Why I'm Not Selling ASML After a 230% Gain

ASML is presented as a long-term winner, with its stock up about 230% since an early-2021 purchase and still viewed as deserving a premium valuation. The article cites gross margin expansion to 52.8% in 2025 from 48.6% in 2020, revenue CAGR of 18% and EPS CAGR of 24% from 2020 to 2025, and analyst expectations for 2025-2028 revenue and EPS CAGRs of 18% and 27%. Despite trading at 39x next year's earnings, the author argues ASML's EUV monopoly and AI-driven demand justify holding the position.

Analysis

ASML remains a structural toll collector on the AI capex cycle, but the more interesting point is that the market is increasingly underwriting it as a duration asset rather than a cyclical equipment name. That re-rating is justified only if EUV exposure stays scarce and high-NA ramps without material execution slippage; otherwise the multiple compresses quickly because the stock already embeds several years of double-digit growth. The premium is less about current earnings and more about the option value of being the only scalable gatekeeper to leading-edge node transitions.

The second-order benefit accrues to the entire advanced-packaging and foundry ecosystem, but particularly to TSM and the leading GPU designers, because ASML capacity constraints effectively ration front-end output and preserve pricing discipline across the chain. If ASML shipsets stay tight, fabs may prioritize the highest-ROI AI wafers, which reinforces a winner-take-most dynamic for NVDA while leaving weaker foundry customers with slower access to bleeding-edge capacity. Conversely, any easing in lead times could signal that AI capex is normalizing faster than consensus expects, which would hit ASML first and then ripple into the broader semiconductor complex.

The main risk is not demand collapse, but a time mismatch: consensus is assuming multiple years of steady EUV absorption, while the supply chain has a history of lumpy installation and qualification windows. A 6-12 month delay in high-NA commercialization would not break the thesis, but it would likely compress the multiple by several turns because the current valuation leaves limited room for schedule slippage. Geopolitical export restrictions are another underappreciated swing factor; tighter controls can protect near-term pricing power while also capping long-term addressable demand, making the stock vulnerable to headline-driven de-rating even if fundamentals stay intact.

Key Decisions for Investors

• Buy ASML on 5-8% pullbacks over the next 1-3 months; target 12-month upside is moderate rather than explosive, but downside is cushioned by scarcity premium. Risk: a schedule slip in high-NA ramp can compress multiple by 3-5 turns.
• Pair trade: long ASML / short a basket of less differentiated semicap equipment names over 6-12 months. Thesis: ASML’s moat and pricing power should outperform if AI capex remains concentrated in leading-edge nodes; risk is a broad capex reset that hits all equipment names.
• Long TSM vs short a second-tier foundry proxy for 3-6 months if you want to express the same AI-fab bottleneck with lower valuation risk. ASML-led capacity scarcity should favor the operator with the best access to leading-edge tool allocation.
• Use NVDA as a downstream beneficiary trade, but size smaller than ASML: 3-6 month upside persists if leading-edge wafer output stays constrained to the highest-return customers. Risk/reward is worse on multiple compression, so keep it as a relative-value complement, not the core long.
• If ASML gaps higher on guidance or AI demand headlines, sell upside via covered calls 2-3 months out to monetize elevated premium. This preserves core exposure while converting part of the valuation rerate into income.