ASML 11 min read

ASML Holding N.V. (ASML) — Investment Tree v1

✅ FULL R2 RE-ANCHOR COMPLETE (2026-05-30) — Bucket-5. Scaffold anchored ~€950†; re-anchored to the live USD line $1,612.76 (2026-05-29 close; ~+56% incl. EUR/USD). Anchor switched to the USD line per the foreign-issuer convention (ordinary shares 1:1 with the NY registry shares — not a depositary ADR). Primary sources now Tier-A: ASML FY2025 results 6-K (accn 0001628280-26-003701) + 20-F (accn 0001628280-26-011378) + Q1 2026 6-K (accn 0001628280-26-025147). Finding: the EUV monopoly is intact and Q1 2026 reinforced it (€8.8B sales, 53% GM, FY2026 guide RAISED to €36-40B, €38.8B year-end backlog) — but the "cyclical-equipment lens underprices the structural monopoly" mispricing has PLAYED OUT. At ~47-50× forward FY2026 EPS the market now pays the structural-monopoly multiple in full; Street median PT ~$1,660 (+2.9%, ≈ at spot). Re-derived band Bull $1,980 / Base $1,620 / Bear $950 → EV −1.2%, asymmetry 0.56× UNFAVORABLE (was +29% / 1.55× favorable at €950). Verdict: Hold quality at full price; new capital WAIT for a pullback (~$1,300-1,400) to restore the margin of safety. Own for the AI-monopoly convexity — do not chase. Structural note: ASML discontinued quarterly net-bookings disclosure in 2026 (the old T1 monitoring metric is dead — annual backlog now). See evidence_2026-05-30.jsonl + scenarios.md + update_2026-05-30.md. The original 2026-05-19 Tier-C build + 2026-05-22 partial-verification are retained in evidence_2026-05-19.jsonl / evidence_2026-05-22.jsonl as the historical baseline.

Stage 7 final essay. Bilingual companion: tree_v1_zh.md. Date: 2026-05-30 (R2 re-anchor) · Anchor: $1,612.76 (US line, 2026-05-29 close) ≈ €1,466 at EUR/USD ≈1.10 · Forward P/E: ~47-50× FY2026E · Dividend yield: ~0.46% Archetype: EUV monopoly with AI-capex super-cycle leverage and China DUV overhang. Closest analogues: TSM 2020-2024 (semiconductor monopoly + AI-cycle leverage); NVDA 2022-2024 (hyperscaler pull-through).

SOURCE QUALITY: R2-VERIFIED Tier-A 2026-05-30 (FY2025 6-K/20-F + Q1 2026 6-K — see evidence_2026-05-30.jsonl). Company financials reported in EUR (ASML's reporting currency); scenario targets + anchor in USD (the line Ming buys). Only forward 2028E scenario EPS remain analytical (†).

0. Company Fundamentals — what ASML is and how it earns

Figures FY2025 (calendar 2025) unless noted; reported in EUR, USD at EUR/USD ≈ 1.10. The line Ming buys is the US listing (ASML).

What it is & how it earns. ASML is the sole maker of EUV (extreme-ultraviolet) lithography machines — and the leading maker of DUV — that every leading-edge chip requires; Canon and Nikon abandoned EUV in 2015, leaving ASML a 100% monopoly at the <7nm node with no credible competitor on a 10-year horizon. It earns two ways: selling systems (EUV €180-380M each, DUV €50-90M) and a high-margin Installed Base Management annuity (service + upgrades on a 200+ EUV installed base). FY2025 net sales were €32.7B (+15.6%) ≈ $35.9B — net system sales €24.5B (EUV €11.6B/+39%; DUV €12.0B/−6%) plus IBM €8.2B (+26%). Top-3 customers TSMC, Samsung & Intel are ~75-80% of EUV revenue; China was ~29% of FY2025 sales (down from ~36%) under US/Dutch export controls.

Cash-flow anatomy. Cash generation is strong but lumpy — system timing and customer down-payments swing it year to year:

Gross margin is structurally high (FY2025 52.8%) and FY2025 net income was €9.6B (EPS €24.73). (FY2025 cash-flow figures are triangulated pending the 20-F cash-flow statement.)

Balance sheet & capital allocation. ASML runs a net-cash balance sheet: cash & short-term investments ~€13.3B against total debt of only ~€2.7B. Capital return: a FY2025 dividend of €7.50/share (+17%) — but a ~0.46% yield at the ~$1,613 anchor, so not a yield name — plus a new €12B buyback program through 2028 (the prior €12B program ended Dec 2025 having repurchased ~€7.6B). R&D of €4.7B (~14% of sales) funds the High-NA roadmap and is the moat.

What drives it. Leading-edge EUV/High-NA demand riding the AI-capex super-cycle (FY2026 guide raised to €36-40B; year-end backlog €38.8B, >50% EUV) set against the China export-control overhang and a full ~47-50× forward multiple. The question the tree resolves: is the structural monopoly already fully paid for at this price?

I. One-sentence verdict

ASML at $1,612.76 is the global monopoly on EUV lithography — the structural single-point-of-presence at the <7nm-node where hyperscaler AI-silicon demand (TSM 2nm + Intel 18A + Samsung 2nm) drives sustained EUV pull-through 2025-2028, High-NA EUV commercializes (first EXE:5200B revenue recognized 2025), China DUV revenue floors ~19-20% of system sales under the export-control regime, and durability is 23/25 HIGH with 0 fatal flags — but at ~47-50× forward FY2026 EPS the cyclical-lens-underprices-the-monopoly mispricing has PLAYED OUT (EV −1.2%, asymmetry 0.56× unfavorable, Street PT +2.9%), so this is now a Hold quality-at-full-price: existing holders keep it for the AI-monopoly convexity; new capital WAITS for a pullback to ~$1,300-1,400; cap 4-5% under the ai-capex-high K.4 sizing rule — but do not deploy fresh ai-capex-bucket capital here at this entry.

II. Company snapshot

ASML Holding N.V. is the global monopoly on extreme ultraviolet (EUV) lithography systems — the only commercially-viable supplier of <7nm-node lithography equipment. FY2025 (Tier-A): total net sales €32.667B (+15.6%), gross margin 52.8%, net income €9.609B, basic EPS €24.73. Revenue split: net system sales €24.474B (EUV €11.6B +39% on 48 systems; DUV €12.0B −6% on 279 systems; metrology/other ~€0.9B) + Installed Base Management (service/upgrades) €8.193B (+26%). Year-end 2025 backlog €38.797B (>50% EUV). R&D €4.7B.

Three classes of lithography: EUV (€180-380M per system), DUV immersion (€50-90M), DUV dry. First High-NA EXE:5200B revenue recognized 2025 (2 High-NA systems in Q4 2025). EUV installed-base ~200+ systems globally; growing ~40-48 systems/yr. Top-3 customers (TSM, Samsung, Intel) = ~75-80% of EUV revenue. China ~29% of FY2025 net sales (down from 36% FY2024; ~19% of Q1 2026 system sales under US/Dutch export controls). Capital return: €7.50/share dividend (+17%) + new €12B buyback through 2028.

Management: Christophe Fouquet CEO (since April 2024 from CCO role); Peter Wennink retired. Roger Dassen CFO. Dutch-domiciled (Veldhoven HQ).

III. Five facts that drive everything

1. ASML is the only commercial EUV supplier globally. Canon + Nikon abandoned EUV 2015. No credible competitor on 10-year horizon. ✅C
2. EUV systems gross margin ~55-60%; recurring services scale with installed-base. Installed-base of 200+ systems generating ~€2B/yr services at ~45% gross margin. ✅C
3. High-NA EUV commercializing 2024-2028 enables <2nm-node single-exposure lithography. TSM, Samsung, Intel all committed. ✅C
4. AI-capex super-cycle (2024-2028) drives sustained EUV pull-through. Hyperscaler aggregate capex ~$340B+ FY2025 → TSM/Samsung/Intel 2nm-node ramp → ASML EUV pull. ⚠️C — cycle-dependent.
5. China revenue mix declining from 46% (FY2024) to ~25-26% (FY2026E) under US/Dutch export controls; floor at ~$4-5B/yr. ✅C bounded.

IV. H-0 thesis (embedded)

H-0: ASML at $1,612.76 is the global EUV monopoly; AI-capex super-cycle drives sustained EUV pull-through 2025-2028 via TSM 2nm + Intel 18A + Samsung 2nm ramp; High-NA EUV commercializes successfully 2026-2028; China DUV revenue floors ~19-20% of system sales; consensus oscillates between cyclical-capital-equipment lens (25-30× forward PE) and structural-monopoly lens (35-45× forward PE) — the AI-capex super-cycle has transitioned ASML from cyclical to structural.

Mispricing taxonomy: Interpretation × Structural. (R2 2026-05-30: the interpretation gap has CLOSED — the market now applies the structural-monopoly multiple in full, ~47-50× forward. The mispricing is played out; the thesis was right and the trade is over at this entry.)

H-0 confidence post-Stage 3: ~78% (post-2026-05-30 R2 re-run: ~68%). The monopoly + AI-pull-through legs are MORE confirmed by Q1 2026 (€8.8B sales, 53% GM, FY2026 guide raised to €36-40B), but two things trimmed confidence from the build's 78%: (1) the 2026-05-25 update's China cliff (system-sales mix 29%→19% faster than the floor expected) + TSM High-NA broader-deployment push to "at least 2029"; and (2) the "market underprices the structural transition" sub-claim is now falsified — the re-rating to ~47-50× forward closed the gap.

Falsification:

• FF1 AI-capex digestion (2-of-4 overlay triggers fire) → cycle compression
• FF2 TSM-specific disruption → customer concentration cascade
• FF3 High-NA EUV technical hurdle / delay
• FF4 Further China export-control tightening collapses DUV floor
• FF5 SMEE breakthrough (extremely low probability)

V. Tree — six branches (embedded; see leaves.md)

A) EUV monopoly durability ✅C strong (90% confidence) B) High-NA EUV commercialization ✅C strong (75%) C) DUV + China exposure ✅C partial (75%) D) AI-capex pull-through ✅C partial — LOAD-BEARING (70%) E) Customer concentration ⚠️C partial (65%) F) Capital allocation ✅C strong (85%)

Verdict tally: 13 ✅ · 7 ⚠️ · 0 ✗ · 0 ⊗ (unchanged post-2026-05-30 R2 re-run — the leaves are business questions and the re-rating changed the valuation verdict, §XI/§XII, not the leaf evidence. The 2026-05-25 update kept the tally flat: the China cliff + High-NA pushback are trajectory shifts within ⚠️ leaves, not state transitions.)

VI. Market consensus + mispricing

Consensus (2026-05-29): Street "Strong Buy"; median 12-month PT ~$1,660 (+2.9% from $1,612.76) — i.e. essentially at spot. Narrative: "AI-capex super-cycle accelerating EUV pull-through; ASML monopoly durable." The Street now sees ASML as fairly-to-fully valued, leaving little median-case headroom.

Mispricing (R2 2026-05-30 — PLAYED OUT): the interpretation lens — ASML oscillated between the cyclical-capital-equipment multiple (25-30× PE) and the structural-monopoly multiple (35-45× PE). The build's thesis was that the AI-capex super-cycle had transitioned ASML to structural and the market hadn't fully repriced it. At $1,612.76 the market now pays ~47-50× forward FY2026 EPS — at or above the top of the structural-monopoly band. The gap is closed: the thesis was correct and has been realized in the price.

Reverse-engineered market-implied probability at $1,612.76 is now ~32/50/18 vs the Stage-4 explicit 30/50/20 — the ~25pp bear-probability gap that existed at €950 (market ~45% bear) has been arbitraged away by the +56% re-rating. See implied_prob.md.

VII. Scenarios (embedded; see scenarios.md — R2 re-anchor 2026-05-30, USD)

Expected value: $1,594; Prob-weighted return: −1.2%; Asymmetry: 0.56× UNFAVORABLE (was +29% / 1.55× favorable at €950 — the +56% re-rating consumed the upside). Street median PT ~$1,660 (+2.9%) corroborates "roughly fairly valued, little headroom."

INDEX_META prob: 30/50/20 (business probabilities unchanged; the EV/asymmetry move is the price re-rating, not a thesis change). Targets in USD (anchor currency); EUR at EUR/USD ≈1.10.

VIII. Risks

Cyclical risk (HIGH; LOAD-BEARING). ai-capex-high cycle exposure per ai_capex_cycle_overlay.md. K.4 sizing rule applies.

Customer concentration (MOD-HIGH). TSM ~30-35% single largest; top-3 ~75-80%.

Geopolitical (MOD). Taiwan + Korea + US/Dutch export-control regime durable.

Execution (LOW). High-NA EUV ramp execution. Strong track record.

Regulatory (MOD). Export controls + China revenue evolution.

Correlated-factor risk (HIGH). cycle_exposure: ai-capex-high — coordinate sizing with NVDA+TSM+AMD+AJNMY.

IX. Historical analogues

TSM 2020-2024 (semi-monopoly + AI cycle leverage). Comparable critical-input monopoly; ASML upstream of TSM.

NVDA 2022-2024 (AI super-cycle). Downstream of TSM; ASML upstream of both. NVDA's cycle-leverage demonstrates the AI-capex super-cycle pattern.

Microsoft 2018-2022 (perpetual-to-subscription reframe). Comparable structural reframe — market reframes ASML from cyclical to structural.

X. When H-0 fails (USD targets, R2 2026-05-30)

Scenario 1: AI-capex digestion (most severe). 2-of-4 overlay triggers fire; ASML EUV pull-through compresses; multiple re-rates to the cyclical lens. Target ~$850-1,000 (≈€770-910) — the dominant bear path.

Scenario 2: TSM disruption. Geopolitical or manufacturing disruption cascade at the single largest customer. Target ~$900-1,050 (≈€820-955).

Scenario 3: High-NA EUV delay (mildest — timing, not technical). TSM broader deployment already pushed to "at least 2029" on cost; a further slip shifts mix toward Low-NA. Target ~$1,300-1,450 (≈€1,180-1,320) — a haircut, not a collapse (the €38.8B backlog + recurring IBM revenue cushion it).

XI. Final verdict (R2 re-anchor 2026-05-30)

Hold quality at full price — existing holders KEEP for the AI-monopoly convexity; new capital WAIT for a pullback to ~$1,300-1,400. Cap 4-5% under the ai-capex-high K.4 sizing rule (coordinate with NVDA+TSM+AMD+AJNMY for the ai-capex bucket <15%) — but at this entry ASML should NOT be the place fresh ai-capex-bucket capital goes.

At €950 this was a Buy/accumulate (+29% prob-weighted return, 1.55× favorable). At $1,612.76 the +56% re-rating consumed that: EV −1.2%, asymmetry 0.56× UNFAVORABLE, Street median PT +2.9%. The business is unchanged-to-stronger (Q1 2026 €8.8B sales / 53% GM / FY2026 guide raised to €36-40B / €38.8B backlog; durability 23/25 HIGH, 0 fatal flags — still best-in-class). What changed is the price: the structural-monopoly multiple the build said was coming has arrived in full. This is the COST/AVGO pattern — a wonderful business at a now-full price. The trade is over at this entry; the business is not. Re-engage on a cycle-driven pullback.

XII. Investment Scorecard (per K.6)

15-question scorecard

K.3.5 derivation (R2 re-anchor 2026-05-30)

The 15-question scorecard measures structural quality, which the re-rating does not change — except Q12 (price reasonable), which drops 4/5 ✅C → 2/5 ⚠️C (entry full at ~47-50× forward; asymmetry 0.56× unfavorable; Street PT +2.9%). Everything else holds: monopoly, balance sheet, ROIC 28-32%, moat, FCF visibility are unchanged-to-stronger after Q1 2026.

• Critical: 5+5 = 10 × 5 = 50
• Load-bearing: 4+5+5 = 14 × 3 = 42
• Important: 5+5+4+5 = 19 × 2 = 38
• Confirming: 4+3+2+5+4+5 = 23 × 1 = 23 (Q12 4→2)
• TOTAL: 50+42+38+23 = 153

Max: 50+45+40+30 = 165.

xii_score = 153 / 165 = 92.7% → 93% (per /39 normalized: 36.2 / 39 = 93%). Still a high-quality structural score — the durability (23/25) and business quality are intact; only the price-reasonableness leg moved on the +56% re-rating. INDEX_META updated 94% → 93%. The gap between xii_score (93%, structural quality) and h0 (68%, thesis confidence) is the headroom that has now CLOSED on the valuation side — the quality is real, but the cheap entry that made it actionable is gone.

Final verdict (R2 re-anchor 2026-05-30)

Hold quality at full price. Existing holders KEEP (AI-monopoly convexity; 23/25 durability, 0 fatal flags); new capital WAITS for a pullback to ~$1,300-1,400. Cap 4-5% under the ai-capex-high K.4 rule — but do not deploy fresh ai-capex-bucket capital at this entry (EV −1.2%, asymmetry 0.56× unfavorable).

2-minute pitch

ASML is the global monopoly on EUV lithography — the critical-input single-point-of-presence at the <7nm-node where hyperscaler AI-silicon demand (TSM 2nm + Intel 18A + Samsung 2nm) drives sustained pull-through 2025-2028. Canon + Nikon abandoned EUV 2015; SMEE/China-domestic is multi-decade behind; High-NA EUV is commercializing (first EXE:5200B revenue 2025). Q1 2026 reinforced it: €8.8B sales, 53% GM, FY2026 guide raised to €36-40B, €38.8B backlog. Durability 23/25 HIGH; 0 fatal flags; xii_score 93%. But at $1,612.76 (~47-50× forward) the market now pays the structural-monopoly multiple in full — the cyclical-lens-underprices-the-monopoly mispricing has PLAYED OUT: EV −1.2%, asymmetry 0.56× UNFAVORABLE, Street PT +2.9%. cycle_exposure: ai-capex-high — coordinate with NVDA+TSM+AMD+AJNMY per K.4. Verdict: a wonderful business at a now-full price (COST/AVGO pattern) — Hold for existing holders; new capital waits for a pullback to ~$1,300-1,400.

Risk types

Cyclical HIGH (LOAD-BEARING); Customer concentration MOD-HIGH; Geopolitical MOD; Execution LOW; Regulatory MOD; Correlated-factor HIGH (ai-capex-high); Valuation MOD-HIGH (NEW — entry full, asymmetry inverted).

When NOT to buy

• ⚠️ Entry full at ~47-50× forward; asymmetry 0.56× unfavorable, Street PT only +2.9% (R2 2026-05-30 binding) — new capital waits for ~$1,300-1,400
• ⚠️ AI-capex digestion in progress (2-of-4 triggers firing)
• ⚠️ ai-capex bucket sum already >15% portfolio (K.4 sizing rule applies)

Long-term holdability

Strongly qualified for long-term hold (5-10 yr). 23/25 HIGH durability. Best-in-class structural monopoly. Sizing must coordinate with other ai-capex-high names.