ASML Holding N.V. (NASDAQ: ASML) — Initiating Coverage NEUTRALScenario Fair Value$1900

1 · Summary & Verdict

FY2025 revenue €32.7B (+16%), then the July 15 print raised 2026 guidance to €43-45B (+35% at mid) with EUV and DUV immersion capacity +30% for 2027. Our Scenario Fair Value of $1,900 sits ~5% above the $1,815 price; we initiate Neutral.

THE THREE RISKS THAT MATTER

2 · Investment Thesis

1. The strongest moat in technology, and it widens every node

ASML is the only EUV supplier on earth (100% share, ~250 systems installed), holds >90% of immersion DUV, and litho intensity rises with every node transition — N5→N3→N2 adds EUV layers, DRAM 1γ and HBM4 insert EUV into memory, and High-NA (2x ASP) begins the next cycle. The moat is a 5,000-supplier ecosystem (Zeiss optics, Cymer light sources) plus two decades and >€10B of program investment no entrant can compress.

• EUV revenue +39% in 2025 to €11.6B (48 systems); 60+ low-NA shipments planned 2026
• Intel shipped the first high-volume High-NA logic product (Panther Lake, 18A) in July 2026

2. The 2026-27 ramp is underwritten by commitments, not sentiment

Guidance went €34-39B → €36-40B → €43-45B in six months on 'extremely strong' H1 order intake and customer long-term agreements. ASML is adding +30% low-NA EUV and +30% DUV immersion capacity for 2027 — a supplier with 100% share does not build capacity speculatively; it builds against committed demand. Backlog was €38.8B before the 2026 order surge.

• FY26E revenue $50.2B (+36%), FY27E $63.2B (+26%) at 56.5% gross margin
• Q3-26 guided €11.0-12.0B at 55-57% GM; implied Q4 ~€15B is a new all-time-record quarter

3. Margin structure is re-rating: 51% → 54-56% → high-50s

Three stacked mix effects lift gross margin: EUV share of systems rising (38% in 2025, ~55%+ by 2028E), High-NA ASPs (~€400M) carrying premium margin as the platform matures, and the service/upgrade annuity (IBM ~28% of revenue, growing >25%/yr) which carries the highest margins in the company. We model 55.0% FY26E → 57.5% FY30E; management's own 2030 frame (56-60%) brackets us.

• Operating margin: 34.6% FY25A → 40.5% FY26E → ~45% FY30E
• EPS: $27.92 FY25A → $45.19 FY26E → $88.18 FY30E (26% CAGR)

4. The valuation tension is the thesis: quality is not in question, entry price is

At $1,815 ASML trades at 64x TTM earnings vs a 40x 8-year mean (+2.5σ) and 19x sales (+2.9σ). Our DCF says $1,292 base, $988-$1,898 across the grid. The gap between intrinsic-cash-flow value and price is the market's bet on the 2028+ capacity option, High-NA's memory insertion, and CMD-2027 re-baselining. We are happy owners of the story at a Neutral entry discipline — and buyers on any digestion-driven derating.

• SFV $1,900 = 30.7x FY27E / 25.7x FY28E EPS — demanding but defensible for a monopoly
• View shifts Positive below ~$1,650 (+15% upside); Cautious above ~$2,000 (bull case fully priced)

3 · Financial Analysis

ASML reports US GAAP in euros; our model is USD-denominated so the financials read coherently against the NASDAQ listing (actuals converted at 1.08 for FY23-24 and 1.13 for FY25 average rates; projections at 1.14). Where we quote reported figures, euros come first.

Revenue: two flat years, then the coil released. Revenue was essentially flat across FY2023-24 (€27.6B → €28.3B; $29.8B → $30.6B) as the industry digested the 2023 China DUV pull-forward and waited on the AI capex cycle to reach lithography. FY2025 broke the range: €32.7B (+15.6%), driven by EUV system revenue up 39% to €11.6B on 48 systems, while DUV declined 6% to €12.0B — exactly the mix rotation the long thesis wants. The definitive tell on business quality: units sold fell from 449 (2023) to 327 (2025) while revenue rose 19% — ASML sells fewer, vastly more valuable machines every year.

The mix beneath the total. Three revenue engines now pull in sequence. EUV ($13.1B in FY25A, 36% of revenue) is the growth engine, compounding ~27%/yr in our model through FY30E. DUV + metrology ($14.5B, 39%) is the cash cow — flattish, with China normalization offset by mature-node breadth. Installed Base Management ($9.3B, 25%) is the annuity: +26% in 2025, +28% YoY in H1-2026, and structurally the highest-margin line in the company. Geographically, 2024’s China peak (36% of revenue) normalized to 29% in 2025, with Taiwan (~30-35% in recent quarters) and Korea re-taking share as N2 and HBM capacity ramps — concentration that mirrors where the leading edge physically lives.

Margins: the 51% floor is broken. Gross margin sat at 51.3% in both FY2023 and FY2024 — the old “EUV dilutes margin” regime. FY2025 printed 52.8%, Q1-26 53.0%, Q2-26 54.0% (above guidance, on service strength), and H2-26 is guided to 55-57%. The drivers are structural, not cyclical: EUV ASPs and maturity, High-NA premium pricing, and the IBM annuity mixing up. Operating margin followed — 32.8% (FY23A) → 34.6% (FY25A) → 36.6% in H1-2026 — while ASML continued to grow R&D to €4.7B (FY25), five times what Nikon and Canon spend on lithography combined.

Earnings and cash: violent quarters, smooth years. FY2025 net income was €9.6B (+27%), EPS €24.71 diluted ($27.92 in our USD model). Cash generation is lumpy by construction — customer prepayments and system-acceptance timing produced Q4-25 CFO of €11.4B followed by Q1-26 CFO of negative €2.2B — but through-cycle conversion is excellent: FY2025 CFO of €12.7B ($14.3B) against capex of just €1.6B, for ~$12.5B of FCF (34% margin). The balance sheet is a fortress: €13.3B cash + short-term investments against €3.6B total debt at YE-2025, even after returning €8.5B to shareholders during the year (€2.55B dividends + €5.95B buybacks) and writing a €1.3B strategic check into Mistral AI.

Capital returns are accelerating. The 2025 dividend of €7.50/share was +17% YoY; the first 2026 interim was set at €1.88/quarter (+17.5% run-rate); and a fresh €12B buyback (2026-2028) replaced the completed 2022-25 program, retiring ~0.8M shares (€1.1B) in Q2-26 alone. Basic share count has fallen from 393.8M (FY23) to 384.5M (Q2-26). Net income compounding plus a shrinking denominator is the quiet second engine of our $88 FY30E EPS.

4 · Projection Assumptions

Our model runs FY2026E-FY2030E in USD (projections at 1.14 EUR/USD). The five-year shape: a step-function 2026, a capacity-fed 2027, then deliberate deceleration into a digestion-aware but structurally growing back half.

FY2026E — the guided step-function: revenue $50.2B (+35.9%). We sit at the midpoint of the raised €43-45B guide. The arithmetic behind it is visible: H1-2026 actual revenue was €18.1B, Q3 is guided €11.0-12.0B, leaving an implied Q4 near €15B — a record by roughly 50%. Management raised full-year gross margin guidance to 54-56%; we model 55.0%. R&D runs ~€1.2B/quarter and SG&A ~€0.4B/quarter per guidance, giving a 40.5% operating margin and EPS of $45.19 (+62%). The tax rate is guided ~17%.

FY2027E — the +30% capacity year: revenue $63.2B (+26%). ASML’s committed 2027 output — low-NA EUV ~85 systems (+30%), DUV immersion ~170 (+30%), High-NA ramping toward ~16 systems — supports revenue in the €54-57B range; we model €55.4B ($63.2B). Mix (EUV to ~47% of revenue) plus High-NA maturity lifts gross margin to 56.5%. EPS: $61.92 (+37%).

FY2028E-FY2030E — deceleration by design, not collapse: +15% / +8% / +6.5%. We deliberately do not extrapolate +30% capacity additions forever. FY2028E revenue of $72.7B assumes the investigated 2028 capacity expansion partially converts; FY2029E (+8%) models a digestion year inside an intact structural uptrend; FY2030E reaches $83.6B — comfortably above the top of ASML’s own (stale, pre-supercycle) €44-60B 2030 model, which the June-2027 CMD is likely to re-baseline. Gross margin plateaus at 57.0-57.5%: the EUV/High-NA/service mix supports it, and we resist modeling management’s 60% aspiration until demonstrated.

High-NA is the swing factor inside the projection. We carry High-NA revenue from ~$0.9B (FY25A, two EXE:5200 recognitions) to ~$4.6B (FY26E, ~10 systems) to ~$14.5B by FY30E — by then ~17% of total revenue. This embeds Intel scaling 18A/14A insertions, Samsung deploying EXE:5200B for SF2 and HBM4, and TSMC adopting only at decade-end (A14P/A10, per its stated plan). A TSMC pull-forward is upside to these numbers; a High-NA cost-per-exposure stall is the main downside.

Opex and capital intensity. R&D grows from $5.6B (FY26E) to $8.0B (FY30E) — ~9.5% of revenue, funding the EXE roadmap and hyper-NA research; SG&A stays under 3% of revenue. Capex steps up to ~$2.7-3.3B/yr through the capacity build (Veldhoven expansion, Berlin optics, San Diego light sources) before normalizing — still under 5% of revenue, because ASML’s suppliers carry most of the fixed-asset burden of its ramps.

Cash flow and returns. We model CFO rising from $17.6B (FY26E) to $34.8B (FY30E) with FCF margins near 30-38%; working capital consumes cash in the ramp years (receivables and inventory build with the Q4-26 shipment wave) and releases later. Capital returns scale with the balance sheet: dividends grow ~15%/yr and buybacks step from the current €12B program toward ~$10-14B/yr by decade-end, taking diluted shares from 388.9M (FY25A) to ~362M (FY30E). We deliberately model returns conservatively; special dividends or an enlarged program at the 2027 CMD are upside optionality.

5 · Scenario Analysis

We model three explicit five-year paths through FY2030E, weighted 25% bull / 55% base / 20% bear. The blend produces a Scenario-Weighted Fair Value of ~$1,905, which we publish as a $1,900 Scenario Fair Value (the base-case anchor). The spread is wide — $1,050 to $2,600 — because semicap outcomes genuinely are: this is an industry where order books can double or halve inside four quarters, monopoly or not.

Base case (55%, $1,900). The guided 2026 lands at €44B; 2027 grows 26% on the committed +30% capacity; 2028-30 decelerates to +15%/+8%/+6.5% as the AI buildout matures without breaking. Gross margin walks 55.0% → 57.5%. EPS: $45.19 → $88.18 (26% CAGR). $1,900 equates to 30.7x FY27E and 25.7x FY28E earnings — a premium to the semicap group, justified by monopoly economics, but no longer a premium to ASML’s own recent trading range. The base case is, deliberately, management’s contracted plan minus the 2028 optionality.

Bull case (25%, $2,600). The supercycle extends: the investigated 2028 +30% capacity converts to commitments, TSMC pulls High-NA forward into A14P, HBM4/1γ DRAM EUV intensity surprises, and China DUV stabilizes. Revenue grows +40%/+32%/+22%/+14%/+12%, hitting ~$106B by FY30E with gross margin reaching 60% (management’s own 2030 aspiration). EPS reaches ~$121 by FY30E. $2,600 is ~29x the bull FY28E EPS of $89 — demanding, but this is the scenario in which ASML posts three consecutive +25% years and the market treats the 2030 CMD model as a floor. Q4-26 backlog >€50B would be the first confirmation.

Bear case (20%, $1,050). The classic semicap air pocket, AI edition: 2026 finishes at the low end (€43B, still +32% — the year is already contracted), but 2027 orders stall as hyperscaler capex digests, memory over-builds HBM, and China service restrictions bite. Revenue: +34%/+4%/−12%/+2%/+8%; gross margin troughs at 50% as under-utilized capacity and pricing concessions bite; EPS bottoms near $32 in FY2028E before recovering to $39 by FY2030E. $1,050 is ~27x the recovered FY30E bear EPS — bear-case multiples stay elevated because the monopoly survives every downcycle; it is the entry price that determines whether the holder does. Note the asymmetry this creates at a $1,815 entry: -42% to the bear anchor vs +43% to the bull.

What moves us between scenarios. Watch, in order: (1) the Q4-26 backlog print and initial 2027 guide (bull needs >€50B / €53-57B); (2) the 2028 capacity decision — commitment converts bull probability, quiet deferral converts bear; (3) High-NA order flow beyond Intel/Samsung; (4) hyperscaler capex guidance revisions, the ultimate upstream signal; (5) export-control headlines in either direction. We would re-weight bull above 25% on any two of the first three confirming.

6 · Valuation — and the Lab

We anchor our Scenario Fair Value of $1,900 on the base case of an explicitly modelled bull/base/bear blend (25/55/20 weights compute to $1,905), cross-checked against an intrinsic DCF that deliberately refuses to chase the price, and trading comparables that place ASML inside — not above — the US semicap complex on forward earnings. Precedent transactions are not a meaningful method here: nothing at ASML’s ~$700B scale and monopoly position has ever traded, and no regulator on three continents would allow it to.

DCF — perpetuity on normalized FCF ($1,292/share). We discount four years of explicit unlevered FCF (FY27-30E: $21.5B / $26.0B / $28.8B / $31.3B) at a 9.5% WACC (cost of equity 9.9% = 4.4% risk-free + 1.15β × 4.75% ERP; near-irrelevant debt at a 98/2 weighting), then apply Gordon growth on a normalized $33.5B FCF at g = 3.5% — a terminal growth rate we set above the usual 2.5-3.0% because lithography intensity and monopoly pricing power justify structural growth above global GDP. PV of explicit FCF: $85.0B. PV of terminal value: $402.0B (82.5% of EV — high, and honestly disclosed: this is a duration asset). Enterprise value $487.0B, plus net cash $6.4B, over 382M diluted shares → $1,292 per share.

DCF sensitivity — the grid cannot reach the price. Across the full 5×5 WACC × growth grid (8.5-10.5% × 2.5-4.5%), per-share value spans $988 to $1,898 (the interactive heatmap in the Valuation Lab below maps the full envelope). The friendliest corner — 8.5% WACC, 4.5% terminal growth, assumptions we would not defend individually — still lands below the $1,815 close. That is the single most important sentence in this report: on standard intrinsic-cash-flow assumptions, the market price already embeds the 2028 capacity option, High-NA’s memory insertion, and a CMD-2027 re-baseline. We do not treat that as a short signal — monopoly duration deserves terminal-value premium — but it disciplines the rating.

Trading comparables. The natural peer set is the US semicap triumvirate (AMAT, LRCX, KLAC), the customer-anchor TSM, and Tokyo Electron. Peer forward P/E runs 21.6x (TSM) to 46.5x (KLAC), median ~36.8x; ASML trades at ~36x forward consensus — inside the US peer range despite monopoly economics and a growth rate (FY27E EPS +37%) nearly double the peer median (~15%). On the PEG view this is the cheapest expensive stock in the group: the premium P/E is fully paid for by growth. Applying 25-35x to our FY27E EPS of $61.92 brackets $1,548-$2,167; peer-median EV/EBITDA logic (18-26x FY27E EBITDA of $29.3B, bridged over net cash) brackets $1,415-$2,040.

Football field & the blend. Method ranges: DCF perpetuity $990-$1,900 (base $1,292); Forward P/E comps $1,550-$2,165; EV/EBITDA comps $1,415-$2,040; scenario range $1,050-$2,600. The scenario blend (25% × $2,600 + 55% × $1,900 + 20% × $1,050 = $1,905) is our published anchor: $1,900 Scenario Fair Value, sitting near the top of the DCF grid, mid-range of both comps methods, and at the base scenario by construction.

Street context. Post-print sell-side consensus clusters near $1,950 (range roughly $1,500-$2,400, with the high end refreshed within hours of the July 15 guidance raise). Our $1,900 sits at consensus — the differentiation of this report is not the number, it is the explicit accounting of which future the number buys.

Rating logic. Kaamos View bands anchor on scenario fair value vs price: Positive above +15% upside, Neutral between -5% and +15%, Cautious below -5%. At +4.7% ($1,900 vs $1,815.27), ASML is Neutral. The View shifts Positive below ~$1,650 (where upside exceeds 15% and the bear case is partially pre-paid) and Cautious above ~$2,000 (where price exceeds our fair value and the bull case is the operative assumption).

Which multiples to actually trust for ASML

P/E (TTM) — primary. The street’s anchor for a monopoly compounder: earnings are high-quality (no commodity write-down cycle, IP-driven pricing, a service annuity underneath), so the earnings multiple carries real signal. Right now it reads 64x TTM against a ~40x 8-year mean (+2.5σ) — the most extended it has been in the band’s history. The honest caveat: TTM earnings lag a violent inflection; on our FY27E the same price is ~29x. The z-score is telling you how much of that future is pre-paid, not that earnings are about to disappoint.

EV/EBITDA (TTM) — secondary. Controls for the net-cash fortress balance sheet and reads the operating inflection fastest, since TTM EBITDA already carries the H1-26 step-up. At 33.5x vs a 30.1x mean (+0.5σ) it is the least alarmed multiple on the board — useful as the counterweight to the P/E signal, and the reason we are Neutral rather than Cautious.

P/Sales (TTM) — secondary. The purest read on how much monopoly is being capitalized, immune to margin flattery: 19.0x vs an 10.7x mean (+2.9σ) is the single most honest “expensive” datapoint we have. When gross margin structurally re-rates (51% → 56%), some P/S expansion is deserved; +2.9σ is more than “some.”

Excluded. P/B (31x, +1.9σ): an IP-and-ecosystem business where book value mostly reflects buyback accounting — economically meaningless. P/CF and P/FCF: customer prepayments whipsaw cash flow at quarterly and even annual granularity (Q4-25 CFO €11.4B; Q1-26 −€2.2B); our DCF normalizes through-cycle FCF instead of trusting any single reading. Dividend yield (0.4% vs 0.8% mean): a payout-policy artifact compressed by the €12B buyback — not a valuation anchor for a growth monopoly.

Net read: across the multiples that matter, ASML prices between +0.5σ (EV/EBITDA) and +2.9σ (P/S) above its own history, with the earnings-based anchor at +2.5σ. The market has pre-paid the contracted 2026-27 ramp; the bull/bear debate from here is entirely about 2028 and beyond. That is precisely a Neutral setup: right company, full price.

7 · Company

Business. ASML Holding N.V. (Veldhoven, the Netherlands; ~44,500 FTEs) supplies photolithography systems — the machines that project circuit patterns onto silicon — to every major chipmaker on earth. It is the sole producer of extreme-ultraviolet (EUV) lithography systems, without which no leading-edge logic or advanced DRAM chip can be economically manufactured, and it holds >90% of the immersion DUV market beneath. Revenue splits three ways: EUV systems (~36% of FY25 revenue, ASPs €235-400M+ per machine), DUV + metrology/inspection systems (~39%), and Installed Base Management — service and field upgrades on the >5,700 systems shipped since inception (~25%, the fastest-compounding line). Customers are the five companies that define the leading edge — TSMC, Samsung, SK Hynix, Intel, Micron — plus the mature-node universe.

History. Founded 1984 as a Philips/ASM International joint venture; IPO’d 1995 (Amsterdam + NASDAQ). Three acquisitions built the modern stack: Cymer (2013, EUV light sources), Hermes Microvision (2016, e-beam inspection), Berliner Glas (2020, optics). The EUV program — two decades and >€10B, part-funded by Intel, TSMC and Samsung through the 2012 Customer Co-Investment Program — produced a literal monopoly: every EUV production system in existence (~250) is an ASML machine, with Carl Zeiss SMT (24.9% ASML-owned) as irreplaceable optics partner. Christophe Fouquet (CEO since April 2024) now runs the company through its steepest ramp ever; CFO Roger Dassen architected the capital-return program (dividend €1.40 → €7.50 over 2017-25; €12B buyback through 2028).

Financial scale. FY2025: revenue €32.7B (+15.6%), gross margin 52.8%, net income €9.6B, backlog €38.8B, €8.5B returned to shareholders. On July 15, 2026, management raised 2026 guidance to €43-45B revenue (+35% at midpoint) at 54-56% gross margin — the largest guidance raise in company history — and committed to expanding EUV and DUV immersion output ~30% each for 2027. The revenue mix by technology and the divisional economics are summarized in the chart and table below; the structural story is EUV and service compounding while DUV cash-cows.

Business mix & divisional economics

8 · Risks to the Target

Export controls & geopolitics — the defining tail risk. China generated 29% of 2025 revenue (down from 36% in 2024), all of it DUV — EUV has never been licensed for export to China, and top-end immersion (NXT:2000i and above) has been restricted since 2023. The risk runs both directions: incremental US/Dutch restrictions on mid-tier tools or on servicing the ~1,400-system China installed base would hit revenue and the IBM annuity directly; and on the other side of the strait, Taiwan takes ~30-35% of shipments — ASML is structurally long the physical security of TSMC’s fabs. There is no hedge for this; there is only position sizing.

Cycle timing — the 2027-28 order cliff scenario. The €43-45B 2026 guide is effectively contracted, but 2027 consensus (~+25%) assumes AI capex compounds without pause. Semicap has never delivered four consecutive years of straight-line growth; ASML itself guided €34-39B for 2026 just six months before guiding €43-45B, which tells you how fast this industry’s visibility regime changes — in both directions. The October 2024 template (one bookings miss, -20% in days) shows how the market treats a 2.5σ multiple when orders wobble. Watch hyperscaler capex guidance, the most upstream signal available.

Customer concentration. TSMC (~32% of system revenue, est.), Samsung (~22%), SK Hynix (~13%), Intel (~10%) and Micron (~7%) fund the model. Single-account decisions move billions: TSMC has already deferred High-NA to ~2029 (A14P/A10); a Samsung foundry retrenchment or an Intel funding crisis would strand capacity ASML is building now. Concentration is intrinsic to the leading edge — it cannot be diversified away, only monitored.

High-NA economics. At ~2.5x the cost per exposure of low-NA, High-NA must win on total cost per good wafer — throughput, yield, and multi-patterning avoidance. Intel’s 18A production insertion (85% cited yield) is proof of technical viability, not yet of industry-wide economics; TSMC’s public skepticism for A14 is the strongest counterparty testimony that the case remains open. A slow EXE ramp would leave ~€400M-ASP capacity underutilized and compress the 2028-30 margin walk.

Supply chain single points. Zeiss SMT is the sole source for projection optics; TRUMPF for drive lasers; final assembly happens in one Veldhoven complex. The +30%/+30% output expansion stresses every tier of a ~5,000-supplier chain running 12-18 month lead times on precision components. ASML’s mitigation — co-investment, multi-year component orders, the January 2026 organizational streamlining — is credible but execution-heavy during the steepest ramp in company history.

Memory cyclicality inside the AI trade. The marginal 2026-27 demand surge is disproportionately HBM/DRAM capacity. Memory has never sustained a four-year upcycle; an HBM oversupply in 2027-28 (all three DRAM makers are adding capacity simultaneously) would hit the fastest-growing slice of the order book first. Litho is later-cycle than WFE peers here — backlog cushions the first shock, then amplifies the second year.

Valuation risk — distinct from business risk. At 64x TTM earnings (+2.5σ vs own history) and 19x sales (+2.9σ), the stock price embeds sustained execution. Our bear case ($1,050, -42%) assumes no catastrophe — just a normal digestion year with a 50% trough gross margin. Investors underwriting at this price are underwriting the absence of a pause, not the presence of growth.

Lower-order risks we monitor: FX (EUR-invoiced revenue vs USD listing — the 1.04→1.14 EUR move flattered USD growth ~8% since early 2025); organizational execution of the Technology/IT streamlining during the ramp; China indigenization (SMEE) eroding restricted-tier DUV demand in the 2030s; Canon nanoimprint remaining confined to non-critical NAND layers; and IP/litigation, historically well-managed (Nikon cross-license 2019).

Catalysts to watch

• Q3 2026 earnings (Oct 14) — first quarter at the €11-12B run-rate; a beat validates the record Q4 implied by guidance
• FY 2026 results + year-end backlog (late Jan 2027) — backlog >€50B and a €53-57B initial 2027 guide are the bull-case data points
• 2028 capacity decision (H2-26/H1-27) — converting 'investigating another +30%' into commitments extends the runway through 2029
• High-NA order flow — a TSMC pull-forward to A14P, or SK Hynix EXE commitments for HBM4, reopens the EXE TAM debate
• Capital Markets Day (Jun 10, 2027) — 2030 model re-baseline from €44-60B; the formal AI-supercycle framing event

Upcoming events

• 2026-10-14 — Q3 2026 earnings (First print of the €11-12B quarterly run-rate; IBM sales guided ~€2.9B)
• 2027-01-27 — Q4 + FY 2026 results (approx.) (Year-end backlog disclosure + initial 2027 guidance; the single biggest catalyst on the calendar)
• 2027-06-10 — Capital Markets Day (2030 model re-baseline (Nov-2024 frame of €44-60B is already stale to the upside))