Sandisk Just Made the Next Memory Crash a Lot Less Scary

Sandisk has secured 10 multiyear flash-memory supply agreements with eight customers that guarantee at least $93.9B of revenue at contractual price floors, supported by $16.5B in customer financial guarantees. The contracts are expected to cover more than half of fiscal 2027 shipment volumes and about two-thirds in fiscal 2028, reducing exposure to a memory-price downturn; management guides for fiscal Q1 2027 revenue of $10.3B-$10.8B and non-GAAP gross margin of 83%-85%. However, nearly half of fiscal 2027 volumes remain exposed to spot pricing, floor levels are undisclosed, and the structure has not been tested through a downcycle, limiting confidence that recent 84.6% gross margins can persist.

Analysis

The key valuation question is no longer peak NAND pricing but whether contracted volumes convert a historically spot-exposed commodity producer into a partially infrastructure-like earnings stream. If enforceable, the agreements should reduce SNDK's beta to NAND spot prices and justify a higher trough multiple; however, the disclosed guarantees are not equivalent to unconditional revenue, and customer credit quality, termination provisions, qualification delays, and ceiling/floor definitions remain the critical diligence gaps. The market will likely withhold rerating until a period of declining NAND prices demonstrates that realized ASPs, utilization, and gross margin decouple from spot.

Near-term, SNDK remains exposed to a nonlinear downside: uncontracted volumes clear at marginal-market prices, while fixed manufacturing costs can amplify the earnings impact if demand weakens. A broad hyperscaler capex retrenchment would also pressure both uncovered demand and customers' willingness to absorb contracted inventory, potentially turning deposits into a dispute/liquidity issue rather than immediate economic protection. Competitors without similar contract coverage—Micron (MU) and Kioxia/Western Digital-linked NAND exposure where accessible—could face greater earnings volatility, but industry-wide supply discipline could make the contracts less differentiated than investors expect.

The contrarian opportunity is that consensus may still value SNDK on prior-cycle trough earnings despite a potentially improved revenue floor. Conversely, an 80% normalized gross-margin assumption leaves little room for execution error: any evidence that floor pricing produces materially lower utilization or margins would compress both estimates and the proposed multiple. The decisive 1-3 month catalyst is contract-detail disclosure and realized pricing versus NAND spot; the 6-18 month catalyst is a down-cycle stress test, not another favorable pricing quarter.

Key Decisions for Investors

• Keep SNDK on a conditional long watchlist rather than chase current strength. Initiate only after the next earnings release confirms contracted-bit mix, customer concentration, and floor-price margin economics; target a 12-month rerating from ~8x toward 10-12x normalized earnings if margins remain above management's long-run framework during softer spot pricing.
• Express the thesis as long SNDK / short MU in modest size after verifying contract enforceability. The pair isolates SNDK's proposed earnings-duration premium from NAND-cycle direction; exit if SNDK's realized ASP declines in line with spot NAND or if contracted coverage for the following fiscal year fails to expand.
• For downside protection around the next print, prefer a defined-risk SNDK put spread rather than an outright short while gross margins remain elevated. A downside trigger is guidance implying utilization cuts, gross margin below the company's normalized range, or receivables/deposit commentary that signals customer-performance risk.
• Do not infer a read-through to NVDA from this development. Monitor hyperscaler capex and enterprise-storage demand instead: a meaningful AI infrastructure slowdown would be bearish for SNDK's uncovered volumes even if accelerator demand remains resilient.