Sandisk Drops 23% From 52-Week High: Buy, Sell or Hold the Stock?

Sandisk shares closed at $1,816.57, down 22.8% from their $2,354.39 52-week high, as investors weigh exceptional AI/datacenter-driven pricing and margins against weak consumer demand and a premium valuation. Fiscal Q4 revenue rose 51% sequentially to $8.97B and non-GAAP gross margin reached 84.6%; Q1 fiscal 2027 guidance calls for $10.3B-$10.8B revenue, 83%-85% gross margin, and $44-$46 EPS. Eight multi-year NBM agreements are expected to cover more than 50% of fiscal 2027 bits, supporting at least $93.9B of floor-priced revenue, while the company retains $15.5B of repurchase authorization. Risks include consumer revenue falling 32% sequentially, rising DRAM/component costs, NAND cyclicality, and a 5.17x forward sales valuation.

Analysis

The key underwriting issue is not demand visibility but the durability and economic quality of that visibility. Contractual floors reduce downside to revenue, yet they do not eliminate the NAND cycle: customers can defer volumes, mix can shift toward lower-margin configurations, and elevated enterprise SSD content raises exposure to DRAM/controller costs. At an implied earnings base above $200 per share, even a modest gross-margin reset from the long-run target toward the low-70s would produce disproportionate EPS and multiple downside over the next 6-18 months.

Near term, the selloff creates a tactical setup only if forthcoming pricing data confirms allocation is translating into realized contract price rather than merely optimistic commentary. The more non-obvious beneficiary of sustained enterprise SSD tightness is MU: DRAM content embedded in high-performance SSDs gives it a second revenue leg, while its broader memory portfolio provides a less concentrated way to express AI-storage demand. Conversely, WDC is likely to remain the higher-beta but less clean expression because HDD exposure benefits from hyperscaler capacity needs but does not receive the same flash-margin rerating.

Consensus may be too focused on consumer weakness and insufficiently attentive to capital discipline as a structural variable. If the industry avoids a capacity race, multi-year agreements can compress the historic NAND earnings volatility and justify a higher through-cycle multiple; however, the market should demand evidence in quarterly bit supply growth, customer prepayment/guarantee enforcement, and capex guidance before assigning that outcome. A break in NAND contract pricing or a material increase in industry wafer-capacity plans would falsify the constructive case quickly.

Key Decisions for Investors

• Maintain a neutral-to-underweight SNDK stance into the next earnings print; do not buy the drawdown solely on valuation. Upgrade only if realized pricing and bit shipments support guidance while gross-margin outlook remains above 80%; a margin guide below 78% is a thesis-breaker and likely reopens 15-25% downside.
• Initiate a 3-6 month pair trade: long MU / short SNDK, sized beta-neutral. MU offers indirect enterprise-SSD component upside with lower single-product-cycle concentration; target 10-15% relative outperformance, with stop-loss if SNDK demonstrates sustained pricing realization and MU DRAM pricing weakens.
• Use WDC only as a tactical high-beta long after confirming hyperscaler storage capex acceleration; otherwise remain underweight versus MU. WDC requires evidence that HDD demand and flash pricing are both improving, and is vulnerable to a broader storage inventory correction.
• Set alerts for NAND contract-price indices, supplier capex/wafer-start announcements, and enterprise SSD controller/DRAM cost trends over the next 1-3 months. A sequential price decline or aggressive capacity expansion shifts the preferred expression from relative-value long MU/short SNDK to outright SNDK downside exposure.