This Memory Stock Has Doubled in 2026. Here's Why It Can Skyrocket Higher (Hint: It's Not Micron or Sandisk)

Lam Research shares have nearly doubled in 2026 as AI-driven memory shortages boost demand for semiconductor equipment. The company derived 39% of Q3 fiscal 2026 revenue from memory, while memory capex estimates for 2027 were raised to $144 billion from $83 billion, supporting a stronger growth outlook. Analysts expect EPS to rise 37% to $5.68 this fiscal year and another 40% next year, though the stock already trades at a forward P/E of 49.

Analysis

The market is still pricing Lam as a cyclical equipment beneficiary, but the more important change is that memory capex is shifting from a normal replenishment cycle to a multi-year industrial buildout. That matters because wafer capacity expansion in HBM and advanced DRAM is not a one-quarter catch-up trade; it implies persistent tool demand, better mix, and a longer tail of consumables/service revenue. The second-order winner is the broader semi-capex complex, but Lam has cleaner operating leverage than memory makers because it monetizes every layer of capacity expansion regardless of which vendor wins the final bits.

What the consensus may be missing is that supply tightness is self-reinforcing: as AI clusters absorb more memory, OEMs and hyperscalers will keep paying for lead-time security, which pushes memory makers to overbuild earlier than they otherwise would. That can keep utilization and pricing elevated for longer than modelers expect, even if end-demand normalizes. The real swing factor is whether this turns into a disciplined capacity cycle or an arms race; in the latter case, Lam’s order book can stay strong for several quarters even if memory pricing begins to decelerate.

The risk is less about near-term demand and more about the fade in 12-18 months if capex gets pulled forward too aggressively. Once fabs start catching up, the market will start discounting peak tool shipments before peak earnings, and the stock’s multiple leaves little room for disappointment. A secondary risk is customer concentration: if one or two memory players pause spending, estimate revisions can compress quickly even though the longer-term thesis remains intact.

Key Decisions for Investors

• Stay long LRCX on pullbacks over the next 1-3 months; use a 10-12% drawdown as entry and target a 15-20% upside if memory capex revisions continue to move higher.
• Pair long LRCX / short a basket of lower-quality semi equipment names with weaker memory exposure; the trade favors the company with the cleanest leverage to the memory buildout and better visibility into order conversion.
• Add MU and SNDK as a hedge against the capex thesis being right but tool margin expectations getting too high; if memory shortages persist, they can outperform on pricing while reducing single-name risk.
• For event-driven exposure, buy LRCX call spreads 3-6 months out rather than stock if implied volatility is cheap; the setup is favorable to upside surprise but vulnerable to multiple compression later.
• Reduce or hedge LRCX if management commentary shifts from capacity expansion to digestion; that would be the clearest signal that the market has moved from shortage narrative to peak-cycle risk.