Better AI Stock: ASML vs. Nvidia

Nvidia and ASML are both benefiting from surging AI infrastructure spend: Nvidia reported fiscal 2025 sales up 114% to $130.5 billion and diluted EPS up 130% to $2.99, with fiscal-2026 Q3 revenue +62% to $57 billion and EPS +67% to $1.30, and holds an estimated ~90% share of the data-center GPU market. ASML, which supplies lithography equipment to foundries, commands an estimated 90% market share in lithography, saw service revenue rise 39% in the first nine months to €6 billion, and reported diluted EPS of $17.38 (up ~40% year-over-year for the same period). Valuation favors Nvidia in the author’s view—Nvidia trading at a P/E of ~23 versus ASML at ~34—though both are described as highly profitable and positioned to capture continued AI-driven capex.

Analysis

Contrarian angles: Consensus underestimates timing risk — the narrative assumes linear capex; mispricing exists where NVDA trades at ~P/E 23 vs ASML ~34 despite NVDA’s faster top‑line compounding (50–100% y/y near‑term). Historical parallel: 2017–18 fab cycles showed equipment stocks lead/lag on order execution — buying ASML pre‑delivery can be riskier than owning NVDA whose revenue books faster. Unintended consequence: excessive optimism may force customers to overbuild specialty capacity, creating a 2–3 year mid‑cycle correction of tooling orders if AI projects de‑prioritize in downturn.

Key Decisions for Investors

• Establish a 3% portfolio long position in NVDA (build to 2–4% over 3 months), accumulate on any 10%+ intraday pullback or if forward P/E compresses to ≤20; horizon 12–36 months, target total return >40% if AI spend ramps as forecast.
• Initiate a 1.5% strategic long in ASML (buy in tranches over 6–12 weeks), hold 12–36 months for service‑revenue tailwinds; trim to zero if quarterly backlog growth stalls or ASML reports >20% quarter‑over‑quarter delivery delays.
• Implement a relative‑value pair: long NVDA (2% notional) and short ASML (1% notional) to capture valuation divergence (NVDA P/E 23 vs ASML 34); maintain for 6–18 months and rebalance if spread narrows to <5 P/E points.
• Options risk‑management: Buy 9–18 month NVDA LEAPS equal to 1% notional to capture convex upside; sell 30–60 day covered calls to finance if willing to cap near‑term gains. Separately, buy 6–12 month ASML puts (cost ≤0.5% notional) as explicit hedge against export‑control or geopolitical shocks.
• Add 1% exposure to TSM (long) to play foundry capacity; monitor two catalysts: ASML quarterly order book releases and any US/Dutch export‑control announcements within the next 30–90 days — if either turns adverse, liquidate ASML and TSM exposure and increase cash by 2–3%.