Up 1,500%, Should You Buy Sandisk Right Now?

SanDisk, which spun off from Western Digital in February 2025 and has rallied ~1,500% through January 2026, reported FY26 Q2 revenue of $3.03 billion, up 61% year-over-year and roughly $360 million above estimates, and EPS of $6.20, a 404% increase that beat the $3.54 estimate. Management cited AI infrastructure demand and price increases, and guided Q3 revenue of $4.4–$4.8 billion (up ~160–183% YoY) and EPS of $12–$14 versus a prior-year loss of $0.30. Despite strong top- and bottom-line beats and bullish guidance, the author cautions that current valuation already prices in substantial growth, making the stock risky for new buyers.

Analysis

Market structure: The surge in SNDK reflects concentrated upside for NAND/NVMe storage suppliers and hyperscalers that train large models; winners are SNDK, NAND fab equipment vendors, and GPU makers (NVDA) via complementary demand, while legacy HDD vendors and commodity DRAM-focused names face relative underperformance. Pricing power is cyclical — current elevated ASPs imply OEMs are absorbing higher costs and SNDK can expand gross margins short-term, but this invites accelerated capex from competitors and inventory builds that can flip pricing within 2–8 quarters.

Risk assessment: Tail risks include a sharp AI-capex pause (20–50% cut by hyperscalers) or a NAND supply shock from fab outages/regulatory export controls, any of which could compress SNDK shares by >50% within months; operational risks include yield issues at partners or concentration (top 3 customers >40% revenue). Immediate horizon (days–weeks) is dominated by sentiment/volatility, short-term (1–3 quarters) by guidance vs. lofty expectations, long-term (2+ years) by NAND cycle mean reversion and vertical integration by hyperscalers.

Trade implications: Tactical direct plays: avoid full-size longs at current froth; prefer sized exposure with downside protection and use relative-value pair trades to isolate idiosyncratic risk. Options: buy 3–6 month SNDK puts (10–25% OTM) as tail hedges and sell covered calls on any near-term position to monetize rich IV; consider long exposure only after a 30–40% pullback or if next-quarter midpoint guidance confirms >120% YoY revenue.

Contrarian angles: Consensus underestimates customer concentration and the ease with which hyperscalers can internalize storage solutions — downside could be faster than models expect. Conversely, the market may underprice durable enterprise demand for inference/edge storage; if SNDK sustains >150% YoY revenue growth for two consecutive quarters, re-rate higher is justified. Historical NAND cycles (sharp booms followed by 40–60% corrections) argue for disciplined entry and explicit risk limits.

Key Decisions for Investors

• Establish a tactical 1–2% portfolio long in SNDK only on a defined pullback: add to 3% total exposure if shares decline 30–40% from current levels or if next-quarter revenue midpoint ≥ $4.6B (consistent with management guide); size to 1–3% notional to limit valuation risk.
• Hedge downside with 3–6 month SNDK protective puts sized to 1–2% portfolio notional (buy 10–20% OTM puts) or buy a 3-month put spread (10%/30% OTM) to cap cost; alternatively sell 6–8 week 15% OTM covered calls on any established long to harvest premium.
• Implement a pair trade: long SNDK (1–2% notional) vs short WDC (1% notional) over 3–12 months to express AI-NAND outperformance vs legacy HDD/restructuring exposure; tighten stops if spread narrows <10% or if WDC announces material NAND divestiture.
• Trim 5–10% from crowded AI beneficiary exposure (e.g., NVDA) and reallocate toward cash/structured opportunities if SNDK or similar AI-infra names run further; target redeployment on volatility spikes >25% in either direction over next 3 months.
• Monitor three catalysts over next 30–90 days before scaling: hyperscaler capex commentary (AWS/GCP/Meta), quarterly NAND ASP trends reported by industry (watch for >10% sequential decline), and SNDK customer concentration disclosures; act if any catalyst deviates by >20% from consensus.