Morgan Stanley updated its research report on SanDisk (SNDK) on June 22, raising the price target from $1,100 to $1,750 and maintaining an overweight rating. The rationale is straightforward: AI inference demand is rewriting the rules of the NAND market. Cloud data center customers are insensitive to price, granting SanDisk pricing power. Moreover, the New Business Model (NBM) agreements lock in gross margins, making the company's future profits largely predictable.
AI Inference Reshapes NAND Demand Structure
SanDisk's cloud business grew 233% in Q1 (after a 64% QoQ increase in Q4), driven by structural demand changes. Cloud providers are paying a premium for AI inference KV cache and context window storage. Morgan Stanley estimates that cloud revenue accounted for a high percentage of SanDisk's Q1 sales, almost entirely driven by TLC (triple-level cell) technology, as customers prioritize density and performance. These customers sign long-term contracts with locked-in prices, providing real support for gross margins.
SanDisk's NBM agreements have locked in over one-third of FY27 bit shipments, typically with 3-to-5-year terms featuring fixed prices or collars. Even at floor prices, these contracts maintain approximately 80% gross margin. For context, SanDisk's gross margin was only 30.3% in FY25, rising to 69.2% in FY26e and 86.7% in FY27e — an improvement that is sustainable. Morgan Stanley believes the company could eventually cover 70%–80% of shipments under NBM, creating a profit buffer. The 80% margin at floor prices means high profitability can persist even in a price war.
Supply Pricing Power and Earnings Resilience
The current NAND supply tightness is likely to persist for an extended period. AI data center expansion is accelerating, and storage demand is far from saturated. SanDisk is locking in long-term contracts to hedge against cyclical risks. Morgan Stanley expects average selling prices (ASP) to continue rising through calendar 2026 and potentially into mid-2027. SanDisk generates 40%–50% of revenue from North America, and data centers have become its largest end market. Against a backdrop of tight supply and high customer loyalty, pricing power rests with the supplier.
The company targets 15%–19% bit growth, primarily through technology transitions (density improvements and process refinements), not capacity expansion. From FY25 to FY27, revenue is projected to rise from $7.355 billion to $48.826 billion (approximately 6.6x), while EPS grows from $2.74 to $14.73. Growth comes from high-margin cloud business, not low-price consumer markets. SanDisk also announced a $6 billion share repurchase program; management believes the current stock is among the lowest-valued in the semiconductor sector. Morgan Stanley outlines three scenarios based on FY27 EPS: base case of 28x PE = $1,750, bull case of 31x PE = $2,635, bear case of 25x PE = $1,100.
Catalysts and Risks
Upside catalysts include: faster-than-expected eSSD penetration in data centers, edge AI applications driving NAND content growth, and returns from advanced technology investments such as HBF (high-bandwidth flash). Downside risks include: below-trend industry growth, increased competitor capex, SanDisk losing data center share, and Chinese storage makers (e.g., YMTC) gaining market share.
Morgan Stanley's bullish thesis rests on three pillars: structural demand change from AI inference, margin protection via NBM agreements, and sustained NAND supply tightness. The price target increase from $1,100 to $1,750 implies ~28x FY27 PE. The forecasts will be updated with earnings and customer feedback, but the logical framework is more valuable than the specific numbers.
This article is a summary and interpretation of Morgan Stanley's research report and does not constitute investment advice. Market risks exist; make independent decisions.