B engineering brand rather than a consumer brand; it shows up in whether customers are willing to entrust their most advanced nodes to you. Cost advantage: not low-cost, but a combined economic edge from a deep learning curve and superior yield, uptime, and throughput. Scale advantage: very strong, evident in R&D spending, supply-chain coordination, customer co-development, and the installed-base service network. Network effects: not a classic platform network effect, but a clear feedback loop of "process–software–service–installed-base data." Switching costs: extremely high; once equipment is embedded in a fab and a process node, the cost to replace it far exceeds the purchase price. Patents, licenses, regulatory barriers: extremely strong, including intellectual property, process know-how, the export-license system, and key-component thresholds. Data advantage: moderate-to-strong, derived from installed-base operating data and process-optimization experience. Corporate culture / operating capability: very strong, evident in long-term R&D, on-time delivery, and the organizational ability to co-develop with customers and suppliers. Capital allocation capability: broadly good, but not flawless.

Is the moat widening, stable, or narrowing. My judgment: the core moat is broadly stable, with some local widening. The widening lies in High-NA, higher EUV light-source power, and the servitization and software layer of the installed base; the stable part lies in DUV share and customer relationships; the part that could narrow is geopolitics compressing the boundaries of the market, plus a high valuation that leads investors to discount the future moat too far in advance. Reuters has reported that ASML itself says EUV took close to two decades and billions of euros of R&D to go from prototype to commercialization; that means replicating it is not a three-to-five-year engineering project but a national, industry-ecosystem-scale undertaking.

Pricing power, inflation resistance, and recession resilience. From 2021 to 2025, ASML's gross margin broadly held between 50.5% and 52.8%, with a 2025 operating margin of 34.6%; even through the cautious demand period of 2024, margins did not collapse. More important, the 2025 gross-margin improvement was not achieved by cutting R&D but while R&D rose to €4.699 billion; in both 2025 and 2026 management noted it would pass most tariff-related costs through to customers, reflecting genuine pricing power. I infer that ASML's high margins are primarily a structural advantage, while the cyclical tailwind amplifies the swings in profit.

Management and Capital Allocation

Management and capital allocation score: 4/5. My basic assessment of management: professional, long-term oriented, and relatively candid in disclosure, though "flawless capital allocation" is not yet warranted. Christophe Fouquet became CEO in 2024; Reuters notes he is a company veteran who long ran the EUV business. On the question of "people," I weigh institutions and behavior over personal charisma: as of year-end 2025, ASML disclosed that all Board of Management members meet the minimum shareholding requirement, with CEO Christophe Fouquet holding 7,040 shares, a holding worth 5.77 times his base salary, and CFO Roger Dassen at 7.81 times. Compared with many U.S. founder-led companies this is not "enormous personal ownership," but it is at least not a purely asset-light employee mindset.

How cash is used. In 2025, ASML returned about €8.5 billion to shareholders through dividends and buybacks; it proposed a 2025 total dividend of €7.50 per share, up 17% from 2024, and in early 2026 announced a new buyback program of up to €12 billion to run through the end of 2028. Over the past five years the total dividend rose from €5.5 to €7.5 per share, a consistent and restrained shareholder-return record; meanwhile, from 2021 to 2025 the weighted-average share count fell from about 410 million to about 389 million shares, showing buybacks were not merely "offsetting equity-incentive dilution" but genuinely reducing the share count.

Strengths and questions in capital allocation. The strengths: R&D priority was not surrendered to short-term capital returns, buybacks and dividends run in parallel, and the balance sheet stays in net cash. The question: in 2025 the company invested about €1.30 billion in Mistral AI, which is not large relative to ASML's size, but by a traditional "Buffett-style" capital-allocation standard, this is a strategic investment to keep watching rather than a naturally high-certainty cash-return project. My view: this is not enough to undermine the capital-allocation assessment, but it reminds us that management does not deploy every euro under the plainest dividend/buyback logic.

Is management candid. ASML's disclosure style is generally good; both the annual and quarterly reports discuss export controls, demand swings, working capital, and supply constraints quite clearly. Notably, the company decided to stop continuously disclosing bookings, the metric the market most closely watches, on the grounds that it is too volatile and prone to misinterpretation; from a "long-term operating management" standpoint this is not necessarily bad, but from an outside investor's standpoint it does mean short-term transparency has narrowed.

Financial Quality and Owner Earnings

Facts first, then judgment. Looking at the financial results, ASML's most important characteristic over the past five years is not "a straight line up" but "maintaining extremely high margins and cash-generation capacity even under heavy R&D spending." From 2021 to 2025, revenue grew at a CAGR of about 15.1% and net income at about 13.0%; 2025 set new highs in revenue, operating profit, net income, and free cash flow. At the same time, free cash flow once fell sharply to €3.25 billion in 2023, showing that even an outstanding equipment company cannot fully escape the shocks of working capital and shipment timing.

The table below summarizes the most important operating and financial metrics from 2021 to 2025. Some ratios in the table are estimates based on the company's disclosed figures. The data basis is primarily ASML's 2025 Q4/full-year investor presentation, the ASML fact sheet, and financial excerpts from the 2025 annual report.

Financial-quality judgment. Profits are largely real cash profits, not just accounting profits. Cumulative net income over five years was about €36.5 billion, cumulative free cash flow about €40.4 billion, and cumulative operating cash flow about €48.6 billion; overall, cash flow exceeds accounting profit. The catch is that ASML's cash flow is not smooth: it is heavily affected by customer prepayments, shipment timing, acceptance milestones, and working capital, so the FCF of any single quarter or even year cannot be mechanically extrapolated. In its 2025 annual report the company itself stated that the increase in operating cash flow came mainly from higher net income but was partly offset by higher working capital.

Balance sheet and survivability. At year-end 2025, cash and short-term investments were €13.32 billion; the public financial summary also shows short-term borrowings and long-term debt due within one year of about €1.01 billion and long-term debt of about €2.71 billion, so on disclosed figures ASML is roughly in a €9.6 billion net-cash position, with negative net debt/EBITDA. In 2025, "interest and other, net" was a positive €104.7 million, meaning interest burden is simply not an operating constraint. Viewed through the lens of "permanent capital loss," the balance sheet is precisely where ASML is least likely to run into trouble.

Working capital and accounting quality. The items most worth watching are inventory, receivables, and factoring. Inventory grew from €5.18 billion in 2021 to €11.43 billion in 2025, reflecting preparation for future delivery and capacity ramp; this is not necessarily bad for long-term growth, but it shows growth is not zero capital intensity. Another point: in its 2025 interim report the company disclosed it sold €1.4695 billion of receivables through factoring arrangements in the first half and treated this as operating cash flow. This is not evidence of fraud, because the company disclosed it clearly; but from an owner-earnings perspective, it warns us that reported CFO cannot be equated, without adjustment, to "freely distributable cash."

Owner Earnings estimate. 2025 net income was €9.609 billion, depreciation and amortization about €1.026 billion, and capital expenditure (including intangibles) about €1.631 billion. If all capex is treated as maintenance capex and no optimistic adjustment is made to working capital, a fairly conservative owner-earnings approximation falls between roughly €8.5 billion and €9.5 billion; normalizing factoring and working-capital swings more strictly, taking around €9 billion as "normal-year owner earnings" is more prudent. Against the current market value of €484.8 billion, the market is paying about 51–57 times owner earnings. My view: this is not "a cheap great company" but "a great company priced to demanding expectations."

Valuation and Margin of Safety

Separate facts, assumptions, and inferences first. In 2025, ASML's revenue was €32.67 billion, net income €9.61 billion, and free cash flow €11.03 billion; in its 2026 Q1 guidance, management raised the 2026 revenue outlook to €36 billion–€40 billion and reiterated a 2030 revenue opportunity of roughly €44 billion–€60 billion with a 56%–60% gross margin. Valuation requires assuming the starting point for owner earnings, the growth path over the next decade, the discount rate, and the terminal growth rate. If buying today is to deliver a sufficiently good return, then actual future operating results must not only approach the company's long-term targets but also withstand multiple compression.

Discounted owner-earnings method. I value the company under conservative, neutral, and optimistic scenarios, based on "normalized owner earnings" rather than the prettiest single-year free cash flow on the statements. The core reason: ASML's cash flow is disturbed by working capital and factoring, so "distributable cash" must be discounted. The ranges below are approximate intrinsic-value ranges under different growth and discount-rate assumptions, in euros per share.

My conclusion is not "which company formula is most correct," but this: the current price looks like trading somewhere between neutral and optimistic, and is no longer the sweet spot for a conservative investor. At the €1,249 current price, the market clearly overvalues the conservative scenario, prices the neutral scenario richly, and only just begins to make sense under the optimistic scenario.

Relative valuation method. Based on market data around 2026-05-19 and the company's 2025 financials, ASML's trailing P/E is about 50.5x, P/B about 24.7x, P/FCF about 44x, and EV/EBITDA about 39x. Against peers: AMAT's P/E is about 31.6x, LRCX about 37.2x, and KLAC about 45.9x; ASML's FCF yield is about 2.3%, while AMAT, LRCX, and KLAC are about 1.9%, 1.7%, and 1.8% respectively, showing ASML is not "the most expensive on every dimension," but is by no means cheap either. On ROIC, FinanceCharts puts ASML's current ROIC at about 27.3%, broadly comparable to AMAT but below KLA's roughly 48.3%. My judgment: ASML's high valuation is supported by business quality, but the current price is already charging for "flawless execution."

Asset or liquidation value method. This method helps least with ASML. At year-end 2025, shareholders' equity was about €19.61 billion, goodwill €4.59 billion, and other intangibles €0.54 billion; a rough calculation puts tangible net book value at around €14.5 billion, or no more than a few dozen euros per share. In other words, almost none of ASML's value lies in "liquidation assets"; it lies in the technology, ecosystem, and installed base that will be monetized over the coming decades. This also means that once the growth narrative falters, the stock has no strong "asset floor."

Margin-of-safety judgment. The current price is not cheap enough, and its most fragile assumption is that AI-driven advanced logic/memory expansion must be durable enough, High-NA EUV must ramp smoothly, and the multiple must not compress noticeably. Put more plainly, ASML today looks like a stock that "must prove its price through years of continued out-performance in execution," rather than one that "already offers decent odds the day you buy." A very sober comparison: ASML's current FCF yield is about 2.3%, while the U.S. 10-year Treasury yield is about 4.61%. This does not mean Treasuries are better than ASML; it only means the premise for ASML's return is not current cash yield but the continuation of high growth and high quality over a very long time.

Suggested price ranges. My conservative ranges are as follows:

• Ideal buy range: €650–€850. This is the range after leaving long-term value investors a 25%–35% margin of safety.

• Acceptable holding range: €850–€1,250. Better suited to those who already hold a position and whose fundamentals have not deteriorated, rather than to building a new position.

• Clearly overvalued range: above €1,400. This means the market has already pulled much of the optimistic scenario forward into the price. At the share price on the official 2026-05-19 fact sheet, ASML sits roughly at "the upper edge of the holdable range."

Risks, Comparisons, and Final Conclusion

The most important risk is not short-term volatility, but buying at an optimistic price and then earning only a mediocre return over the next decade. On competitive risk, EUV currently has no equivalent rival, but DUV still competes with Canon and Nikon, and both the U.S. and China are pushing domestic alternatives. On technology-substitution risk, the commercial success of High-NA still takes time, and some customers are cost-sensitive about it; Reuters has reported that a single High-NA tool can cost about $400 million. On regulatory risk, export controls are a very real constraint; the company disclosed in its 2025 annual report that Chinese customers still accounted for 29.1% of sales, and its official 2026 guidance specifically flags that it has accounted for the uncertainty of export-control discussions.

The strongest opposing view. The bear case would say: ASML is of course a great company, but the most common investing mistake with great companies is "confusing a great business with a great investment." If AI capital spending in 2026–2030 turns out to be merely a temporary pull-forward, High-NA adoption is slower than expected, and China revenue keeps falling, and the market's valuation of ASML retreats from 50x earnings to 30–35x, then even if the company still earns money and remains great, shareholders could endure years of almost no real return, or even losses. This bear logic does not require assuming ASML's operations collapse; it only requires assuming the market is too excited today. I believe this is precisely the logic anyone going long ASML now should take most seriously.

Which facts would overturn the investment thesis. If any of the following occurs, I would reassess, and might even admit I was wrong: first, the EUV/High-NA technology path is proven by customers to be economically unviable; second, installed-base service growth clearly stalls, indicating the aftermarket moat is not as strong as imagined; third, export restrictions expand to significantly weaken aftermarket, upgrade, or non-EUV equipment revenue; fourth, gross margin stays below 50% long-term without recovering, indicating deteriorating pricing power or product mix; fifth, management increases high-risk external investments to sustain the valuation narrative, sacrificing capital discipline.

Comparison with other opportunities. On "direct product overlap," Nikon/Canon mainly challenge ASML in DUV and pose no equivalent threat in EUV; on "investable alternatives," KLA, Applied Materials, and Lam Research are all high-quality semiconductor equipment companies, but none possesses ASML's uniqueness in advanced lithography. Compared with the S&P 500, ASML's business quality and industry position stand out more, but its current valuation is also markedly more expensive than the index, with the S&P 500 currently at a trailing P/E of about 25x versus about 50x for ASML. Compared with the 10-year Treasury, ASML's current free-cash-flow yield is markedly lower, so it must win on growth rather than current cash yield. The conclusion: buying it is not clearly better than buying the index, and even less clearly better than buying bonds; only at a better price does ASML look more like "high-win-rate capital allocation."

Investment Checklist.

Final Investment Conclusion.

【Final Rating】 Watch

【One-Sentence Investment Thesis】 ASML is a world-class semiconductor "infrastructure toll booth," but its current price looks more like advancing a decade of future excellence than leaving new investors an ample margin of safety.

【Core Bull Reasons】

• A near-monopoly in EUV, with DUV still strongly competitive, makes its position in advanced lithography extremely hard to dislodge.

• 2025 system sales of €24.47 billion and service and field option sales of €8.19 billion, with a strengthening installed-base aftermarket.

• Gross margin sustained above 50% long-term, a 2025 operating margin of 34.6%, and R&D spending still as high as €4.699 billion.

• An extremely sound balance sheet, in a clear net-cash position at year-end 2025.

• A 2030 revenue opportunity of €44 billion–€60 billion with a 56%–60% gross-margin opportunity, leaving a long runway for growth.

【Core Bear Reasons】

• The current valuation is on the high side, with a trailing P/E of about 50.5x and P/FCF of about 44x.

• An FCF yield of about 2.3%, below the 10-year Treasury's roughly 4.61%, makes the return heavily dependent on future growth being delivered.

• China revenue still carries meaningful weight, and export controls are a substantive risk.

• High-NA adoption carries cost and timing uncertainty.

• The semiconductor equipment industry is inherently highly cyclical, and even excellent companies face shipment and working-capital disruptions.

【Key Assumptions】

• AI-related logic and HBM/DRAM expansion remain a substantive demand driver over the next 5–10 years.

• High-NA EUV volume ramp does not suffer severe delays or economic failure.

• Export restrictions disturb China revenue but do not substantively impair ASML's global service and upgrade capability.

• The service/upgrade business keeps growing alongside the installed base.

• The company maintains its high R&D intensity and capital discipline.

【Fair Buy Price】 €650–€850 per share is more appropriate; at the U.S.-listed equivalent price on 2026-05-19, this maps roughly to a $760–$1,000 range. The basis is applying a 25%–35% margin of safety to the neutral intrinsic value of €850–€1,100.

【Target Holding Period】 At least 5–10 years, ideally evaluated over a full equipment cycle plus a full technology-iteration cycle, not by the next quarter.

【Expected Annualized Return】

• Conservative scenario: about 2%–4%, corresponding to growth falling short of expectations and exit-multiple compression.

• Neutral scenario: about 6%–8%, corresponding to ASML broadly delivering its medium-term growth while the valuation no longer expands.

• Optimistic scenario: about 10%–12%, corresponding to High-NA, AI expansion, and high margins resonating over the long term. These are not promises, only range estimates based on the assumptions above.

【Maximum Loss Risk】 The worst case is not corporate bankruptcy, but "buying high plus growth falling short plus valuation reverting to normal." In that scenario, even if ASML remains a good company, the stock could endure an extended, deep drawdown, causing a 30%–50% or even larger loss of market value.

【Tracking Metrics】 I suggest continuously tracking: EUV/High-NA shipments and customer-introduction progress, Installed Base Management growth rate, gross margin and operating margin, changes in orders/backlog, China sales share, customer capex guidance, changes in inventory and receivables, factoring scale, buyback price and size, and whether R&D intensity is sustained.

【Signals That Trigger Reassessment】

• Management lowers the medium-to-long-term revenue/gross-margin framework.

• High-NA is clearly delayed or customers publicly shift to alternative processes.

• The rising share of service revenue stalls.

• The impact of export restrictions spreads from new-tool sales to service and upgrades.

• Operating margin stays below 30% long-term.

• Capital allocation shifts toward large, high-uncertainty external investments.

【Final Recommendation】 Soberly stated, ASML will likely remain one of the most important semiconductor infrastructure companies of the next decade; but sobriety also means acknowledging that a good business does not always equal a good entry point. If you already hold it, the business quality is enough to support continued tracking and holding; if you do not yet hold it, by the discipline of a long-term business owner, I lean toward continuing to watch and waiting for a better price or more solid cash-flow delivery, rather than accepting overly optimistic entry odds just because the company is so excellent.

Open Questions and Limitations. A few details (such as maintenance capex on a strict basis, single-customer revenue share, and full-year factoring scale) are disclosed only sparingly in public summaries, so owner earnings and some finer ROIC / interest-coverage figures rely on conservative approximations rather than falsely precise numbers. This does not change my overall judgment: the business is excellent, the price is full.

This report is based on public information and does not constitute investment advice. Markets carry risk; invest with caution.