Sandisk's Memory Prices Could Double Again in Fiscal 2027, but Has the Market Already Priced That Into the Stock?

The article highlights that Sandisk (SNDK) is benefiting from a severe supply-demand imbalance in flash memory driven by AI data-center storage demand, enabling price hikes and potential further gains (Morningstar projects >100% pricing growth in fiscal 2026 and nearly +100% in fiscal 2027). However, it stresses the memory sector’s extreme cyclicality and warns that as supply comes online, Sandisk’s margins and profitability can swing sharply—citing a decline from a $1B net profit in fiscal 2022 to a $2B net loss in fiscal 2023. While the stock is down more than 25% from its peak, the piece argues it still looks expensive given risk of a subsequent demand/supply reversal and weaker longer-dated earnings (Micron fiscal 2028 estimates below 2027, with 2029 pointing to a major revenue drop).

Analysis

The market is likely still underestimating how fast peak-cycle earnings get discounted in memory. Even if near-term ASPs keep moving higher, this is a business where fixed R&D and fab/JV commitments make incremental downside asymmetric once supply catches up; the equity can de-rate before the income statement rolls over. The key distinction is that pricing power today is not the same as durable pricing power, and the spread between those two is where the short opportunity lives.

Second-order, the bigger winner may be the better-capitalized, more diversified memory names with differentiated AI exposure, while the pure NAND levered names face the most valuation risk. If storage costs keep inflating, cloud buyers can stretch replacement cycles, shift mix toward lower-density configurations, or delay non-critical AI deployments; that is a 1-3 quarter demand elasticity effect, not a long-term collapse. Over 6-18 months, the more important catalyst is capacity coming back online, which typically flips the market from scarcity to inventory digestion quickly.

Consensus appears to be treating the current pricing wave as semi-structural because AI demand is strong. That likely overstates durability: AI does support storage growth, but it also incentivizes supply response and customer procurement discipline, which caps the upside after the initial squeeze. The thesis is falsified if management or peers show another leg of disciplined supply plus sustained ASP acceleration into the next two reporting cycles; absent that, the more likely path is multiple compression before earnings peak.

Key Decisions for Investors

• Short SNDK on further strength, ideally after a gap-up or pre-earnings run; thesis is 1-3 month multiple compression rather than immediate fundamentals. Risk: if memory ASP guides keep inflecting higher, cover quickly.
• Prefer a relative-value short SNDK / long MU pair. MU should absorb more of the AI-memory spend with a more diversified product mix, while SNDK is more exposed to NAND cyclicality and peak-cycle de-rating.
• Use SNDK 3-6 month put spreads instead of outright shorts if borrow or squeeze risk is a concern. Structure for a 2:1 to 3:1 payoff if the stock gives back just a portion of its recent rerating.
• Watch for a reversal signal: any peer capex announcement or inventory build that suggests supply discipline is breaking. If NAND pricing still rises but company guidance stops outperforming, that is the point to press shorts.
• If you need a cleaner basket expression, short a memory-heavy basket via SOXX/XSD against a long AI compute leader instead of single-name exposure; the trade works best if the market rotates from scarcity winners to durable compounders.