If You Invested $30,000 in Sandisk Stock at Its Spinoff, Here's How Much You'd Have Today (Hint: You'd Be a Millionaire)

Sandisk is benefiting from a severe NAND memory shortage tied to AI data center build-outs, with analysts expecting 332% year-over-year revenue growth in fiscal Q4 2026 and 118% growth in fiscal 2027. The article argues that sustained hyperscaler capex—about $650 billion in 2026 and potentially over $1 trillion in 2027—could keep memory pricing elevated for years. While highly positive for Sandisk’s fundamentals, the piece also cautions that the stock’s upside is dependent on the ongoing AI infrastructure cycle.

Analysis

The key second-order read-through is that this is not just a Sandisk story; it is a pricing-cycle signal for the entire storage stack. When NAND tightens this hard, gross margin expansion usually cascades first to the most supply-constrained vendors, then to module makers and SSD assemblers with less pricing power, and finally to hyperscalers through higher data-center capex per unit of compute. The market is still underestimating how sticky pricing can be once procurement teams lock in multi-quarter supply agreements, which can keep spot strength feeding into reported results for several quarters even if end-demand cools modestly.

The more interesting question is duration. If AI infrastructure spend stays on its current trajectory, the bottleneck shifts from accelerators to memory density and storage refresh cycles, which means the shortage can persist longer than the usual 6-9 month commodity upcycle. That said, the trade becomes fragile once capacity additions, wafer starts, or inventory normalization begin to compound; memory is one of the fastest industries to overshoot on the downside after a period of scarcity. The most likely reversal catalyst is not demand collapse but supply response plus customer inventory de-risking, which can hit the stock before fundamentals visibly roll over.

Consensus is probably too linear on both revenue and terminal multiple. The market is likely extrapolating peak-like growth rates into a “new normal,” but the right framing is that Sandisk is a high-beta beneficiary of a temporary but potentially multi-year bottleneck, not a structurally defensible moat compounder. That argues for owning the trade, but with disciplined expression: upside is still substantial if the cycle extends, yet the valuation can de-rate violently once investors start modeling normalization rather than scarcity.

The cleaner expression may be to prefer the scarcer, more levered exposure to the shortage over the broad AI semiconductor basket, while hedging against a storage-cycle mean reversion. The biggest knock-on risk to hyperscalers is budget crowd-out: higher memory costs can force more selective deployment of lower-ROI AI projects, which would eventually pressure the more speculative parts of the AI trade even if near-term capex remains elevated.

Key Decisions for Investors

• Maintain a tactical long in SNDK for 1-3 months, but size it as a cycle trade rather than a core holding; upside remains meaningful while supply remains tight, but trailing stops should be tight because memory names can lose 20-30% quickly on any inventory or capacity headline.
• Pair trade: long SNDK / short WDC over the next 2-4 quarters to isolate pure NAND pricing leverage versus the broader legacy storage mix; the spread should outperform as the market rewards the tighter asset base and stronger operating torque.
• Buy NVDA puts or finance them with call spreads against a SNDK long if you think memory inflation will eventually crowd out AI capex; this hedges against a second-order slowdown in lower-return data center projects over the next 6-12 months.
• Use strength to harvest gains: trim 25-50% of SNDK on any further multiple expansion without a corresponding guidance beat, because the stock can overshoot fundamentals long before the earnings base catches up.
• For aggressive accounts, consider a call spread in SNDK with 3-6 month expiry to capture continued upside from contract repricing while limiting the blow-up risk if supply normalizes faster than expected.