Evercore ISI reiterates Lam Research stock rating on AI tailwinds

Lam Research remains constructive: Evercore ISI reiterated Outperform with a $275 target, citing 56% YTD stock outperformance, ~37% exposure to NAND spending, and AI-driven demand tailwinds. Street models now call for 31% product revenue growth in 2026 and 24% in 2027, while Lam also guided current-quarter sales to about $5.7B with a 49% gross margin. Offseting the positives, Erste Group downgraded the stock to Hold on margin-risk concerns, even as the company announced leadership changes effective March 6, 2026.

Analysis

The crowdedness of the capex complex is now as important as the fundamentals: when positioning is already overweight and the stock has re-rated to a high-teens to high-30s earnings multiple, incremental good news mostly supports multiple durability rather than outsized EPS-driven upside. That makes Lam a quality leader, but also a late-cycle beneficiary of a theme that is increasingly self-financing through sentiment, not just orders. The second-order winner is likely the broader NAND supply chain, while more price-sensitive equipment peers may lag if investors rotate toward the cleanest AI storage exposure.

The bigger risk is not demand disappearing; it is estimate compression from margin pressure and mix. If the company keeps winning share in a stronger spending tape, customers can demand faster lead times, pricing concessions, and heavier service commitments, which can cap operating leverage even as revenue grows. In that setup, the market can still reward the name for visibility, but the multiple becomes vulnerable if gross margin or book-to-bill stops expanding over the next 1-2 quarters.

Consensus appears to be underestimating how much of the upside is already embedded in the stock through both valuation and positioning. The catalyst path from here is narrower: continued upside likely requires either a NAND capex step-up that beats the current 2026-27 trajectory or a clear margin surprise that validates premium economics. Absent that, the stock may continue to grind higher, but with poor risk/reward versus cleaner ways to express AI infrastructure exposure.

The contrarian setup is to be long the spend cycle but less exposed to multiple compression. If agentic AI really drives a storage-intensity wave, the most asymmetric expression may be suppliers one step removed from the obvious consensus long, where revenue inflects but ownership is lower. That also argues for avoiding chasing Lam after a strong run unless there is a pullback or a fresh catalyst that resets the risk/reward.

Key Decisions for Investors

• Trim or avoid adding to LRCX here; use rallies to de-risk because the stock’s upside is increasingly dependent on multiple expansion, not just earnings revisions, over the next 3-6 months.
• Prefer a pair trade: long a lower-owned NAND-capex beneficiary / short LRCX for 1-3 months, targeting performance dispersion if positioning in LRCX stays crowded and margins become the focus.
• If maintaining LRCX exposure, own it via call spreads rather than stock over the next 2-4 months to cap downside from valuation compression while preserving upside from any NAND spending surprise.
• Watch for the next quarterly print or guidance update; if gross margin fails to improve despite strong demand visibility, use that as the trigger to short into strength with a 5-8% stop.
• For AI infrastructure exposure, shift incremental capital toward names with less consensus ownership and more operating leverage to storage capex rather than adding to the most obvious winner.