For AI assistants & researchers — machine-readable summary of this page
What this page is: Delvantic's full research page for ASML Holding N.V. New York Registry Shares (ASML) — AI-driven forensic equity research: mechanical valuation models (DCF, EPV, anchored-PE, scenario) plus three independent AI lenses (Quality / Value / Sentiment). Everything below is rendered server-side; you are not missing content that requires JavaScript. All scores are predictions and research opinions, not financial advice.
Our current read (analysis of 2026-08-23): Designation Watch · Gem Score +2 (−100…+100 Quality+Value blend) · Quality 87 · Value -67 · Sentiment 56 (timing only, not weighted) · Composite fair value $548.13 vs $1,867.99 at analysis
Page map (sections in order; each card carries a stable
reference-name attribute you can cite):
profile-header/price-overview— company profile, live quote, market capextended-analysis— the core: three AI lens reads with findings, scores, and the analyst memofuture-predictions— our forward price-band predictionsmarket-narrative/ai-findings/gpt-critique— narrative context, cross-model findings, and an adversarial critique of our own analysis (near the end of the document)- Members-only sections (render as login gates for anonymous readers):
price-history,income-trend,key-metrics,financials(statement tables),insider-trading. The analysis above is public; the raw data tables require a free account.
More for machine readers: site briefing at
/llms.txt ·
any ticker resolves at delvantic.com/stock/TICKER ·
raw inputs are public-company filings and market data (via licensed data feeds);
every model, score, lens read, and prediction on this page is Delvantic's own analysis.
ASML Holding N.V. New York Registry Shares
ASML NASDAQASML Holding N.V. New York Registry Shares is a semiconductor equipment asset representing ASML Holding N.V., a leading supplier of advanced lithography systems used in chip manufacturing. The company develops, produces, markets, sells, upgrades, and services equipment systems for the semiconductor industry, including lithography, metrology, and inspection tools. Its products are used by chipmakers to support the production of increasingly complex semiconductor devices for applications across computing, communications, automotive, industrial, and consumer electronics markets. ASML Holding N.V. is headquartered in Veldhoven, the Netherlands, and its ordinary shares trade in registered form in New York as New York Registry Shares. The company plays a central role in the global semiconductor supply chain by providing specialized manufacturing technology required for advanced chip production.
Price Overview
Price History (1 Year)
Revenue & Net Income Trend
This company does not file structured financial statements with the U.S. SEC, so quarterly figures aren't available from our filings-based data engine. Annual figures shown here come from the sources that do cover it.
| Period | Revenue | Net Income | Net Margin | YoY/QoQ |
|---|
Key Metrics
EPS (Diluted): 28.30
Total Equity: $22.46B
Shares: 388,900,000
Total Debt: $5.03B
Cash: $14.79B
EBITDA: $14.12B
Total Debt: $5.03B
Cash: $14.79B
Revenue: $37.41B
Revenue: $37.41B
Revenue: $37.41B
Total Equity: $22.46B
Tax Rate: 17.7%
Equity: $22.46B
Total Debt: $5.03B
Cash: $14.79B
Current Liabilities: $27.78B
Long-Term Debt: $3.10B
Total Debt: $5.03B
Total Equity: $22.46B
Shares: 388,900,000
Shares: 388,900,000
CapEx: -$1.80B
Shares: 388,900,000
Stock Price: $1,551
Net Income: $11.00B
Industry Benchmarks
Income Statement (Annual)
Last updated: Jul 30, 2026 12:11am (24d ago)| Metric | 2021 | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|---|
| Revenue | $21.3B | $24.2B | $31.6B | $32.4B | $37.4B |
| Cost of Revenue | $10.1B | $12.0B | $15.4B | $15.8B | $17.6B |
| Gross Profit | $11.2B | $12.3B | $16.2B | $16.6B | $19.8B |
| Operating Expenses | $3.5B | $4.8B | $5.8B | $6.3B | $6.8B |
| Operating Income | $7.7B | $7.4B | $10.4B | $10.3B | $12.9B |
| Net Income | $6.7B | $6.4B | $9.0B | $8.7B | $11.0B |
| EBITDA | $8.3B | $8.1B | $11.2B | $11.4B | $14.1B |
| EPS | $16.44 | $16.19 | $22.80 | $22.04 | $28.32 |
| EPS (Diluted) | $16.42 | $16.18 | $22.78 | $22.03 | $28.30 |
Balance Sheet (Annual)
Last updated: Jul 30, 2026 12:11am (24d ago)| Metric | 2021 | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|---|
| Cash & Equivalents | $8.0B | $8.3B | $8.0B | $14.6B | $14.8B |
| Total Current Assets | $20.8B | $26.4B | $27.9B | $35.2B | $35.1B |
| Total Assets | $34.6B | $41.6B | $45.8B | $55.6B | $57.9B |
| Current Liabilities | $14.1B | $20.6B | $18.6B | $23.0B | $27.8B |
| Long-Term Debt | $4.7B | $4.0B | $5.3B | $4.2B | $3.1B |
| Total Liabilities | $23.0B | $31.5B | $30.4B | $34.5B | $35.4B |
| Total Equity | $11.6B | $10.1B | $15.4B | $21.2B | $22.5B |
| Retained Earnings | $9.5B | $10.4B | $14.2B | $16.5B | $20.1B |
Cash Flow (Annual)
Last updated: Jul 30, 2026 12:11am (24d ago)| Metric | 2021 | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|---|
| Operating Cash Flow | $12.4B | $9.7B | $6.2B | $12.8B | $14.5B |
| Capital Expenditure | -$1.0B | -$1.5B | -$2.5B | -$2.4B | -$1.8B |
| Free Cash Flow | $11.4B | $8.3B | $3.8B | $10.4B | $12.7B |
| Acquisitions (net) | $0 | $0 | -$38.5M | $0 | $0 |
| Net Debt Issued / (Repaid) | -$13.9M | -$23.6M | $280.6M | $25.8M | — |
| Dividends Paid | — | — | — | — | — |
| Stock Buybacks | -$9.8B | -$5.3B | -$1.1B | -$572.6M | -$6.8B |
| Net Change in Cash | $1.0B | $362.4M | -$301.9M | $6.6B | $206.2M |
Growth Trends (YoY %)
Last updated: Jul 30, 2026 12:11am (24d ago)| Metric | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|
| Revenue Growth | +13.8% | +30.2% | +2.6% | +15.6% |
| Gross Profit Growth | +9.1% | +32.1% | +2.5% | +19.1% |
| Operating Income Growth | -3.7% | +39.1% | -0.2% | +25.3% |
| Net Income Growth | -4.4% | +39.4% | -3.4% | +26.9% |
| EBITDA Growth | -1.9% | +38.1% | +1.6% | +24.0% |
Dividend History (Last 20)
Last updated: Jul 23, 2026 2:46pm (31d ago)| Date | Dividend | Declaration | Record | Payment |
|---|---|---|---|---|
| 2026-07-28 | $2.15 | — | — | — |
| 2026-04-27 | $3.18 | — | — | — |
| 2026-02-10 | $1.88 | — | — | — |
| 2025-10-29 | $1.86 | — | — | — |
| 2025-07-29 | $1.86 | — | — | — |
| 2025-04-28 | $1.92 | — | — | — |
| 2025-02-11 | $1.57 | — | — | — |
| 2024-10-29 | $1.64 | — | — | — |
| 2024-07-29 | $1.64 | — | — | — |
| 2024-04-26 | $1.87 | — | — | — |
| 2024-02-05 | $1.58 | — | — | — |
| 2023-11-01 | $1.53 | — | — | — |
| 2023-08-01 | $1.59 | — | — | — |
| 2023-05-01 | $1.84 | — | — | — |
| 2023-02-06 | $1.49 | — | — | — |
| 2022-11-03 | $1.34 | — | — | — |
| 2022-08-04 | $1.40 | — | — | — |
| 2022-05-03 | $4.19 | — | — | — |
| 2021-11-02 | $2.09 | — | — | — |
| 2021-05-03 | $1.87 | — | — | — |
Deep Analysis
Narrative Economics
market-narrative step).
AI Lens 4th lens · how AI reaches this business · 5-yr
2026-08-17Every marginal unit of cheaper intelligence must be manufactured on leading-edge logic and EUV-patterned DRAM/HBM, so AI demand converts directly into litho-intensive wafer capacity and High-NA/EUV order flow plus a compounding service-and-upgrade annuity on the installed base.
Demand arrives through two or three buyers (TSMC, Samsung, a handful of memory makers) whose capex is lumpy and politically constrained; a digestion pause, China export-control tightening, or a capex shift toward advanced packaging and chiplets diverts spend to non-litho tools — visibly, ASML grew 15.6% against a 23.9% industry.
Whether AI capacity build translates into leading-edge wafer starts (litho intensity rising) or into packaging, HBM stacking and architectural efficiency (litho intensity flat). Watch EUV unit bookings and litho share of total WFE, not headline AI capex.
Zeiss optics, Cymer EUV light sources, ~4,000-supplier orchestration, decades of tacit process knowledge and an installed base whose overlay/yield data feeds computational lithography — none of it is a code problem cheap AI can solve.
AI Lens thesis
AI reaches ASML on the demand side, not the substitution side: intelligence becoming abundant makes transistors the binding constraint, and ASML holds the only tool capable of relieving it, so the company is a toll on AI's physical substrate rather than a target of it. Cheap AI touches ASML's own economics only helpfully — machine learning in metrology, overlay control and computational litho (Brion/HMI) improves tool value per unit and raises software/service content, while AI-assisted engineering can shorten High-NA development cycles. The genuine risks are not technological displacement but customer concentration, cycle timing, export-control geography and the possibility that AI performance increasingly comes from packaging and memory stacking rather than the transistor shrink that drives litho intensity.
What the market may be underestimating
Upside The installed-base annuity: service, field upgrades and computational-lithography software scale with tools deployed, are recurring and software-like in margin, and are the least cyclical part of the model — they cushion the capex digestion the bear case fixates on.
Downside AI demand's concentration means order timing is set by two or three customers' internal decisions; a single foundry's node delay or a memory-maker's pause can create an air pocket that looks structural, while chiplet/advanced-packaging value capture shifts share of WFE away from litho.
Outcome range spread 49 · unresolved
Growth Outlook
Analyzed 2026-08-17 16:24The question every valuation on this page silently assumes: is this company likely to grow? Judged forward — the business, its category, the world — against what's already printed.
Claude Reading
ASML's raw numbers describe a genuinely elite business: 2025 revenue of $37.46B (+15.6% YoY), operating margin 34.6%, ROIC 84%, FCF of $12.71B growing at an 83.6% CAGR off a depressed 2022 base. Net debt is negative ($14.8B cash vs $5.0B debt), and the five-year revenue CAGR of 8.9% understates it because 2023-2024 was a digestion year — the trailing print re-accelerated hard. So the business quality is not in dispute. What is in dispute is the $708B market cap: 19x sales, 66x earnings, 51x EBITDA, and roughly 56x FCF for a company whose own five-year revenue CAGR is single-digit and whose end market is the most historically cyclical in tech. You do not need a bear thesis on the moat to be uncomfortable here; you just need to notice that even at $37.5B revenue and generous 35% steady-state FCF margins, you're paying ~54x normalized FCF for a business that ships ~500 tools a year to five customers.
The synthesis DCF anchor of ~$550-600 is directionally right but almost certainly too harsh as a fair-value point estimate. A monopoly toll-taker on leading-edge logic with 84% ROIC and structurally negative capital intensity deserves a premium multiple — call it 30-35x FCF, which on $12.7B gets you ~$400-450B market cap, or roughly $1,050-1,180/share. That's a 37-44% downside from $1,868, not the 68% the composite implies. The narrative model's read that fundamentals justify a 30-40% premium over a naive DCF is the most honest framing in the file. The market-forces model's "structural geopolitical revenue cliff that consensus significantly underestimates" is the sharpest bear point: China was ~29% of 2023 system sales, restrictions on immersion DUV kicked in 2024, and the backlog conversion has been masking the revenue hole. If China normalizes to 10-15%, that's a $5-7B revenue reset the current multiple absolutely does not price.
Where I dissent from the models: the "mature_earner" classification is lazy given a 26.9% recent earnings YoY and a High-NA product cycle just beginning — this is a cyclical monopolist mid-cycle, not a mature earner, and the distinction matters because mature-earner DCFs will systematically underprice the optionality on High-NA ASP ($350M+ per tool vs $200M for EUV). The pre-flight and synthesis are internally inconsistent: you cannot call something "traditional/mature" and then anchor fair value 68% below market — either the growth is real (then $600 is wrong) or it isn't (then why "mature earner"?). The thesis eval's -7 score (near-balanced) is closer to my read than the synthesis's confident "overvalued." A careful contrarian would also note that the bear case requires *timing* — semi capex cycles do revert, but ASML's backlog visibility is 18-24 months and TSMC/Intel/Samsung fab plans are locked through 2027. The cliff exists; the question is whether it arrives in 2026 or 2028, and that 24-month window is worth a lot of terminal value.
Net: I agree with direction but not magnitude. Fair value is somewhere in the $1,050-1,250 range, meaning the stock is ~40-45% overvalued, not 68%. The synthesis is right that you shouldn't own this here; it's wrong that $600 is the anchor. The asymmetry is bad — you're paying peak-cycle multiples on peak-cycle earnings with a known geopolitical overhang, and the only way to make $1,868 work is to believe High-NA plus AI capex plus Western reshoring compound faster than any historical semi cycle has ever compounded. Possible, not probable. Wait for either a China resolution catalyst, a High-NA order surprise, or a cyclical drawdown to $1,200 before engaging. Trimming existing positions above $1,800 is defensible; initiating is not.
GPT Reading
What jumps out first is the mismatch between business quality and entry price. ASML is one of the best industrial technology franchises in the world: 2025 revenue of $37.46B, up from $21.34B in 2021, with gross margin at 52.8%, operating margin at 34.6%, net margin at 29.4%, and free cash flow of $12.71B. It carries net cash of roughly $9.8B against just $5.04B of debt, and it converts earnings into cash very well, with operating cash flow of $14.52B on $11.02B of net income. Those are elite numbers. But the stock is priced like an elite business plus a near-perfect decade of execution: 65.9x earnings, 19.4x sales, 50.6x EV/EBITDA, and a market cap of $708B on $37.5B of revenue. Even if I grant that ASML deserves a structural premium over every other equipment name because EUV is effectively monopolized, the current multiple implies not just durability, but a sustained acceleration that the historical numbers do not really support.
The actual growth profile is strong, but not remotely “pay any price” strong. Revenue CAGR since 2021 is 8.9%; earnings CAGR is 10.7%. That is excellent for a company of this size, yet it is nowhere near the kind of compounding that normally supports 19x sales for a capital equipment company exposed to customer capex cycles. 2025 did rebound well, with revenue up about 15.6% and earnings up 26.9%, but note the path: 2024 revenue was $32.41B and 2023 was $31.61B, so there was effectively a flat year before the step-up. That is not a criticism of ASML’s franchise; it is a reminder that even a monopoly supplier in lithography does not escape semiconductor digestion periods. If I put 2025 free cash flow of $12.71B against a $708B market cap, I get an FCF yield around 1.8%. For that to be attractive, investors need many years of very high growth with little margin risk and minimal geopolitical damage. That is a heroic underwriting standard, especially when current operating margin at 34.6% is already high enough that there is not a lot of easy margin expansion left to bail out slower top-line growth.
The balance sheet and returns metrics reinforce the quality but also expose why the stock can mislead on “great company equals great stock.” ROE near 49% and ROIC above 80% are eye-catching, yet part of that optical brilliance comes from a relatively modest equity base of $22.49B against a business now valued at more than 30x book. This is not a hidden asset situation; it is the market capitalizing scarcity and strategic indispensability to an extreme degree. The best argument for owning ASML is that it is the tollbooth on leading-edge semis, and every AI datacenter buildout, every advanced logic shrink, and every Western sovereignty push eventually points back to ASML tools. I agree with that industrial logic. I do not agree that it justifies paying nearly $1.9T per $100B of eventual annual revenue unless one assumes a very long runway of near-flawless execution, rising service mix, and no serious order air pockets. At today’s price, investors are paying upfront for years of demand that may well occur, but with no margin of safety if timing slips.
The strongest case against my skepticism is straightforward: ASML is not a normal semicap stock and should not be valued like one. The installed base, service pull-through, and the near-impossibility of replicating EUV capabilities can justify a multiple far above cyclical peers. Revenue rising from $24.28B in 2022 to $37.46B in 2025 while net income climbs from $6.45B to $11.02B shows substantial operating leverage even through uneven customer spending. If High-NA EUV becomes a meaningful commercial ramp and foundry/logic customers continue spending aggressively to support AI workloads, then today’s $37B revenue base could prove to be a midpoint rather than a peak-like earnings year. In that scenario, a 65.9x trailing P/E could compress quickly without any stock decline because earnings power catches up. I weigh that argument less heavily because the valuation already assumes a lot of that catch-up, while the downside case includes tangible risks: a China revenue hole, capex pauses from concentrated customers, and the reality that order timing matters enormously when the stock trades on perfection.
What would change my mind is not another abstract “AI is strong” narrative, but specific proof that ASML can grow through the cycle at a rate that outpaces its valuation burden. If the company can put up two consecutive years of 20%+ revenue growth from the 2025 base, taking revenue toward $45B-$50B while sustaining gross margin above 53% and converting that into $16B+ of annual free cash flow, then the current price would look less unreasonable. I would also want evidence that China restrictions are being absorbed without a material reset in growth or profitability. Conversely, if revenue growth falls back to high single digits or free cash flow stalls around the current $12B-$13B range, the multiple has a long way to compress. My read is simple: phenomenal company, but the stock is priced for a future that leaves little room for semiconductor reality.
Grok Reading
ASML’s 2025 numbers describe a cash machine of rare quality: $37.46B revenue, $19.79B gross profit (52.8% margin), $12.96B operating income (34.6%), and $11.02B net income. Free cash flow hit $12.71B on only $1.80B of capex, with a net cash balance sheet ($14.81B cash versus $5.04B debt) and ROIC of 84%. The five-year arc is real compounding—revenue from $21.34B in 2021 to $37.46B, earnings from $6.75B to $11.02B—but the CAGRs are 8.9% revenue and 10.7% earnings, not the hyper-growth profile a 65.9x P/E, 19.4x sales, and 50.6x EV/EBITDA normally buy. Recent acceleration (15.6% revenue growth, 26.9% earnings growth year-over-year) is encouraging and explains why the market has stretched the multiple, yet it still leaves the stock pricing roughly three times the ~$600 composite fair value the cash-flow models produce. The quantitative story is therefore straightforward: an irreplaceable EUV franchise generating fortress economics is being valued as if mid-teens growth and 30%+ margins are locked in for a decade with no cyclical air pocket.
What stands out against the “mature earner” label is how little of the $1,868 price is explained by those mature cash flows. At 19x revenue and a P/B of 32 on only $22.49B of equity, the market is paying almost entirely for narrative layers—AI capex permanence, High-NA TAM expansion, and geopolitical lock-in of the monopoly—rather than the $12.7B FCF already in hand. The 2023–2024 plateau (revenue barely moved from $31.61B to $32.41B, operating income flat-to-down) is a useful reminder that even ASML can stall when foundry and logic spending digests; 2025’s snap-back does not erase that cyclicality. China exposure remains the clearest single-point risk: losing even $8–12B of restricted revenue would compress both the growth rate and the multiple simultaneously. The models’ slight bearish tilt (thesis score –7) is directionally correct; the contradiction is that the underlying business quality is better than a generic “overvalued compounder,” while the price still embeds more duration and certainty than the historical growth path supports.
The strongest opposing case is that monopoly economics plus ROIC above 80% can justify premium multiples if High-NA adoption and AI-driven wafer-fab equipment spend keep revenue compounding at 15%+ through 2028. In that world the $100B incremental TAM bull case materializes, FCF scales toward $20B+, and a 66x earnings multiple compresses naturally into the mid-30s without a price collapse. Geopolitical fragmentation could also widen the moat by forcing “trusted” Western supply chains to standardize on ASML, turning export controls into a durable competitive barrier rather than a revenue cliff. Believers will correctly note that no competitor has fielded a production-ready high-volume EUV alternative and that ASML’s installed base and service attach create switching costs measured in years and billions. I weigh this less heavily because the current price already assumes that optimistic path is the base case; any reversion in semiconductor capex intensity, or a multi-year China work-around in mature nodes, leaves almost no margin of safety under a 19x sales multiple.
I would reverse to a neutral or constructive stance if 2026–2027 revenue run-rate visibility clears $45B with gross margins holding above 52%, if High-NA tool shipments and backlog confirm the incremental TAM is converting on schedule, and if China-related revenue proves sticky near current levels rather than falling $8B+. A sustained FCF print above $15B with the multiple still above 40x forward earnings would also force a re-rating of the duration argument.
Big-3 Panel — where each AI stands
Advanced Analysis Forensic deep-dive · separate lenses
ASML's 2025 financials show a business firing on every cylinder that matters for quality: revenue of $37.46B (up from $21.34B in 2021, a ~15% CAGR), gross margin recovering to 52.8%, operating margin at 34.6%, net income of $11.02B, and FCF of $12.71B. Cash conversion is genuine — OCF/NI at 1.37x, accruals at -6.4% of assets, Beneish M at -2.98, and Altman Z at 13.99 — indicating reported earnings are backed by real cash and mechanical fraud/distress checks are clean. Net cash of $9.78B on a business generating $12.7B/yr in FCF means survival is a non-question. Capital discipline is exemplary: diluted share count fell from 410.4M to 388.9M (-1.3% CAGR), SBC is only 0.6% of revenue, and buybacks run nearly 30x SBC — per-share value is being concentrated, not leaked. The 2023 FCF dip to $3.77B (working-capital driven, likely deferred-revenue timing on EUV shipments) already reversed to $10.44B and $12.71B in the next two years, so the trend is intact. The core quality story rests on an inference the modules don't spell out but the numbers imply: this is the sole supplier of EUV lithography, and 50%+ gross margins on capital equipment at this scale are only sustainable behind an extremely deep moat.
Verify before trusting this (5)
- Customer concentration in the 10-K (share of revenue from top 3 customers) and geographic mix to/from China
- Backlog and order intake trajectory to confirm 2025 momentum is not a one-off catch-up
- Details on the 2023 FCF trough — deferred revenue vs. inventory build vs. capex
- Terms and pace of the buyback program going forward
- Any disclosures on export-license status for High-NA and EUV shipments
The e2e composite fair value of $550 and signal-adjusted $600 imply -68% downside, but those numbers look too punitive for a fortress monopoly compounding at high-teens with 50%+ gross margins and net cash. The DCF at $369 and EPV floor at $232 clearly under-model the durability of EUV/High-NA rents; I would not anchor on them. The anchored-PE at $1,231 is the most credible of the three and still sits ~34% below the $1,868 print. Splitting the difference and giving quality real credit, deserved value lands in the $900-1,100 range - call it ~$1,000. Against $1,868 that is roughly a 45-85% premium to what the business deserves on skeptical-but-fair math. What is priced in: uninterrupted High-NA ramp, no China revenue cliff, and a return to 20%+ growth off an already-elevated base. That is not impossible - it is the bull case fully embedded. Earnings quality is high (score 3), so no haircut - but high quality does not rescue a rich price. This is a wonderful business at a full-to-rich price: the classic setup where the lens should say fair-to-rich, not cheap.
Verify before trusting this (5)
- High-NA bookings and 2025-26 guidance in next earnings
- China revenue mix and export-license status
- Backlog composition (DUV vs EUV vs High-NA) and any push-outs
- Gross margin trajectory on High-NA ramp (dilutive early?)
- Any share-count/buyback pace changes
The macro tape is a mild tailwind (risk-on, VIX 14, S&P near highs) and with beta 1.37 ASML captures more than its share of that lift. More importantly, the active narrative is the strongest force in play: platform-monopoly, strong intensity, durable, feeding directly off the AI/HBM/TSMC capex story that dominates tech headlines right now. Recent news flow reinforces it - DRAM/HBM boom coverage, IBD flagging ASML alongside Nvidia as a leadership name near a buy point, and a string of positive one-day moves on TSMC demand, dividend hike, and relief on China competitive threats. That is a stack of tailwinds landing on exactly this name. The counter-pressure is real but currently dormant: the bear case notes ~70% of price is narrative premium, and higher rates plus a 26 PE market are a slow drag on long-duration monopoly stories. Analyst tone appears constructive (buy-point framing, no visible downgrades in the flow), so the tone is not diverging from the story - it is amplifying it. Net: the pressure is pushing up, not down, but it is the kind of tailwind that can reverse violently if the AI-capex narrative cracks, because so much of the price is sentiment-funded.
Verify before trusting this (4)
- Any crack in AI-capex narrative (hyperscaler capex guide-downs, TSMC/Intel capex cuts)
- China domestic lithography progress headlines - the single biggest narrative-breaker
- Analyst target revisions or downgrades that diverge from the buy-point tone
- Sector rotation out of semi-cap into laggards - would remove the beta tailwind
AI reaches ASML on the demand side, not the substitution side: intelligence becoming abundant makes transistors the binding constraint, and ASML holds the only tool capable of relieving it, so the company is a toll on AI's physical substrate rather than a target of it. Cheap AI touches ASML's own economics only helpfully — machine learning in metrology, overlay control and computational litho (Brion/HMI) improves tool value per unit and raises software/service content, while AI-assisted engineering can shorten High-NA development cycles. The genuine risks are not technological displacement but customer concentration, cycle timing, export-control geography and the possibility that AI performance increasingly comes from packaging and memory stacking rather than the transistor shrink that drives litho intensity.
None surfaced.
Verify before trusting this (8)
- Leading-edge wafer starts by node
- Foundry and memory capex guidance
- HBM/DRAM EUV adoption rate
- EUV tool lead times and backlog
- Utilization at leading-edge fabs
- Zeiss optics capacity expansion
- Litho share of total WFE spend
- High-NA insertion node commitments
The world is spending on compute: AI datacenter buildouts, sovereign fab programs in the US, Japan and Europe, and HBM/DRAM scaling all convert into advanced-node wafer capacity, and advanced-node capacity converts into lithography. Geopolitical fragmentation is duplicative — the same design gets fabbed in more places — which is net additive to tool counts even as it caps the China channel. Offsetting: a 4.63% 10-year raises the hurdle on multi-billion fab projects and macro headwinds could stretch capex phasing, and the current cycle leg favours deposition/etch/metrology suppliers before litho. Net structure: demand pull is real and multi-year, but its slope is cyclical and far below what the price assumes.
When we made this prediction on Aug 18, 2026, ASML was $1,883.12. We expect it to be $1,720.00 by Feb 2027, and we consider it great value under $1,250.00. This is an early model (v0.6.0) — the direction is more reliable than the exact price. Made Aug 18, 2026.
Blue is our prediction, starting the day we made it. Grey is a slower route to the same place — the same destination, taking longer. Black is the actual price, so you can see how we are doing.