I frequently ask AI for its views on hot investment themes and events, and find that AI's reasoning offers some unique perspectives; I record them promptly and share them with everyone—this is the original intent of this column. Of course, the content is for reference only and does not constitute investment advice.
This Episode's Topic: Why Is ASML Weak? Has the Memory Expansion Benefit Been Priced In?
Recently, ASML's stock has been weak ahead of its Q1 earnings; on the surface the market is trading earnings uncertainty, but fundamentally it is reassessing a question: can the optimistic expectations already baked into the stock price continue to be validated?
Why Has ASML Underperformed Some Equipment Peers?
Prior gains have already front-run a portion of the optimistic expectations.
After the Q4 earnings, the market has already seen AI-related demand, EUV growth, and record orders, so it is normal for ASML to enter a "wait for verification" phase. For a company of this size and valuation, whether the stock can continue to rise often depends on whether orders and full-year guidance can keep being revised up, not just on whether a single quarter's revenue slightly beats expectations.
ASML is not the purest play on memory expansion beta.
Per ASML's 2025 annual report, full-year revenue of €32.7B comprised €16.1B from Logic, €8.4B from Memory, and €8.2B from Service and Field Options. The math: Logic ~49.2%, Memory ~25.7%. This means ASML certainly benefits from HBM and DRAM expansion, but its revenue structure is not like some front-end equipment companies that the market can simply understand as "pure memory beta." From this angle, ASML's short-term stock elasticity lagging behind equipment names more directly levered to memory capex is not surprising.
The market will continue to apply a policy and geographic structure discount.
The quality of ASML's orders and revenue is judged not just by volume but also by geographic mix. The company explicitly stated in its Q4 earnings that 2026 growth will be driven primarily by EUV sales and Installed Base growth, but the market will keep reassessing variables such as export restrictions, tariffs, and Mainland China revenue share. This is why ASML's fundamentals look solid but its stock performance is not always the strongest.
Why I Think the Memory Expansion Benefit Has Not Been Fully Priced In
The core lies in the transmission chain: memory expansion "first turns into orders, then settles into backlog, and finally gradually converts into revenue." ASML explicitly wrote in its Q4 earnings that 2026 growth will be "driven primarily by a significant increase in EUV sales," and customers' medium-term judgment on AI-related demand has become more positive, reflected in upward capacity plan revisions and record orders. The annual report also clearly states that Memory growth momentum comes from HBM and DDR5.
Therefore, to judge whether Q1 remains strong, one cannot look only at headline revenue; more important is whether Q1 net bookings continue to exceed Q1 system revenue. Per company guidance, Q1 total revenue is €8.2B–€8.9B, of which Installed Base Management is approximately €2.4B, so Q1 system revenue is roughly €5.8B–€6.5B. As long as Q1 net orders are clearly above this range, the backlog will most likely be revised upward again.
Now look at the latest industry signals. On March 24, SK Hynix announced it will procure approximately $7.97B of ASML EUV equipment by the end of 2027—a very strong incremental signal in the public domain. It confirms that DRAM and HBM expansion is indeed continuing to transmit to ASML. However, note that such a large order may not all be reflected in ASML's quarterly net bookings or backlog in a single quarter; it ultimately depends on ASML's own written authorization and booking accounting. Directionally positive, but the pace may not be all at once.
How to View High NA?
High NA is important, but I prefer to view it as a medium-to-long-term option for ASML, not the core trading catalyst for this Q1 earnings. Public info shows ASML has delivered 5 High NA systems to Intel, TSMC, and Samsung; Intel's progress is the most explicit, while TSMC's public stance remains "continuing evaluation," and TSMC management has explicitly stated that A14 does not necessarily require High NA, and previously said A16 would not adopt High NA. In other words, TSMC has not publicly confirmed formal full-scale adoption on a production node.
High NA will enhance ASML's long-term optionality, but what is more directly impactful for the near-term stock is the order conversion of low-NA EUV in DRAM and advanced logic. The market right now wants confirmation of "whether orders in the next two quarters continue to rise," not "how large the High NA space is post-2027."
Six Key Numbers to Watch in the Q1 Earnings
Conclusion
Putting all clues together, ASML's recent stock weakness is not because the long-term thesis is broken, but because the market is waiting for harder order verification. I believe the memory expansion benefit has not been fully reflected, but this benefit will first show up in net bookings and backlog, then gradually convert into revenue and profit. For this Q1 earnings, the most critical sentence is: don't stare at headline revenue first; watch whether DRAM and HBM expansion continues to settle into ASML's orders and backlog.
Personal View
AI said a lot, but the key is that at the start of the year the market hyped semiconductor equipment on the core expectation of a memory WFE capex comeback. In prior years of the memory downcycle, lacking the memory leg, ASML relied on Mainland China DUV stockpiling and TSMC logic volume to support growth.