This semiconductor stock has more than doubled in a year. Bernstein sees more gains ahead

Bernstein raised its ASML price target on U.S.-listed shares to $2,623 from $1,971, implying 48% upside from Thursday’s close, and reiterated an Outperform rating. The note cites materially higher topline forecasts driven by unprecedented AI-driven expansion in advanced logic and DRAM, with high numerical aperture EUV likely adopted first in DRAM due to lower exposure costs. ASML shares were up ~3% in premarket as the stock has gained 123% over the past year, supported by broad sell-side buy ratings (19/19 covering analysts).

Analysis

The main second-order effect is not just higher ASML revenue, but a shift in the AI capex mix toward memory-heavy nodes where the tool intensity is highest. That should make ASML’s order growth more durable than the market assumes, while also tightening DRAM/HBM supply and extending pricing power for memory makers like MU before the next upcycle is fully visible in reported earnings. The more interesting spillover is that memory suppliers may see margins inflect first, while logic/foundry exposure benefits later and less cleanly.

The risk is crowding. ASML already trades like a consensus AI winner, so upside now depends on quarterly bookings and backlog re-acceleration, not the broad AI narrative. Over the next 1-3 months, the key catalyst is whether customer capex commentary validates HNA EUV adoption in DRAM; over 6-18 months, the risk flips to overbuild if HBM capacity is added too aggressively, which would eventually pressure memory pricing and compress the setup.

Contrarian view: the Street may be overpaying for the timing of HNA EUV adoption. If DRAM makers can meet demand with process tweaks and existing tools longer than expected, ASML’s revenue recognition lags the enthusiasm and the stock can digest even after a bullish target reset. In that case, the cleaner expression is relative value in memory versus broad semis, not chasing ASML after a large run.

Key Decisions for Investors

• Long MU / short SOXX for 3-6 months: express the view that AI memory scarcity, not broad semiconductor beta, is the cleaner near-term beneficiary if DRAM/HBM capex accelerates.
• Accumulate ASML only on pullbacks; use a 6-12 month call spread rather than outright chase. Risk/reward is still positive, but the name is crowded and needs order-intake confirmation to re-rate further.
• Long MU / short TSM for 6-12 months: if HNA EUV adoption starts in DRAM before logic, memory should outperform foundry exposure that is already more fully owned in AI portfolios.
• Watch item: if ASML’s next 1-2 quarters do not show higher bookings/backlog from memory customers, or if MU’s DRAM ASPs fail to inflect, the thesis is weakened and the trade should be reduced.