13F 15+/9-×0.0 · short↓0.18
Current snapshotASML Holding N.V.
AI Compute · Physical AI
AI Compute exposure in Physical AI.
EUV lithography monopoly - the capital-equipment bottleneck of advanced-node compute; parked pending panel.
Catalysts
Q3 2026 earnings on calendar 2026-10-14 (MarketScreener). Last print Q2 2026 (2026-07-15 IR): net sales €9.3B, FY26 guide raised to €43-45B; Intel Foundry first High-NA EUV HVM on select Panther Lake/18A layers. Low-NA EUV ~65 systems 2026. Parked pending panel.
Invalidation
needs primary source validation — pending deep-dive research. Placeholder: (1) thesis chain breaks on primary-source evidence, (2) named revenue/customer concentration risk, (3) margin compression or competitive displacement
Latest change
· Target moved from $2,376.89 to $2,378.00.
Mapped funds and investors
Direct relationships supported by published evidence.
Reported holdings as-of
| Fund / investor | As of | Fund portfolio weight | Evidence |
|---|---|---|---|
| VanEck Semiconductor ETF SMH | 4.47% | Source | |
| iShares Semiconductor ETF SOXX | 1.5% | Source |
Disclosed investment history
No sourced relationships mapped yet.
Showing up to four relationships in each group. ETF weights describe the fund’s portfolio, not ownership of this company. Private investment history does not establish current ownership.
needs primary source validation — no X/Reddit signal yet
Lower-confidence snapshotTarget price: $2361.80
Dated research targetMoat & position
- Evidence Grade
- Inferred
- Moat Durability
- Qualification barrier
- Toll Road Tier
- DIRECT_BOTTLENECK
Deep research update
3,709 words · Updated 19 Oct 2026 · 8 sources
| Metric | Value | Source |
|---|---|---|
| FY2025 revenue | €32,667M (+15.6% vs FY24 €28,263M) | 20-F companyfacts RevenueFromContractWithCustomerExcludingAssessedTax |
| FY2025 gross margin | 52.8% (€17,258M) | 20-F companyfacts GrossProfit |
| FY2025 operating income | €11,301M (34.6% margin) | 20-F companyfacts OperatingIncomeLoss |
| FY2025 net income | €9,609M (29.4% margin) | 20-F companyfacts NetIncomeLoss |
| FY2025 R&D | €4,699M (14.4% of revenue) | 20-F companyfacts ResearchAndDevelopmentExpense |
| Q1 2026 net sales | €8,767M; GM 53.0%; net income €2,757M; 67 new systems | 6-K EX-99.1 |
| Q2 2026 net sales | €9,326M; GM 54.0%; net income €2,918M; 86 new / 5 used systems | 6-K EX-99.1 |
| Q2 2026 EPS (basic) | €7.59 | 6-K EX-99.1 |
| Q3 2026 guidance | net sales €11.0–12.0B; GM 55–57%; R&D ~€1.2B; SG&A ~€0.4B | 6-K EX-99.1 |
| FY2026 guidance (raised) | net sales €43–45B; GM 54–56% | 6-K EX-99.1 |
| Shareholder return | €1.1B buyback in Q2 (2026–2028 program); €1.88 interim dividend (paid 2026-08-05) | 6-K EX-99.1 |
| Headcount | 44,500+ FTE | 6-K EX-99.1 |
- monetization model: One-time lithography system sales plus recurring service/field options/upgrades/spares; no subscription/SaaS. (ASML FY2024 20-F, Business & Note 8; asml.com ; EDGAR CIK 0000937966; accession UNKNOWN in model.) ASML files 20-F/6-K, not 10-K/10-Q.
- revenue mix: FY2024 net system sales ~77%, net service/field option sales ~23% of total net sales. (20-F FY2024, Note 8/segment note.)
- who pays: Foundry, logic, and memory semiconductor fabs via capital equipment budgets. (20-F FY2024, Business.)
- margin_value_capture: FY2024 gross margin 51.3%; value accrues in EUV/High-NA monopoly and installed-base service. (20-F FY2024, MD&A.)
- customer classes: Semiconductor Fabs - IDM Buyers
ASML does not report multiple GAAP operating segments; the 20-F presents one reportable segment and breaks revenue into system sales and services/field option sales. That split matters because ASML is effectively a “razor/razorblade” business: it sells a scarce, extremely expensive lithography tool and then earns recurring service/upgrade revenue on the installed base.
ASML FY2024 revenue mix, as filed:
| Line item | Value | Share of total net sales | Source |
|---|---|---|---|
| Total net sales | €28,263 million | 100.0% | ASML 20-F FY2024, SEC EDGAR, accession no. 0001104742-25-000012, filed 2025-02-26, accessed 2026-04-27 |
| Net system sales | €21,795 million | 77.1% | Same ASML 20-F FY2024 |
| Net services and field option sales | €6,468 million | 22.9% | Same ASML 20-F FY2024 |
The same filing’s comparative FY2023 data shows total net sales of €27,559 million, with system sales of €21,227 million and services/field option sales of €6,332 million. That implies FY2024 total net sales growth of approximately 2.6% year over year. The mix stayed nearly constant at roughly 77/23.
The revenue mix also has a product-technology dimension. ASML groups lithography systems into EUV, ArFi immersion, ArF dry, KrF, and i-line. From the quarterly 6-K disclosures, FY2024 unit shipments can be derived as follows:
| Technology | FY2024 units sold | Source |
|---|---|---|
| EUV | 68 | ASML 6-K Q4 2024 unit table, SEC EDGAR, accession no. 0001104742-25-000003, filed 2025-01-29, accessed 2026-04-27; derived by summing quarterly EUV unit disclosures |
| ArFi immersion | 160 | Same 6-K unit disclosures |
| ArF dry | 70 | Same 6-K unit disclosures |
| KrF | 87 | Same 6-K unit disclosures |
| i-line | 33 | Same 6-K unit disclosures |
| Total systems | 418 | Derived sum |
This unit count is the basis for the worked ASP calculation in Section 2.4. Because the 20-F does not provide the same year-sum unit table on a single page in the copy I used, the 418-unit figure is derived from the quarterly 6-K disclosures rather than quoted as a 20-F line item. If your downstream parser requires a single reported line, mark the exact 20-F unit count as needs primary source validation.
ASML is paid by semiconductor manufacturers, including integrated device manufacturers, foundries, and memory producers. The 20-F’s business description identifies the customer set as leading semiconductor manufacturers and specifically names TSMC, Samsung, SK Hynix, Micron, and Intel among the customers in its disclosed market description. The company’s revenue comes from:
- System sales — a capital equipment sale at very high ticket prices. The filed revenue and derived unit count imply a blended average selling price of roughly €52 million per lithography system for FY2024, but this is a blended figure across EUV and DUV; EUV tools sell at much higher prices than the blended ASP.
- Services and field option sales — maintenance, spare parts, upgrades, relocations, productivity enhancements, metrology/software options, and applications support. These revenues recur over the life of the installed tool.
Because the installed base is long-lived and process-qualification-dependent, service revenue is not optional for customers: a fab cannot simply drop the service contract without risking yield and uptime on a tool that may be the single most expensive machine in the fab.
ASML FY2024 profitability as filed:
| Metric | Value | Source |
|---|---|---|
| Total net sales | €28,263 million | ASML 20-F FY2024 |
| Gross profit | €14,498 million (derived from 51.3% gross margin on above sales) | ASML 20-F FY2024 / ASML 6-K Q4 2024 |
| Gross margin | 51.3% | ASML 20-F FY2024 / ASML 6-K Q4 2024 |
| Operating income | €8,850 million | ASML 20-F FY2024 |
| Operating margin | 31.3% | ASML 20-F FY2024 |
| Net income | €7,569 million | ASML 20-F FY2024 / ASML 6-K Q4 2024 |
| Net margin | 26.8% | ASML 20-F FY2024 |
These margins are high for semiconductor capital equipment and are consistent with a sole-source supplier at the leading edge. In contrast, competitor filings show that etch/deposition-focused suppliers operate with narrower gross margins, though direct margin comparability is limited by product mix.
This is a filing-supported, derived unit economics calculation.
The 20-F discloses that ASML’s customers include semiconductor manufacturers across the industry:
- Foundries — customer example named in filing: TSMC.
- Logic/integrated device manufacturers — customer examples named in filing: Intel, Samsung.
- Memory manufacturers — customer examples named in filing: Samsung, SK Hynix, Micron.
These buyer archetypes have different lithography purchase patterns:
- Leading-edge foundry/logic buyers buy EUV and High-NA EUV for sub-7 nm and sub-3 nm nodes.
- Memory buyers buy both EUV and DUV immersion; memory demand is historically more cyclical.
- Mature-node and specialty fabs buy DUV dry, KrF, and i-line tools.
ASML’s 20-F identifies a concentrated customer base. The precise customer-concentration percentage for 2024 is in the filing and should be referenced directly if it is material to an investment decision. Third-party summaries that quote a different figure should not be used unless they are traceable to the 20-F.
Lithography is a qualified process step. A fab cannot easily switch suppliers on an existing node without re-qualifying masks, resists, overlay models, and process recipes. That creates high switching costs and makes ASML’s installed base a recurring-service annuity. The service and field option revenue of €6,468 million in FY2024 is the monetization of that installed-base lock-in.
The longer the tool lasts inside a fab, the more service revenue ASML captures. This is why ASML disclosures often emphasize installed-base growth, not just quarterly unit shipments.
Calc 1 — The EUV capacity ceiling (the number that gates the AI industry).
Low-NA EUV capacity ≈ 65 systems (2026) → +30% ≈ 85 (2027, "nearly sold out") → investigating +30% ≈ 110 (2028). DUV-immersion ≈ 130 (2026) → +30% ≈ 169 (2027). Source: CEO Christophe Fouquet, Q2 2026 EX-99.1. A TSMC/Samsung/Intel can commit $25–40B to a new fab, but the number of EUV scanners they can install is determined entirely by ASML's (and Zeiss's) output — leading-edge capacity growth is supply-capped, not capex-capped.
Calc 2 — The H2 2026 revenue step-up (inflection math).
H1 2026 net sales = €8,767M + €9,326M = €18,093M. FY2026 guide = €43–45B ⇒ H2 2026 = €24.9–26.9B (midpoint €25.9B), i.e. H2 runs ~+43% above H1 — the back-half step-up that drives the +35% YoY guide. This is the capacity ramp + High-NA + service all compounding at once.
Calc 3 — The recurring installed-base annuity.
Installed Base Management (service + field options) sales = €2,488M (Q1) → €2,762M (Q2) = 29.6% of Q2 net sales; annualized ≈ €11B of high-margin, contractual revenue that scales with the ~200+ EUV + thousands-of-DUV installed base. This is the counter-cyclical toll that smooths system cycles.
Calc 4 — Cost of the moat (R&D intensity).
FY2025 R&D €4,699M ÷ revenue €32,667M = 14.4% of revenue to defend a 100% monopoly — the annual reinvestment no rival can match (Nikon's entire lithography segment is smaller than ASML's R&D budget).
Calc 5 — Margin trajectory = pricing power, not just volume.
Gross margin: FY2025 52.8% (€17,258M ÷ €32,667M) → Q1 2026 53.0% → Q2 2026 54.0% → Q3 guide 55–57% → FY2026 guide 54–56%. Rising GM while revenue accelerates +35% is the empirical signature of a toll with pricing power.
Calc 6 — Blended system ASP (with the mix caveat).
Q2 2026: 86 new systems on system sales ≈ €9,326M − €2,762M = €6,564M ⇒ ~€76M blended ASP. This mixes low-ASP DUV (€20–60M) with EUV (public estimates $200–300M low-NA, ~$350M High-NA, up to $400–500M fully loaded); the EUV/high-NA mix, not the DUV volume, is the value driver.
| Parameter | Value | Source Confidence |
|---|---|---|
| Q2 2026 net sales | €9,326M | measured (6-K EX-99.1) |
| Q2 2026 gross margin | 54.0% | measured (6-K EX-99.1) |
| FY2025 revenue | €32,667M | measured (20-F companyfacts) |
| FY2025 gross / op / net margin | 52.8% / 34.6% / 29.4% | measured (20-F companyfacts) |
| FY2026 net-sales guide | €43–45B (+35% YoY) | measured (6-K EX-99.1) |
| Low-NA EUV capacity | ~65 (2026) → ~85 (2027) → ~110 (2028e) | measured (CEO, EX-99.1) |
| DUV-immersion capacity | ~130 (2026) → +30% (2027) | measured (CEO, EX-99.1) |
| Installed Base Mgmt sales | €2,762M Q2 (29.6% of net sales) | measured (6-K EX-99.1) |
| EUV market share | 100% | measured (supply-chain analyses; 20-F "sole supplier") |
| Overall lithography share | ~94% | measured (supply-chain analyses) |
| EUV system ASP | $200–300M low-NA; ~$350M High-NA | assumed (public estimates) |
| Installed EUV base | ~200+ systems | assumed (public estimates) |
| Zeiss equity stake | ~24.9% | measured (20-F / supply-chain) |
| Cumulative EUV R&D | >€10–20B | assumed (supply-chain analyses) |
| China share of revenue (pre-restriction) | ~30% | assumed (needs primary-source validation) |
\[
\text{Blended system ASP} = \frac{€21,795 \text{ million}}{418 \text{ systems}} = €52.1 \text{ million per system}
\]
| Parameter | Value | Source (URL + document ID + filing date + access date) | Tier | Measured/derived/assumed | Cross-verified |
|---|---|---|---|---|---|
| ASML FY2024 total net sales | €28,263 million | ASML 20-F FY2024, sec.gov , accession no. 0001104742-25-000012, filed 2025-02-26, accessed 2026-04-27 | Tier 1 | Measured | Yes, against ASML 6-K Q4 2024, same issuer |
| ASML FY2024 net system sales | €21,795 million | Same ASML 20-F FY2024 | Tier 1 | Measured | Yes, against ASML 6-K Q4 2024, same issuer |
| ASML FY2024 services and field options | €6,468 million | Same ASML 20-F FY2024 | Tier 1 | Measured | Yes, against ASML 6-K Q4 2024, same issuer |
| ASML FY2024 gross margin | 51.3% | ASML 20-F FY2024 / ASML 6-K Q4 2024 | Tier 1 | Measured | Yes, same issuer |
| ASML FY2024 operating income | €8,850 million | ASML 20-F FY2024 | Tier 1 | Measured | Single-source |
| ASML FY2024 net income | €7,569 million | ASML 20-F FY2024 / ASML 6-K Q4 2024 | Tier 1 | Measured | Yes, same issuer |
| ASML FY2024 R&D expense | €4,398 million | ASML 20-F FY2024 | Tier 1 | Measured | Single-source |
| ASML FY2024 total lithography units | 418 | Derived from ASML quarterly 6-K unit tables | Tier 1 | Derived | Single-source |
| ASML FY2024 EUV units | 68 | Derived from ASML quarterly 6-K unit tables | Tier 1 | Derived | Single-source |
| ASML FY2024 ArFi units | 160 | Derived from ASML quarterly 6-K unit tables | Tier 1 | Derived | Single-source |
| ASML blended lithography system ASP | €52.1 million | Derived from ASML 20-F net system sales ÷ derived total units | Tier 1-derived | Derived | No, derived from two same-issuer filings |
| Low-NA EUV wavelength | 13.5 nm | ASML technical disclosures / 20-F product description | Tier 1 / Tier 2 | Measured/technical | Single-source |
| High-NA EUV numerical aperture | 0.55 NA | ASML technical disclosures | Tier 1 / Tier 2 | Measured/technical | Single-source |
| Low-NA EUV resolution at k1=0.4 | 16.4 nm | Calculated using Rayleigh formula | Tier 1-derived | Derived | No |
| High-NA EUV resolution at k1=0.4 | 9.8 nm | Calculated using Rayleigh formula | Tier 1-derived | Derived | No |
| ASML lithography market share | >80% | ASML 20-F / ASML investor materials | Tier 1 / Tier 2 | Reported/claimed | Single-source |
| Lam Research FY2024 revenue | $14,905 million | Lam Research 10-K FY2024, sec.gov , accession no. 0000707549-24-000044, filed 2024-08-13, accessed 2026-04-27 | Tier 1 | Measured | Single-source |
| Applied Materials FY2024 revenue | $27,176 million | Applied Materials 10-K FY2024, sec.gov , accession no. 0000006951-24-000043, filed 2024-12-13, accessed 2026-04-27 | Tier 1 | Measured | Single-source |
| Standard EUV ASP | €170–200 million | ASML investor disclosures / industry reporting | Tier 2 / needs primary source validation | Assumed/needs primary source validation | No |
| Worldwide wafer fab equipment market size 2024 | ~$105 billion | Lam/AMAT 10-K references to market data | Tier 1 | Reported/needs primary source validation | Needs primary source validation |
ASML is the sole supplier of extreme-ultraviolet (EUV) lithography systems on Earth — 100% of commercial EUV scanner supply, ~94% of overall lithography. Every leading-edge chip (sub-7nm) made by TSMC, Samsung, Intel, SK Hynix, and Micron is patterned by an ASML scanner. The binding constraint is not demand, not design, not packaging — it is the physical rate at which ASML (and its exclusive optics supplier Carl Zeiss SMT) can assemble EUV machines: ~65 low-NA EUV systems in 2026. That number is the hard ceiling on how much leading-edge fab capacity the entire AI industry can add in any given year.
- EUV monopoly, zero commercial alternative. ASML is the only company that manufactures EUV lithography systems; Nikon and Canon compete only in DUV (i-line/KrF/ArF) and have no EUV capability. Canon has not released a new DUV tool since 2012; Nikon posted its largest-ever loss in FY2026 with its Precision Equipment (lithography) segment struggling. China's SMEE holds ~4% of the global i-line (legacy-node) market; a Shenzhen prototype can generate EUV light but cannot print a commercial chip. Next-gen alternatives (e-beam, actinic EUV) are 5+ years from deployment.
- Exclusive Zeiss optics — "two companies, one business." ASML's 2025 annual report (20-F, Note 26): "Carl Zeiss SMT GmbH is our single supplier, and we are their single customer, of optical columns for lithography systems." Zeiss SMT is the sole source of EUV projection optics (multilayer mirrors, illuminators, collectors); ASML holds a ~24.9% equity stake. ASML's own filing states its scanner output is directly limited by Zeiss SMT's production capacity, and that a prolonged Zeiss disruption "would effectively cease to be able to conduct our business."
- Vertically integrated light source. Cymer (acquired 2013) makes the laser-produced-plasma (LPP) EUV source; Trumpf (Ditzingen) is the sole supplier of the high-power CO₂ lasers firing each machine.
- Brute-force complexity. Each EUV system is a ~457,000-component assembly from 5,000+ suppliers across 17 countries, holding optical alignment to picometer tolerances in vacuum while firing 50,000 tin droplets/second into a laser at ~500,000°C. High-NA projection optics contain 40,000+ parts and weigh ~12 tons.
- Sunk-cost barrier. >€10–20B cumulative EUV R&D over two decades, with TSMC/Intel/Samsung having co-invested equity in the EUV program — the entrant would have to amortize a national-scale R&D program with no installed base.
- product_service: EUV and DUV lithography systems, metrology/inspection tools, and service.
- technical function: Photolithography: projects 13.5nm EUV or 193nm DUV through masks/optics onto photoresist-coated wafers to print IC patterns.
- why it is hard: Sole EUV source, Zeiss optics, vacuum/mechanics/software, multi-decade R&D, patents, supplier ecosystem, customer co-development; High-NA adds complexity. (20-F FY2024, Business/Risk Factors.)
- substitutes: Leading-edge EUV: none at volume. Non-leading-edge: DUV multi-patterning, e-beam, nanoimprint, directed self-assembly. (20-F FY2024, Business/Risk Factors.)
- pricing power: HIGH — sole EUV supplier, high switching costs, scarce High-NA capacity.
ASML sells:
- Lithography systems
- EUV systems — low-NA EUV at 0.33 numerical aperture (NA) and High-NA EUV at 0.55 NA.
- DUV immersion systems — argon-fluoride immersion (ArFi) at 193 nm with water immersion.
- DUV dry systems — ArF dry at 193 nm, krypton-fluoride (KrF) at 248 nm, and i-line at 365 nm.
- Metrology and inspection systems — tools that measure overlay and critical dimension to enable process control.
- Computation and software products — computational lithography, process optimization, and application software.
- Services and field options — installation, service, spare parts, upgrades, relocations, and productivity options.
The 20-F describes the company as a manufacturer of semiconductor wafer processing equipment focused on lithography. The technical product description in ASML’s annual report and investor materials states that EUV lithography uses 13.5 nm wavelength light and reflective optics inside a vacuum chamber.
A lithography system transfers an integrated circuit pattern from a mask/reticle onto a silicon wafer. The key physical limit is resolution. Resolution is governed by the Rayleigh equation:
\[
CD = k_1 \cdot \frac{\lambda}{NA}
\]
Where:
- \(CD\) = critical dimension / resolution
- \(k_1\) = process factor
- \(\lambda\) = exposure wavelength
- \(NA\) = numerical aperture
ASML has demonstrated pricing power by holding gross margin above 50% despite heavy R&D and supply-chain complexity. The company can raise prices at the leading edge because there is no second EUV supplier. At the mature DUV end, however, pricing power is weaker because Nikon and Canon compete in KrF and i-line systems.
The 20-F’s risk factors still frame pricing power as cyclical: if memory or foundry customers defer capacity additions, ASML cannot force shipments and revenue recognition. Therefore, pricing power should be understood as structural at the leading edge, cyclical in timing, not as an unconditional ability to push price every quarter.
- Lithography: ASML ~94% share; sole EUV (100%). Nikon (ArF/i-line/KrF, record FY2026 loss, Intel-sluggish) and Canon (no new DUV since 2012) are DUV-only. China SMEE ~4% i-line (legacy); Shenzhen EUV prototype cannot print commercially.
- Adjacent equipment (not lithography): Applied Materials, Lam Research, KLA, Tokyo Electron — compete in deposition/etch/inspection, not patterning. No cross-threat.
- Hyperscaler exposure: hyperscalers (Google/Amazon/Microsoft/Meta/NVDA) are downstream — they buy from TSMC/Samsung/Intel who buy ASML. No hyperscaler can in-source EUV lithography (unlike networking/cooling/servers); hyperscaler capex is the demand, not a substitute.
- The real substitute threat is geopolitical: Dutch/US export controls restrict EUV and (now) DUV-immersion to China — historically ~30% of ASML revenue — and China's indigenous lithography push (SMEE, Shenzhen prototype) is the only long-run non-ASML path, still 5+ years from EUV-class capability.
- China demand shock. China was historically ~30%+ of revenue; export controls (Dutch, coordinated with US) cap that market. A full China decoupling (or a faster-than-expected SMEE catch-up) removes a growth wedge ASML cannot replace at equal margin.
- Customer concentration. TSMC, Samsung, and Intel dominate the EUV order book; a memory/logic downturn (e.g. HBM oversupply) or a single mega-customer's capex pause would compress the order book sharply, given "nearly sold out 2027" capacity leaves no buffer.
- Zeiss single-point-of-failure. ASML's 20-F itself warns a prolonged Zeiss disruption "would effectively cease to be able to conduct our business" — the toll is one German optics house deep.
- High-NA yield risk. High-NA mask defectivity (>10%) and anamorphic-optics complexity are still ramping; a delayed High-NA HVM ramp would slow the sub-2nm monetization the +35% guide partly assumes.
- The capacity constraint cuts both ways. ~65→85→110 EUV systems/yr means ASML cannot oversupply to chase share, but it also means the +35% guide is already near the physical ceiling — upside beyond 2027 is bounded by Zeiss + ASML factory expansion, not by demand.
The ASML thesis would be materially weakened if any of the following appears in primary filings or technical reality:
- EUV monopoly breaks. If Nikon, Canon, or a new entrant ships a production-ready EUV system at acceptable yield and throughput, ASML’s sole-source pricing power erodes. No competitor 10-K reviewed supports this today, but Canon’s nanoimprint push is a nearby substitute risk for certain DUV applications.
- Lithography intensity stalls. If advanced chipmakers rely more on deposition, etch, 3D packaging, chiplet interconnects, or backside power delivery, then Lam and Applied Materials’ narrative wins and ASML’s share of total equipment spend could decline even without losing lithography market share.
- Export-control revenue loss. ASML’s 20-F risk factors discuss Dutch and international export controls. A tightening of restrictions on DUV immersion sales to China would remove a material source of demand. The 2024 China revenue percentage is in the 20-F; if it is large and cut off, earnings fall.
- High-NA EUV adoption delays. If TSMC, Intel, or Samsung push out High-NA EUV insertion, ASML’s most expensive tools sit in backlog longer, revenue recognition slips, and R&D payback deteriorates.
- Memory capex downturn. The ASML filing discloses customer concentration among memory and logic leaders. A simultaneous memory capex freeze by Samsung and SK Hynix would cut EUV and DUV shipment volumes.
- Service revenue fails to re-accelerate. If customers defer upgrades and relocations, the €6,468 million service/field option stream could stagnate, reducing the recurring revenue cushion.
- Margins compress. The ASML gross margin is already high. If EUV manufacturing costs outrun price increases, or if DUV mix rises, gross margin could fall from 51.3% toward 45–48% quickly, and the market’s valuation multiple may compress.
- Customer concentration converts to pricing pressure. A very small number of buyers—TSMC, Samsung, Intel—control a large share of purchases. If one large customer delays orders, it affects results more than a diversified equipment supplier.
The disconfirming evidence is not that ASML has competition today; it is that ASML’s value is concentrated in the assumption that leading-edge semiconductor manufacturing will continue to depend disproportionately on lithography—and that EUV/High-NA EUV will remain the only practical path.
Sources
8 sources preserved from the latest qualified research update.