Sandisk Stock Is Up Nearly 635% in 2026. Can It Still Go Higher?

Sandisk (SNDK) has surged ~635% YTD and trades at 9.3x forward earnings after a prior spike to 35x, driven by sustained NAND/SSD tight supply amid data-center SSD demand for AI workloads. Industry peer Micron expects tight memory conditions to persist beyond calendar 2027, supporting continued revenue/earnings growth for Sandisk. The article frames the recent AI-related selloff as a potential buying opportunity given the ongoing supply-demand imbalance.

Analysis

This is a classic squeeze where the first-order trade is not AI demand, but pricing power in a brutally tight commodity-like market. The near-term winners are the names with the most direct exposure to NAND spot pricing, but the better risk-adjusted beneficiary is the larger diversified memory operator, because it can monetize the cycle without being fully hostage to one product line. The second-order effect is that every incremental dollar of storage cost raises the hurdle rate for AI infrastructure returns, which can eventually slow ordering at the margin even if unit demand stays strong.

The market’s biggest mistake is usually assuming shortage = straight line. In memory, the cure for high prices is capex, and that cure tends to show up with a lag; the key question over the next 1-3 months is whether channel inventories stay lean and whether management commentary confirms pricing discipline. Over 6-18 months, the risk is that capacity additions and process improvements compress margins faster than consensus models, especially if AI deployment pauses or hyperscaler spend shifts toward GPUs and networking rather than storage.

My contrarian read is that the move in the pure-play names may be ahead of fundamentals, while the cycle itself is likely underappreciated in duration. If the market is right that tight conditions persist beyond next year, the more durable expression is through the better-capitalized peer rather than the most extended stock. What would falsify this thesis is any evidence of ASP moderation, lead-time normalization, or commentary that wafer starts and capex are ramping materially sooner than expected.

Key Decisions for Investors

• Pair trade: long MU / short SNDK over the next 1-3 months. Thesis: MU offers cleaner exposure to the same supply-driven uplift with less multiple risk; SNDK is more vulnerable if the market starts discounting a mid-cycle capex response. Stop if SNDK outperforms MU by another ~15-20% without follow-through in pricing data.
• Do not chase SNDK after a large run; wait for a pullback or for confirmation in the next pricing/earnings update. Risk/reward is only attractive if the stock de-risks faster than forward estimates rise.
• Buy MU on weakness into any AI semis selloff as the better balance-sheet-and-cycle expression. Time horizon: 3-12 months, with the thesis invalidated if memory pricing or gross margin guidance rolls over.
• Watch NVDA as a secondary beneficiary/loser indicator: if storage costs start to bite AI ROI assumptions, the market can rotate from compute winners into supply-chain beneficiaries. A break lower in NVDA alongside rising SNDK would argue the AI capex trade is becoming more selective.
• Set an alert on NAND pricing and management commentary from MU/SNDK: if evidence emerges that supply additions are accelerating, reduce exposure quickly; that is the highest-probability reversal catalyst within 6-12 months.