Why Sandisk Stock Was on Fire This Week

Sandisk’s fiscal Q3 2026 revenue more than tripled year over year to $5.95 billion, with adjusted EPS of $23.41 versus a $14.66 consensus and a swing to nearly $3.68 billion in adjusted net income. Management guided Q4 revenue to $7.75 billion-$8.25 billion and adjusted EPS to $30-$33, both well above expectations, driven by very strong storage demand tied in part to AI infrastructure. The stock was up more than 11% week to date, and the results also lifted sentiment across the storage peer group.

Analysis

The key second-order read-through is that storage is moving from a cyclical commodity lane into a capacity-constrained AI infrastructure bottleneck. When both flash and disk peers are printing outsized upside simultaneously, the market should think less about isolated earnings beats and more about a multi-quarter pricing regime where supply tightness, not just unit growth, is the dominant variable. That tends to support gross margin durability across the chain, but it also raises the risk of over-ordering in the back half of the cycle as customers lock in inventory to avoid allocation risk.

For SNDK specifically, the blowout quarter is likely to compress forward estimate dispersion rather than just lift the stock. The bigger implication is that NAND pricing strength can spill into capex decisions for cloud and OEM customers: if storage remains elevated, some AI buildouts may shift spend toward compute later than expected, creating a temporary relative tailwind for storage vendors but a medium-term brake on broader AI hardware sentiment. Seagate’s strength reinforces that the market is rewarding scarcity, not just execution, which can keep multiple expansion going for a while even as absolute growth normalizes.

The contrarian risk is that consensus is extrapolating peak scarcity into a sustained supercycle. Storage is historically one of the fastest sectors to self-correct once pricing signals attract capacity or customer procurement pauses; that reversal often shows up with a 2-3 quarter lag, not immediately. The cleaner trade is to stay with the trend near term, but be ready for a sharp reset if management commentary turns from "supply constrained" to "normalizing lead times" or if hyperscaler procurement becomes more selective.

NVDA and INTC are only marginally implicated today, but the broader AI hardware stack matters: if storage spend is absorbing more of the budget than expected, there is a potential rotation away from the most expensive compute names into underappreciated infrastructure beneficiaries. That argues for being long the picks-and-shovels with direct pricing power now, while avoiding names where the narrative already assumes endless AI capex acceleration.

Key Decisions for Investors

• Long SNDK for 4-8 weeks on momentum and estimate revisions; use a tight stop if the stock fails to hold post-earnings gains, since the upside is driven by multiple expansion plus upward guidance revisions, but the trade will unwind quickly if the market senses peak pricing.
• Pair trade: long SNDK / short a basket of high-multiple AI compute beneficiaries (e.g. NVDA on a relative basis) for 1-3 months; thesis is that storage scarcity is being underowned while compute valuations already discount a longer-duration capex boom.
• Overweight STX versus the broader semiconductor/AI hardware complex for the next quarter; the peer confirmation suggests durable industry pricing power, and STX offers a cleaner way to express the storage supercycle with lower execution risk than a smaller-cap name.
• Buy SNDK downside protection 2-3 quarters out via put spreads if you are long cash equity; the asymmetry is attractive because the market can keep rewarding scarcity now, but supply response or customer inventory digestion could re-rate the group abruptly.
• If you want a cleaner macro hedge, pair long storage names against short a basket of enterprise hardware that relies on cheap storage inputs; if storage costs stay elevated, margin pressure should show up first in less pricing-powerful downstream equipment makers.