Sandisk's Q3 Earnings: The Market Is Still In Denial (Review)

Sandisk delivered a standout Q3 2026, with revenue up 252% YoY and adjusted gross margin of 78.4%, both far above expectations. Management guided Q4 revenue to $7.75B-$8.25B and EPS to $30-$33, authorized a $6B share buyback, and said the company remains debt-free. The results were driven by a shift into higher-value data center SSDs and multi-year customer agreements with firm commitments.

Analysis

The market is likely underestimating how quickly this shifts Sandisk from a cyclical storage supplier into a quasi-contractual infrastructure beneficiary. When a memory vendor locks in multi-year volume with firm financial commitments, the earnings stream behaves less like spot NAND and more like a capacity-constrained toll road; that should compress the appropriate discount rate and support a meaningfully higher multiple than peers still exposed to inventory swings.

The biggest second-order effect is on competitors that remain heavily retail/consumer weighted or lack committed hyperscale mix. If Sandisk can sustain this margin structure, it forces the rest of the NAND complex to either accept lower utilization or chase AI/data-center demand at less favorable terms, which can pressure industry discipline in the next 1-2 quarters. The buyback also matters: with a zero-debt balance sheet and a large repurchase authorization, management has a credible mechanism to absorb volatility and amplify per-share EPS even if revenue growth normalizes.

The main risk is that the market extrapolates peak conditions too far out. Gross margin this high likely reflects an unusually favorable mix and tight supply-demand balance; if procurement normalizes or customers re-negotiate on volume, the stock could re-rate sharply in either direction over the next 2-3 earnings prints. Another tail risk is concentration: if a handful of large customers slow deployment or defer qualification cycles, the step-function earnings profile can revert faster than consensus expects.

The contrarian angle is that this is less about a one-quarter beat and more about structural proof of pricing power in a market long assumed to be commoditized. The move may still be underowned by long-only investors because memory has historically been traded as a mean-reversion business; if that framework is wrong, the multiple expansion could continue for several months, especially if the company keeps converting guidance into cash and capital returns.

Key Decisions for Investors

• Go long SNDK on any post-earnings consolidation over the next 1-2 weeks; risk/reward favors buying pullbacks if the stock holds above the post-print breakout level, with upside driven by multiple expansion rather than just earnings revisions.
• Pair trade: long SNDK / short a more retail-exposed memory name or broad hardware supplier over the next 1-3 months to isolate the premium for contracted data-center exposure and capital-return credibility.
• Use SNDK Jan/Apr calls to express upside while capping gap-risk; the setup favors convexity because the stock can re-rate quickly if the market starts capitalizing the next few quarters as structurally different.
• Fade crowded skepticism only after checking customer concentration and forward bookings; if those remain intact, maintain a tactical long, but reduce if the next quarter shows any sign of margin normalization or guide conservatism.
• Watch for supplier read-throughs in NAND equipment and data-center storage peers over the next 4-8 weeks; a positive read-through would confirm that this is an industry upcycle with Sandisk as the premium winner.