Q2 Print Tops Guide But Management Cannot Confirm 2026 Growth on Tariff Uncertainty, €1.4B China Backlog Cancellations, EUV Growth Revised to ~30% From ~40% — Initiating Hold (CB) on the Cleanest EUV Monopoly Plus Concrete Near-Term Uncertainty

Results vs. Consensus

Q2 2025 Scorecard (EUR)

Year-over-Year Comparison

Backlog Detail

Revenue assessment. Q2 net sales of €7.7B reflect strong execution: revenue recognition of one High NA system + additional upgrade business pushed revenue above guide. The 23% YoY growth reflects the 2024-2025 recovery cycle. Installed Base sales of €2.1B above guide reflects strong upgrade momentum on bringing NXE:3800E tools to the 220 wph configuration. The mix shift toward Logic (84% of Q2 system sales) reflects continued AI-driven advanced logic demand at TSMC and other leading-edge customers; Memory at 31% remains supported by HBM and DDR5 demand.

Margins assessment. Gross margin 53.7% (vs 50-52% guide) reflects three favorable drivers: (a) the increase in upgrade business (high-margin), (b) one-off items resulting in lower cost, and (c) lower-than-expected tariff impact — partially offset by the dilutive effect from the High NA system revenue recognition (High NA is currently lower-margin than mature Low NA EUV). For Q3, gross margin guide of 50-52% reflects (a) more High NA tools recognized in H2, (b) lower upgrade revenue contribution, and (c) absence of the Q2 one-off cost benefits.

EPS assessment. €5.90 EPS reflects the gross margin upside flowing through to net income, plus continued share count reduction from the ongoing €12B buyback program. The 47% YoY EPS growth reflects both operating leverage on revenue + the ~3% share count reduction over the past year.

Segment Performance

System Sales by Technology (Q2 2025)

System Sales by End Use (Q2 2025)

EUV (Low NA + High NA)

EUV system sales of €2.7B in Q2 reflect continued strong leading-edge adoption. The structural mechanism: customers add EUV capacity to support AI demand; the higher productivity of NXE:3800E (220 wph) means the same capacity addition needs fewer tools but at higher ASP and improved gross margin. ASML expects ~30% more EUV capacity addition in 2025 vs 2024 (revised from earlier ~40% framing) with approximately the same number of system units. FY2025 EUV revenue growth expected at ~+30% (revised lower from ~+40%).

Assessment: The EUV growth revision is technically a mix-shift accounting story (upgrades counted in Installed Base, not system sales), not a demand deceleration. But the optics on EUV unit count are negative against pre-print expectations. The structural EUV monopoly is intact — DRAM customers adding EUV layers, logic customers ramping leading-edge, High NA milestones progressing. The Q2 EUV bookings of €2.3B were light vs Street expectations but management noted the mix included High NA orders that customers asked not be specifically called out.

Non-EUV (Deep UV + Application)

Non-EUV revenue of €2.9B in Q2 reflects continued strength in advanced DUV nodes (NXT:2100, NXT:2150) plus the KrF NXT:870B carrier system. The DUV mix is shifting toward China (now >25% of total revenue, up from prior >20% framing) as China customers continue purchasing for legacy node capacity expansion under the export restriction framework. The €1.4B China backlog cancellation in Q2 was almost entirely Deep UV, reflecting customers who had ordered tools before 2024 export restrictions deciding they cannot deploy those tools at the configurations they need.

Assessment: Deep UV demand profile reflects two countervailing forces: (a) accelerating leading-edge logic + memory adoption requiring advanced DUV nodes (positive), and (b) China legacy capacity expansion deceleration as customers work through export restriction implications (negative). Net flat-to-slightly-positive for FY2025; the H2 2025 China component is the swing factor.

Installed Base Management

Installed Base sales of €2.1B in Q2 (+37% YoY) reflect strong upgrade business momentum on bringing NXE:3800E tools to the full 220 wph configuration plus growing EUV service revenue as the installed base of EUV systems grows. FY2025 Installed Base growth is now expected at >+20% (raised from prior ~+10% framing).

Assessment: The Installed Base business at €8B+ annualized is now structurally meaningful — comparable to ASML's total 2017 revenue. This provides resilience: even in a slower system-sales year, Installed Base service + upgrades generate ~€8B of recurring revenue at high margins.

Key Topics & Management Commentary

Overall Management Tone: Confident on the structural framework, measured on near-term visibility. Christophe Fouquet was operationally precise on the EUV growth math (mix shift to upgrades + higher productivity per tool = same capacity with fewer units). Roger Dassen was explicit on the €1.4B China cancellation mechanics and the tariff uncertainty driving customer caution. The combined tone reflects a management team confident in the long-term thesis but appropriately cautious about the multi-quarter trajectory given the macro overhang.

1. Cannot Confirm 2026 Growth — The Binding Negative

The most consequential disclosure of the print: Christophe explicitly stated that ASML cannot confirm 2026 growth at this stage. This is a regression from Q1 2025's clearer pre-growth language and reflects the materially higher level of customer uncertainty around tariff and geopolitical developments.

"As we look ahead to 2026, we continue to see strong demand related to AI for both Logic and Memory, and we see the positive impact of a growing number of EUV layers. On the other hand, as we said before, customers are facing increasing uncertainties based on macroeconomic and geopolitical developments. Further, some customers are navigating specific challenges that might affect the timing of their capital expenditure. Against this backdrop, while we are still paying for growth in 2026, we cannot confirm it at this stage. We will continue monitoring developments over the coming months."
— Christophe Fouquet, CEO

The 2026 visibility regression is the binding factor in the post-print sentiment compression. With Q3-Q4 booking data still ahead, the framework will be re-tested over the next 6 months.

Assessment. The 2026 regression is structurally meaningful but timing-bound. The Q3 print will provide the first checkpoint — if Q3 bookings re-accelerate above Q2's €5.5B level + management restores the 2026 growth language, the Hold (CB) becomes upgrade-able. If Q3 bookings remain in the €5B range and 2026 still cannot be confirmed, the rating moderates further.

2. The €1.4B China Backlog Cancellation

Roger disclosed a €1.4B adjustment in the backlog related to customers' response to 2024 export restrictions. Almost entirely Deep UV. The reduction is from China customers who had orders booked before the late-2024 restrictions came into force and have now made up their minds that they cannot deploy those tools at the configurations they had ordered. The backlog goes from ~€34B to €33B as a result.

"The adjustment I specifically talked about was related to China. So there is a €1.4 billion adjustment in the backlog that you need to understand, and that is related to customers now in light of the export restrictions of last year. Customers have now made up their mind what they want to do and that has led to the debooking or the cancellation of orders for about €1.4 billion."
— Roger Dassen, CFO

The cancellation is structurally a one-time event — customers are clearing orders that were no longer executable under the new restriction framework, not signaling broader demand weakness. Management noted that China demand remains healthy at ~25% of total revenue.

Assessment. The cancellation is contained: one-time, almost entirely Deep UV, and does not signal broader China demand deterioration. China at >25% of FY2025 revenue is actually higher than prior >20% guidance, indicating underlying customer demand remains. But the visible reduction in backlog plus the timing (mid-cycle reset) compounds the 2026 visibility compression.

3. EUV Growth Revision — Mix-Shift Math Not Demand Deceleration

FY2025 EUV revenue growth revised from ~+40% to ~+30%. Roger explained the math carefully: the delta sits in the Installed Base business (where upgrades to the 220 wph configuration are counted) rather than in system sales. The 30% capacity addition framework is unchanged — customers are adding capacity, just with fewer units because the per-tool productivity is materially higher.

"At the beginning of the year, that percentage was a little bit higher. And there was a bit of confusion maybe on our side, so I apologize for that, but the big delta is actually with installed base business… the revenue that is associated with bringing those tools to the 220 wafers per hour level is not in system sales, but is in the upgrade business and therefore, is in the installed base business. So the delta that you would see, so this 10% delta, if you like, so 30% growth rather than 40% growth, the delta of that number sits in the installed base business."
— Roger Dassen, CFO

Combined with the Installed Base growth revision to >+20% (from ~+10%), the math works out to roughly flat FY2025 revenue against the prior framing. The EUV unit count is also lower than expected due to the mix shift toward NXE:3800E (220 wph) vs NXE:3600D (160 wph) — same capacity needed with fewer units.

Assessment. The revision is mechanically clean and not structurally negative. EUV demand is intact; the accounting moved between line items. But the optics on EUV unit count are negative against pre-print expectations and contribute to the sentiment compression.

4. China Revenue Profile >25% of FY2025 — Above Prior Guidance

China revenue is now expected to account for >25% of FY2025 (vs prior >20% guidance). The structural mechanism: China demand has remained strong through H1 2025 as customers continue investing in legacy node capacity. Management noted the dynamic of China backlog mix-shifting toward the natural ~25% level that China represents of the total backlog.

"I don't think that dynamic has materially changed, right? So the 25% that we have been talking about now for the second quarter in a row is what we expect for the year. I think that's still the — over 25% is still what we expect for the year. I don't think that dynamic has necessarily changed in the past quarters. That's still there… I think there is a customer need that customer need continues to be strong as we see it. I mean there have been discussions in the past is China falling off the cliff? Well, it is not."
— Roger Dassen, CFO

Assessment. The China >25% framework is a near-term positive (demand remains healthy) but creates a 2026 deceleration risk vector — as China backlog normalizes to the proportional ~25% level, 2026 China revenue likely declines from 2025 levels. This is the structural reason 2026 visibility has compressed.

5. Tariff Uncertainty — The Dominant Macro Overhang

Roger and Christophe both repeatedly cited tariff uncertainty as the dominant overhang on customer capex decisions. The direct impact (tariffs on US system sales, US import of materials, US export of parts to other countries) is being managed actively with customers and suppliers. The indirect impact (broader GDP / market demand) is more complex and undeterminable.

"What you have there is, as you probably know, we shipped quite some tools last year, but also this year to customers at a lower configuration… The clarity from the customers, the uncertainty that they have to navigate is quite substantial, and that leads to them holding the cards a little bit closer to the chest than they typically would do in this time frame."
— Roger Dassen, CFO

The tariff overhang is the binding factor in customer hesitation. Customers are maximizing the waiting time on their capex decisions until tariff clarity emerges. The 232 review remains open with no line of sight to outcomes.

Assessment. The tariff overhang is real and undeterminable. Until the 232 review concludes and country-level tariff frameworks stabilize, customer capex decisions will remain cautious. This is structurally a multi-quarter overhang that will persist through H2 2025 and into 2026.

6. EUV Productivity Math — 30% Capacity With Fewer Units

The mechanical driver of the EUV unit revision: NXE:3800E at 220 wph delivers ~38% more capacity per tool than NXE:3600D at 160 wph. As FY2025 EUV shipments shift toward 3800 (now the dominant configuration), customers can add 30% more capacity using approximately the same number of tools as 2024 — but at higher ASP and better gross margin.

"This year is completely dominated by the 3800, only a handful of 3600 tools. And that's the reason why the tool number is lower simply because people get a much higher productivity on a per tool basis and that still gives them the 30% capacity increase that they were looking for, albeit with a lower number of tools. So it's the tool mix that has led to a lower number of unit numbers."
— Roger Dassen, CFO

Assessment. The productivity math is structurally positive for ASML's margin profile (higher ASP per tool, better GM) but creates an "EUV unit count is decelerating" optics problem against pre-print expectations. This is the kind of operational nuance that matters more to the multi-quarter framework than the headline number suggests.

7. High NA EUV: First 5200B Shipped, HVM Insertion Track

Christophe disclosed that ASML shipped the first EXE:5200B (the High NA HVM tool) and commenced install at customer. The 5200B targets 175 wph — a 60% productivity improvement over the EXE:5000 R&D tool. High NA insertion timeline for high-volume manufacturing remains 2026-27, with progress on customer qualification ongoing.

"This quarter, we also shipped and commenced the install of the first EXE:5200B system, which is intended to support the High NA technology insertion into high-volume manufacturing. As a reminder, the system is capable of achieving at least 175 wafers per hour, which is approximately a 60% productivity improvement compared to the EXE:5000."
— Christophe Fouquet, CEO

The High NA qualification is described as 3-phase: Phase I (R&D customer qualification with 5000) is progressing well; Phase II (HVM qualification with 5200) just commenced with the first 5200 install; Phase III (HVM production at scale) targeted for 2026-27.

Assessment. The High NA roadmap is on track. The 5200 shipment is the key structural milestone for the HVM insertion path. Multi-quarter execution will determine whether High NA can begin contributing meaningfully to bookings/revenue in 2026.

8. DRAM EUV Intensity Increasing — Multi-Year Tailwind

Christophe emphasized that DRAM customers are increasing the number of EUV layers on their latest and future nodes. This is a structural multi-year tailwind for EUV demand as the DRAM technology roadmap requires more EUV exposures per wafer.

"For example, our DRAM customers have recently mentioned an increase in the number of EUV layers on their latest and future nodes. In 2025, we expect our advanced customers to add about 30% more EUV capacity compared to 2024."
— Christophe Fouquet, CEO

Assessment. The DRAM EUV layer intensity expansion is a structural multi-year tailwind. Each EUV layer add at a DRAM node means incremental EUV system demand. With HBM and DDR5 capacity expansion ongoing, the DRAM EUV demand trajectory supports the 2030 framework.

9. 2030 Framework Reiterated: €44-60B Revenue / 56-60% Gross Margin

Despite the near-term uncertainty, management reiterated the 2030 framework from the 2024 Capital Markets Day: revenue opportunity €44-60B and gross margin 56-60%. The structural drivers remain intact: AI-driven advanced logic + DRAM demand, increasing EUV intensity per node, High NA insertion enabling further node shrinks, and the conversion of multi-patterning DUV to single EUV exposure.

"In line with our 2024 Capital Market Day, we expect a 2030 revenue opportunity between €44 billion and €60 billion with gross margin expected between 56% and 60%."
— Christophe Fouquet, CEO

Assessment. The 2030 framework reiteration is structurally important. It signals that the multi-year thesis remains intact even as the near-term 2026 confidence has compressed. The framework supports continued investment in the structural compounder thesis at the long-term horizon.

10. Capital Return: Q2 €1.4B Buybacks + Dividend

Capital return discipline continues: Q2 share repurchases of €1.4B; cumulative 2022-2025 program at €5.8B through Q2; final 2024 dividend €1.84/share paid in Q2; first 2025 interim dividend €1.60/share declared (payable Aug 6). Total 2024 dividend was €6.40/share.

Assessment. The capital return discipline is structurally important. Even with the 2026 uncertainty, ASML's FCF generation supports continued meaningful buybacks + dividend at scale. The combined capital return is ~€2-3B per quarter on top of operating investment.

Guidance & Outlook

FY2025 framework. The total ~+15% revenue growth framework is unchanged from Q1 2025 guidance, even though component allocations shifted (EUV +30% vs +40%; Installed Base +20% vs +10%). H2 2025 is expected to be weighted heavier than H1 with the Q4 ramp being seasonally strongest.

FY2026 framework. "Cannot confirm growth" is the binding signal. Management will provide more 2026 specifics in subsequent quarters as customer capex visibility improves. The structural drivers (AI demand, EUV intensity expansion, High NA insertion) all argue for growth; the timing and magnitude depend on tariff clarity + China normalization + customer-specific capex timing decisions.

Analyst Q&A Highlights

EUV Growth Revision — Mix Shift vs Demand

The opening question on the EUV mix revision pressed on whether the lower EUV unit count signaled demand deceleration. Roger framed the answer mechanically: the delta sits in Installed Base business (upgrades to 220 wph configuration), not in underlying demand.

Q: "My first question would be for the mix of this year. It seems that China is a bit a stronger than expected and even though you said that EUV capacity is up 30% and your revenues is up 30%. It seems consensus was at 49% and flat units still like much lower units than maybe we expected in the beginning of the year when we said below 50%. So we thought it would be closer to 50% rather than 40%. So I was just wondering if you had any change in EUV underlying, although still strong, but is there any change?"
— Francois-Xavier Bouvignies, UBS

A: "In the EUV mix, I think what you see indeed is that we expect growth on EUV this year to be about 30%. At the beginning of the year, that percentage was a little bit higher. And there was a bit of confusion maybe on our side, so I apologize for that, but the big delta is actually with installed base business… the revenue that is associated with bringing those tools to the 220 wafers per hour level is not in system sales, but is in the upgrade business and therefore, is in the installed base business. So the delta that you would see, so this 10% delta, if you like, so 30% growth rather than 40% growth, the delta of that number sits in the installed base business."
— Roger Dassen, CFO

Assessment: The mix-shift mechanics are clean. EUV demand intact at the 30% capacity addition level. The accounting moved between line items, not the underlying demand. But the optics on EUV unit count remain negative for pre-print modeling.

2026 Cautiousness — What Changed Over 90 Days

The most important Q&A exchange of the call: a direct question on what changed since Q1 2025 to drive the 2026 visibility regression. Christophe and Roger jointly framed the answer as tariff uncertainty + 232 review + customer capex hesitation.

Q: "I wanted to ask one on 2026, your commentary. I guess I wanted to try to better understand what has changed over the last 90 days just in your customer conversations. It sounds like you kind of mentioned customers navigating specific challenges that might impact their CapEx. But can you just help us understand what's changed since 90 days ago?"
— Joe Quatrochi, Wells Fargo

A: "When we talked 90 days ago, I think this was a couple of weeks after both Liberation Day, but also the announcement that the tariff we have put on hold for 90 days. And I think we said back then that we are not at the end of discussion, we are most probably at the beginning of it. And I think today, we are still not at the end of it, 3 months have passed. And I think those discussions are in all our customer minds and they are trying basically to understand what it means for them in the short term… this type of uncertainty usually invites people who make investment to be more cautious. And I think what we report is the results of the discussion we are having basically with those people."
— Christophe Fouquet, CEO

Assessment: The "tariff overhang invites customer caution" framing is operationally correct. The unresolved 232 review + ongoing tariff uncertainty are structural overhangs that compress customer capex visibility. This is the binding factor in the 2026 regression and will persist until tariff clarity emerges.

The €1.4B China Backlog Cancellation

A pointed question asking Roger to break down the China backlog adjustment by technology and customer count. Roger framed it: almost entirely Deep UV, multiple customers, related to 2024 export restrictions playing through customer order decisions.

Q: "So, Roger, just a clarification first. I think you said at the beginning of the call that you had order adjustments in the backlog. Can you explain a bit around what proportion of this order adjustment was EUV-related. And what proportion was DUV and whether this was across multiple customers or whether this was just 1 or 2 customers?"
— Adithya Metuku, HSBC

A: "No, the adjustment I specifically talked about was related to China. So there is a €1.4 billion adjustment in the backlog that you need to understand, and that is related to customers now in light of the export restrictions of last year. Customers have now made up their mind what they want to do and that has led to the debooking or the cancellation of orders for about €1.4 billion. That's the comment that I made. And that's a data point you need in order to understand the €33 billion backlog that we ended the quarter with. So this is all Deep UV and a bit of application business, but in essence, most of this is Deep UV."
— Roger Dassen, CFO

Assessment: The cancellation is contained, one-time, and structurally clean. Almost entirely Deep UV, related to a specific regulatory event (2024 export restrictions). Not a signal of broader China demand deterioration.

Bookings Needed to Confirm 2026 Growth

An attempted question to extract the bookings threshold management would need to see in Q3 to confirm 2026 growth. Roger declined directly.

Q: "Roger, first one on calendar '26. I understand that it's hard for you to confirm at this stage. But I'm just curious, when you look at the bookings you had in Q2, what kind of bookings run rate should we assume you need to get in Q3 to get some conviction on calendar '26 as a growth year or not?"
— Krish Sankar, TD Cowen

A: "Krish, very nice try. I think some tried this before. We're not going there, right? So we've made a comment that we've made on bookings. We believe bookings are lumpy, because we believe bookings are not necessarily a good reflection of the business momentum. So I don't think it would be wise to entertain that. So we move away from giving guidance on '26 and I don't think it would be appropriate to indirectly give guidance on '26. So I'm going to pass on this one."
— Roger Dassen, CFO

Assessment: Roger's refusal to anchor on bookings as a 2026 indicator reflects the structural "bookings are lumpy" framing that ASML has maintained for years. The Q3-Q4 framework restoration will be driven by aggregate revenue/EBITDA visibility, not bookings thresholds.

Logic Litho Intensity — Rise Before 2030?

A long-term question on whether litho intensity rises before 2030 in Logic, given that 2nm is a node where EUV layers don't materially increase. Christophe framed the answer: 2nm is a transition node (FinFET to gate-all-around) where EUV layers pause, but the subsequent nodes will see EUV layer expansion driven by AI demand for higher density.

Q: "Just a slightly different topic. But recently, there's been this thesis going around that within the Logic end market, there won't be a rise in litho intensity now until 2030 when your largest customer is expected to adopt High NA EUV. I just — this seems a bit pessimistic to me, but I just wondered how you see if you would agree with the statement?"
— Adithya Metuku, HSBC

A: "This statement should we start with the fact that 2 nanometer, we discussed that many times, is a node basically where we don't see an increase of EUV layers. Now of course, what's happened after that is that gate-all-around is the new architecture and customers are going to go back to drive more density. Many ways to do that. One way is to use, of course, more litho intensity. So I think that after the 1.4 nanometer node, we will see again some litho intensity increase some more EUV layers. If you look at the long term also there, the Logic customers are extremely bullish about the need for more EUV layers."
— Christophe Fouquet, CEO

Assessment: The Logic litho intensity framework is structurally correct. 2nm is a pause node (transistor architecture transition); 1.4nm and beyond see EUV layer expansion. This supports the 2030 framework at €44-60B revenue.

High NA HVM Milestones

A question on the milestones still needed for High NA volume manufacturing. Christophe framed the 3-phase qualification: Phase I R&D qualification (good progress), Phase II HVM qualification (5200 just shipped, install commenced), Phase III HVM production (2026-27 target).

Q: "And then a question for Christophe just on High NA adoption, 175 wafers an hour, an impressive achievement. Clearly, you're shipping qualifying a lot. Maybe just in sort of layman's terms, could you share what are the milestones that are still needed for your customers to cross before we can be more confident about when volume manufacturing begins and just the dilution on High NA. Is it — is the program already above breakeven?"
— Timm Schulze-Melander, Rothschild & Co Redburn

A: "On High NA, so I think we talked about the 3 phases in the past. So I think we're still in what I call Phase I, which is basically R&D customer really qualifying the technology with the 5000. There, I think, the good news is the imaging is doing great, overlay is good. So performance of High NA has been validated… The shipment of the first 5200 means that we are heading towards Phase II, which will be the qualification of the tool for high-volume manufacturing insertion. And there, the key word is maturity, right? Can the tool run without major interruptions so that the customer not only like the performance, but can trust the performance to basically be repeatable in high-volume manufacturing. So I think the next key milestone is about the maturity of the tool."
— Christophe Fouquet, CEO

Assessment: The High NA roadmap is on track for the 2026-27 HVM insertion. The 5200 shipment is the structural milestone. Customer maturity validation will be the binding watch item over the next 4-6 quarters.

Market Reaction

• Pre-print setup (July 15 close): approximately $815. Stock had rallied from ~$650 in March 2025 (post-Q1 recovery) to $815 through April-June on AI capex narrative + Q1 strong order book. YTD return entering print: ~+17%; trailing 12-month return: ~+8% from ~$753 in July 2024 (volatile path through Oct 2024 -16% miss).
• Options-implied move: Approximately 7-9%.
• July 16 close (US session reaction to overnight Amsterdam print): $738, down −9.5% (−$77). Intraday low $732 (−10.2%). Volume ~22M ADR shares, ~4x average.
• Friday July 17 session: Modest recovery to ~$748.
• Sell-side reaction: Multiple price target cuts. Goldman, Morgan Stanley, BNP Paribas all reduced targets ~10-15%. "Cannot confirm 2026 growth" was the lead headline.
• Peer reactions: Other semicap names (LRCX, KLAC, AMAT) down 2-4% on the day on read-through; TSMC, NVDA flat.

The -9.5% reaction reflects forward-framework compression rather than print quality. The Q2 print itself topped guide on every metric. The forward framework regression (cannot confirm 2026 + EUV growth lowered + €1.4B China backlog cancellation) reset position sizing across long-only books. At $738 post-drawdown, the entry asymmetry has materially improved — the long-term thesis is structurally intact (EUV monopoly, AI-driven litho intensity, 2030 €44-60B framework), but the near-term sentiment is digesting the multi-vector uncertainty.

Street Perspective

Debate: Does the 2026 Visibility Regression Justify Multi-Quarter Sentiment Compression?

Bull view: The 2026 regression is timing-bound, not structural. AI demand drivers (advanced logic + DRAM litho intensity) remain intact; the tariff overhang will resolve over coming quarters as the 232 review concludes and country-level frameworks stabilize. China backlog cancellations are a one-time event, not a broader China demand deterioration. The Q2 print itself was strong (revenue + GM + EPS above guide). At $738 post-drawdown, the multi-year thesis is materially under-priced.

Bear view: The 2026 regression signals genuine multi-quarter visibility compression. Tariff uncertainty persists indefinitely; China revenue normalization continues; specific customer capex timing concerns reflect real near-term headwinds. The 2030 framework is years away; the next 4-6 quarters carry meaningful execution risk. Multiple compression is appropriate given the framework reset.

Our take: Bear view captures the near-term framework correctly; bull view captures the long-term framework correctly. The multi-quarter timing matters: Q3 print + Q4 framework will determine whether 2026 can be confirmed. We initiate at Hold (CB) reflecting both views — structural confidence in the multi-year thesis, near-term caution on the visibility compression.

Debate: Is the EUV Growth Revision a Mix-Shift Story or a Demand Deceleration?

Bull view: The revision is purely mechanical: upgrades to 220 wph configuration are counted in Installed Base, not system sales. EUV unit count appears lower because higher productivity per tool (NXE:3800E at 220 wph vs NXE:3600D at 160 wph) means customers add the same capacity with fewer units. Underlying EUV demand is intact at the 30% capacity addition framework.

Bear view: The mechanical math is correct but the optics matter. EUV unit count is the binding metric the Street uses to model multi-quarter revenue. A lower unit count reduces the visibility window for 2026-27 EUV revenue. Even if the dollar-revenue math holds in 2025, the 2026 trajectory becomes more sensitive to High NA insertion timing (currently 2026-27).

Our take: Bull view captures the mechanics correctly. The revision is a mix-shift accounting story, not a demand deceleration. But the Street's modeling framework is built on unit count and the reset creates real near-term visibility compression. Net: contained negative but worth flagging for multi-quarter framework restoration.

Debate: Is the €1.4B China Cancellation a One-Time Reset or a Multi-Quarter Deceleration?

Bull view: The cancellation is one-time, mechanical, and clean. Almost entirely Deep UV, related to a specific regulatory event (2024 export restrictions playing through customer decisions). China revenue at >25% of FY2025 is actually higher than prior >20% guidance, indicating underlying demand remains healthy. The backlog reset is the result of customers making final decisions on orders that were no longer executable under the restriction framework.

Bear view: The cancellation may be one-time but the trend is structural. China revenue normalization to the proportional ~25% backlog level implies 2026 China revenue declines from 2025 levels (which are inflated by China pull-forward). Multi-year China deceleration is a structural overhang on ASML's revenue trajectory.

Our take: Bull view captures the immediate event correctly; bear view captures the multi-year trajectory correctly. The cancellation itself is contained, but China normalization is a real 2026-27 framework drag. Management's "China is not falling off a cliff" framing is appropriate but does not eliminate the multi-year deceleration concern.

Model Update & Valuation Framework

Valuation framework. At $738 post-drawdown, ASML trades at approximately 22x FY2026E EBITDA (base case ~€13B). The historical EV/EBITDA range is 25-35x for the structural EUV monopoly. Current multiple compression reflects the 2026 visibility regression + China deceleration concerns. Multi-quarter recovery to 25-28x is supported by 2026 framework restoration if Q3-Q4 customer capex visibility improves.

Risk-reward. At $738: base case PT $800-900 implies +8-22% upside; bull case $950-1,050 implies +29-42%; bear case $600-700 implies -5-19% downside. The up-to-down ratio is approximately 2:1 in base-to-bear scenarios — favorable enough for Hold (CB) initiation pending Q3 framework restoration evidence.

Thesis Scorecard & Initiating Coverage Framework

As this is our initiating coverage report on ASML, we establish the foundational thesis pillars and signposts.

Bull Case Pillars (5)

Bear Case Pillars (3)

Upgrade Triggers (to Outperform)

• Q3 2025 print restores 2026 growth language with management confirming directional growth
• Q3 bookings re-accelerate above €5.5B with EUV bookings >€3B
• Tariff clarity emerges (232 review concludes; country-level frameworks stabilize)
• China revenue stabilizes at >20% baseline through 2026 (vs. deceleration to ~15%)
• Stock retraces to $650-700 range with fundamentals undamaged

Downgrade Triggers (to Underperform)

• Q3 bookings <€4.5B (lower than Q2's €5.5B), confirming demand deceleration
• FY2026 revenue guide at flat-to-negative growth when issued
• Material new China restrictions imposing additional backlog cancellations
• Tariff outcome materially negative for ASML or its customers (32% blanket tariff range)

Overall verdict: Long-term thesis structurally intact. Near-term visibility compressed. Hold (CB) reflects the appropriate balance — the EUV monopoly + 2030 framework support the long-term compounder thesis, but the 2026 visibility regression + China normalization + tariff overhang create multi-quarter sentiment compression that argues against immediate Outperform.

Action: Initiate at Hold (CB). Existing holders: hold. New positions: wait for Q3 framework restoration or pullback to $650-700. Sized positions: consider trimming top quartile if portfolio weight has expanded materially.