ASML Prints Into a Tape That Already Won: Q3 Needs to Show 2027, Not Just Beat Q3

The setup heading into Wednesday is a 60% ADR rally and a 52% Taiwan index rally compressed into ten months. The TAIEX closed at a record 49,712.04 on Oct 6 — a +2.55% session in which TSMC alone contributed an estimated 595 points of the 1,236-point advance on turnover of NT$1.15 trillion. TSMC's Taipei listing (2330.TW) hit an intraday record of NT$2,590 the prior session. The US-listed ADR is at $453.31 (Oct 9 close) versus the $485.80 Oct 5 all-time high, a 7% pullback. The bid resumed on Sep 22 when the TAIEX first pierced 48,600, and accelerated after Taiwan's Sep 5 nonfarm payrolls print (+29k vs +84k expected) cut the implied probability of a Fed hike to 22.7% from 64.2%.

Sell-side has converged near the high end of ASML's own guide. Zacks' consensus calls for $13.18B in revenue and $12.47 EPS; ChartMill's consensus is €11.85B in revenue and €10.84 in EPS — both inside the top quartile of management's €11.0–12.0B range. The company set that range on July 15 alongside Q2 results, and Q2 itself printed €9.33B versus a guide of €8.8–9.4B. Translation: in any normal quarter, a print at €11.7–12.0B would be a clean beat. On Wednesday, it will be the baseline.

Every name in ASML's five-customer stack — TSMC, Samsung Electronics, Intel, SK Hynix, Micron Technology — has stepped up capex commitments in 2026, and the four most aggressive have all done so since ASML's Q2 print on July 15. The aggregate 2026 spend now runs at roughly $300B from the top three foundries and memory players, of which ASML's addressable lithography cut is a high-teens-percent share.

Two of those line items are direct ASML signals. ASML guided July 15 to ~75% growth in memory-customer revenue in 2026 and ~45% growth in EUV system sales, alongside plans to add 30% to its 2026 low-NA EUV capacity (~65 systems) for 2027. End-2025 backlog of €38.8B — of which €25.5B (66%) is EUV — represents roughly 16 to 17 months of forward revenue. Memory customers are now the marginal buyer: Samsung Electronics signed an expanded High-NA EUV partnership with ASML on Sep 8, and TSMC targets High-NA EUV in advanced node mass production by 2030 (still a 2027–28 mass-production start for the broader industry, per CEO Christophe Fouquet).

ASML trades at 17.3x TTM EV/sales and 44.4x TTM EV/EBITDA, with a 13.7% R&D-to-revenue ratio and 52.4% return on equity TTM. Those are not bubble multiples on absolute terms, but they are rich on a five-year view. The rally is also outpacing TSMC on a relative basis: TSM is at 21x forward earnings versus ASML at 29x, leaving TSMC the cheaper AI-capex proxy. Three near-term overhangs argue against chasing the print.

China is half a problem. ASML's China share has stepped down sharply: from 33% of 2025 sales to a guided ~20% in 2026, with Q1 and Q2 2026 European revenue share at 0% and Taiwan share climbing to ~30%. The runway for further mix-down is largely priced, but the composition of 2027 growth is now entirely dependent on logic and memory, where visibility is good but not asymmetric. Any new US or Dutch export-control tightening on legacy DUV would be a binary risk; the High-NA EUV system is already prohibited for China.

Geopolitics has not gone away. Trump and Xi met in Beijing the week of Sep 22. Trump called the pending $14B Taiwan arms package a \"very good negotiating chip.\" The market read this as a reprieve rather than a threat, but ASML sits at the intersection of US export controls, Dutch licensing, and Taiwan-strait risk — the three lines that can move the multiple as fast as any earnings number. President Lai's Oct 10 National Day speech promised record NT$1.1T in 2027 defense spending but did not produce a fresh Taipei-Beijing escalation.

The easy comps end in Q4. Q3 2025 revenue was just €7.52B; the +58% consensus Q3 2026 number looks heroic but the comp is the trough. Q4 2025 was €9.72B — so the 2026 Q4 print must clear roughly €12B to maintain the YoY growth narrative. That is exactly the level ASML itself has guided to. If Q4 2026 lands between €10.5B and €11.5B, growth optically slows to a low double-digit print at a time when the multiple has already priced in a 25%+ grower.

Short-term (days to one quarter). ASML's options market is flashing a defensive tone: total open interest of 187.24K with a 1.30 put/call ratio skewed bearish, and TSMC options imply a ±4.26% one-day move into the Oct 15 print. The first move after ASML's call will be a function of two numbers — Q3 net sales versus the high end of guide (€12B) and the FY2026 guide versus the current €43–45B range. A raise to €45–48B would extend the rally; a hold or worse would invite a 5–8% pullback. The TSMC Oct 15 print, with the actual Q3 net profit figure and Q4 outlook, will set the tape for the rest of the month.

Long-term (one to three years). The structural bull case is intact: AI capex from the top three foundries is on track to push global semi capex above $300B in 2026 and ASML's 2027 EUV book is essentially full. The open question is the shape of 2028 — when 30% of incremental EUV capacity lands, when the High-NA EUV installed base scales from 10 to 20+ systems, and when the China mix is closer to 15%. Watch for management commentary on High-NA unit economics, 2nm tool readiness, and any pull-forward of memory orders from Samsung Electronics and SK Hynix.

The non-obvious read: this is the first ASML print where consensus and management are aligned, the supply chain is screaming \"buy,\" and the tape has already done the work. A clean beat at €11.8–12.0B with a FY2026 raise to €46–48B is the only outcome that extends the rally; anything in the middle of the guide or any whiff of a 2027 pause would test the 50-day moving average, which sits roughly at $1,650 (around 7% below current). The fact that the call goes out two days before TSMC's Q3 print — and the same week as a TSMC capex reaffirmation at the 2026 Open Foundry Forum — means the conversation moves from \"is AI capex real?\" to \"how much of 2027 is already locked?\"