Sandisk Is Still Not Peaking

Sandisk is maintained at Buy despite a 23% pullback from its highs, with an analyst price target of $2,200 implying 22-23% upside. The bullish view is supported by multi-year net bookings, record free cash flow and robust enterprise/AI demand. Management is targeting 80% gross margins, 75% EBIT margins and 50% FCF margins in FY2028-30, aided by BiCS10 and HBF next-generation products.

Analysis

The valuation case hinges on whether SNDK can sustain a structurally higher NAND pricing regime rather than merely benefit from an upcycle. Enterprise SSD mix and AI storage demand can improve realized pricing and operating leverage, but NAND remains vulnerable to rapid supply responses from Samsung, SK hynix, Kioxia (285A), and Micron (MU). The key read-through is quarterly bit-supply discipline: if industry output rises faster than nearline/enterprise demand over the next 2-3 quarters, SNDK's earnings multiple will compress before reported margins materially weaken.

The margin framework should be treated as a long-duration execution aspiration, not a near-term valuation anchor. Achieving it requires successful qualification and yield ramps for next-generation technology, continued enterprise mix gains, and no meaningful customer concentration or hyperscaler inventory correction. A more attractive second-order beneficiary may be MU: it has AI-memory exposure but less dependence on a single enterprise SSD narrative, making it a cleaner relative hedge if the market is overpaying for SNDK's terminal economics.

Near term, the recent drawdown creates a tactical setup only if management commentary confirms stable contract pricing, improving enterprise SSD attach rates, and restrained capex across the NAND industry. Over 6-18 months, the principal downside is a return to commodity-like NAND conditions; high fixed-cost manufacturing means even modest ASP declines can produce disproportionate FCF and estimate risk. The thesis is falsified by sequential NAND ASP declines, weaker gross-margin guidance, elevated inventory days, or evidence that AI customers are adopting lower-cost HDD/tiered-storage architectures instead of premium flash capacity.

Consensus may be underestimating the duration of AI-related storage demand, but is likely over-crediting a smooth margin path. Storage bottlenecks are real, yet hyperscalers optimize total cost of ownership aggressively and can shift workload tiers quickly. The stock should therefore trade as a high-beta semiconductor-cycle position rather than a secular software-like compounder until margin durability is demonstrated through at least two clean pricing and technology-ramp quarters.

Key Decisions for Investors

• Maintain SNDK as a tactical long only on post-earnings confirmation of sequential enterprise revenue growth and stable-to-higher NAND ASP commentary; use a 3-6 month horizon and size modestly given cyclicality. Exit if management guides gross margin lower sequentially or cites inventory accumulation.
• Express relative upside through long SNDK / short MU only if SNDK's enterprise SSD growth materially exceeds MU's storage-related growth for two consecutive quarters; otherwise MU is the preferred less-idiosyncratic memory exposure.
• For existing SNDK longs, buy 3-6 month downside puts or reduce exposure into earnings if implied volatility is below the stock's recent realized volatility. The asymmetric risk is a pricing or yield-ramp disappointment that can reset forward estimates faster than spot demand data signals.
• Create an industry alert around Kioxia (285A), Samsung memory capex disclosures, and Micron NAND output commentary. Any coordinated indication of capacity expansion is a catalyst to cut SNDK exposure before a potential 6-12 month ASP downturn.