The Artificial Intelligence (AI) Inference Market Could Reach $255 Billion by 2030. This Stock Is Best Positioned to Win.

Key numbers: ASML's EUV tools can cost up to $400M, maintenance made up ~23% of net sales in 2025, and Markets & Markets forecasts a 19% CAGR for the AI inference market through 2030. The piece argues ASML is effectively a monopoly in EUV—critical to TSMC (which controls ~71% of the foundry market)—and benefits from manufacturing in the Netherlands, reducing Taiwan-related geopolitical risk and limiting exposure to Chinese rivals amid export restrictions. Conclusion: ASML is positioned as a durable, high-barrier 'pick-and-shovel' AI play with recurring maintenance revenue and limited technical competition.

Analysis

ASML’s real optionality isn’t just marquee wins but the way a concentrated installed base converts lumpy capex into multi-year, high-margin aftermarket cash flows. That annuity characteristic compresses revenue volatility: a 10–20% miss in new tool deliveries historically translates to only a mid-single-digit EPS impact over 12 months because service and spare-part revenues persist. Treat ASML as a quasi-utility within semicap exposure rather than a pure cyclical equipment vendor.

The largest asymmetric risks are policy and supply fragility, not pure technology competition. A targeted export stop or single-supplier failure for critical EUV subassemblies could curtail deliveries by a material share (order-of-magnitude: tens of percent of annual shipments) inside 6–12 months, forcing foundries to defer node transitions and creating a temporary oversupply in legacy tools. Conversely, accelerated foundry rollouts for AI inference could front-load orders and compress lead times, producing above-consensus revenue in the next 4–8 quarters.

Second-order winners and losers are underpriced today. Legacy lithography and legacy-node vendors would capture incremental demand if leading-edge rollouts slow, while metrology and optics players with high spare-part exposure (and low incremental capex) will see margin expansion. Finally, geopolitical repricing favors firms with globally dispersed manufacturing footprints and after-sales networks; that divergence should widen over the next 12–36 months and can be harvested via relative-value trades.

Key Decisions for Investors

• Long ASML (ticker: ASML) — size 3–6% of tech equipment sleeve via outright shares or 9–18 month call spreads; target asymmetric upside of 30–50% if demand front-loads, with downside capped by allocating 1% to 9–12 month put protection to limit a policy/ban shock to single-digit portfolio drawdown.
• Pair trade: long ASML / short TSM (ratio 1:0.6 notional) over 6–12 months — monetizes differential geopolitical and aftermarket exposure. Risk: 20–40% short squeeze/correlation convergence if a broad semicap rally lifts both; trim at a relative outperformance of 15% or if semicap ETF flows accelerate.
• Long metrology & legacy equipment suppliers (tickers: KLAC, LRCX, AMAT) — buy 12-month out-of-the-money call spreads (small notional) to capture subsidy-driven legacy-node demand in China or accelerated capacity builds elsewhere. Expect 2:1 asymmetry: limited premium vs >40% upside if legacy-piece demand surges.
• Protective hedges: buy 9–12 month puts on Taiwan foundry exposure (TSM) sized to cover estimated value-at-risk from a geopolitical shock (~portfolio tail risk). Cost of hedge acceptable if it avoids a 20–30% catastrophic P&L hit; unwind if implied vol for Taiwan rises above realized by >150bps.