AI Can't Thrive Without This Stock. Hint: It's Not Nvidia.

ASML is positioned as a critical enabler of AI chip production, producing the only extreme ultraviolet (EUV) lithography systems used by leading foundries (TSMC, Samsung, Intel). Analysts expect ASML revenue and EPS CAGRs of 26% (2025–2028) and 39%, respectively, while noting the stock trades at about 29x next year’s earnings after a ~130% rally over the past 12 months. The article argues ASML’s EUV/DUV monopoly-like position and pricing power could support continued growth as ASML is restricted from selling EUV and higher-end DUV systems in China.

Analysis

ASML is the cleanest toll-collector on the AI capex stack: it monetizes every incremental node shrink regardless of whether demand is served by NVDA GPUs, AVGO ASICs, or custom silicon. The second-order implication is that the scarcity rent sits upstream—tool lead times and service mix matter more than unit chip volumes—so the best risk-adjusted exposure is not the highest-beta AI end market, but the constrained equipment layer with pricing power.

Near term, this is mostly a multiple/positioning story, not an operating inflection. The stock can rerate if order intake and guidance confirm that foundry capex is still accelerating, but after a strong move the market will punish any sign of digestion faster than it rewards more good news. The main losers are late-cycle capacity seekers like INTC, where any delay in advanced-node ramp extends depreciation drag and pushes out turnaround economics; TSM is more insulated, but its earnings upside is still gated by tool availability rather than demand.

Contrarian view: consensus may be over-anchored to AI chip brands and underweight the fact that lithography is the binding constraint. The bearish counter is that export controls, customer concentration, and a capex pause at one or two hyperscale/foundry buyers can hit ASML’s growth path abruptly, so the bull case needs visible bookings, not just narrative. Over 6-18 months, the thesis fails if advanced-node spending rolls over or if ASML’s multiple compresses below the high-20s as the market reclassifies AI capex as cyclical rather than structural.

Key Decisions for Investors

• Initiate or add to ASML on 3-5% pullbacks over the next 1-2 weeks; 6-12 month upside is driven by scarcity pricing and installed-base service growth, but trim if the next order/bookings print misses consensus or management turns cautious on 2025-26 capex.
• Pair trade: long ASML / short NVDA (beta-adjusted) for 1-3 months to express the view that upstream bottlenecks capture more durable margin than downstream AI hardware; risk is a renewed GPU demand surge that lifts both legs, so cover if NVDA data-center growth reaccelerates above expectations.
• Use INTC as the clearest relative loser if you want a cleaner expression of advanced-node delay risk; maintain a short or underweight into any rally unless Intel shows evidence of process-ramp acceleration and capex discipline.
• Watch TSM and AVGO earnings for capex commentary rather than revenue headlines; a sustained increase in foundry tool budgets is the best confirmation for ASML, while any capex pause is the fastest falsifier.