Why Sandisk Stock Soared 29% in August

SanDisk’s fiscal Q4 (ended July 3) delivered a 372% YoY revenue surge and widened gross margin to 84.6% (from 26.2%), while EPS rose 91% sequentially to $43.97. For fiscal Q1 2027, management guided to $10.5B revenue at the midpoint (up 357% YoY) and gross margin of 83%–84.9%, with NAND demand still outstripping supply. Despite earlier investor concerns about how long the memory upcycle can last, the quarter’s results appear to have reset sentiment and helped the stock regain momentum, though it remains well below its prior peak.

Analysis

SNDK looks like a textbook example of a cycle turning into a narrative premium: the fundamental inflection is real, but the market is likely extrapolating peak scarcity into a multiyear growth story. The key mechanism is that memory is becoming a strategic AI enabler, yet memory is still a supply-response business; once margins stay this rich for a few quarters, competitors and idle capacity will re-enter, which is what eventually compresses valuation even if earnings keep rising.

The biggest second-order winner is probably the broader memory stack, but not all at once. If SNDK is able to convert more customers to longer-dated commitments, that lowers volatility and could re-rate the equity from a pure cyclically depressed multiple toward a quasi-utility on the first leg; however, it also telegraphs to rivals that pricing power is attracting supply, which is negative for the next phase of the cycle. WDC is the cleaner read-through loser on residual exposure and lower-quality earnings, while NVDA beneficiaries are more indirect: AI deployment keeps pulling storage content per rack higher, but this is still a smaller dollar pool than compute.

Timing matters: over the next 1-3 months, the stock is most vulnerable to any signal that gross margins or backlog growth are peaking faster than expected; over 6-18 months, the risk is capacity additions from larger memory vendors and a normalization of spot pricing. The contrarian view is that consensus may be underestimating how long inference-driven storage demand can persist, but it may also be overpaying for permanence in a business that has historically mean-reverted hard once supply catches up.

My base case is not to chase strength outright here; the setup is better as a relative-value trade than a naked long. The bullish thesis is only falsified if management’s commitment model stops converting into durable backlog, or if margin guidance rolls over before the next two quarters, which would signal the cycle is more advanced than the market wants to believe.

Key Decisions for Investors

• Long SNDK / short SOXX for 1-3 months: express memory-specific upside while reducing broad semi beta; target is SNDK continuing to outperform if backlog converts into visible earnings revisions, with the short leg cushioning a sector de-rating.
• Consider a tactical long SNDK only on a pullback rather than after strength: the risk/reward is better if the stock de-rates on any macro wobble and then re-rates into the next guidance update.
• Short WDC vs long SNDK on a 3-6 month horizon: SNDK has the cleaner pricing-power narrative, while WDC carries more residual exposure to memory normalization and less optionality from the AI inference story.
• Buy downside protection on SNDK into the next earnings/guidance window if implied vol is still cheap: the main risk is not earnings miss, but a sharp multiple contraction once the market starts pricing in supply response.
• Watch for a reversal trigger in spot NAND pricing or commentary from competing memory suppliers; if pricing stops tightening, reduce exposure quickly because the supercycle thesis can unwind faster than reported EPS suggests.