Back to News
Market Impact: 0.25

What Sent Sandisk Shares Tumbling Nearly 20% Last Month

Artificial IntelligenceCredit & Bond MarketsAnalyst EstimatesAnalyst InsightsCompany FundamentalsTechnology & Innovation
What Sent Sandisk Shares Tumbling Nearly 20% Last Month

Sandisk (SNDK) shares fell nearly 20% over the past month amid a late-July semiconductor sell-off that wiped out about $1.3T in market value across major chip stocks. The article attributes the pullback partly to concerns about AI infrastructure spending efficiency and potential future supply catch-up for memory/storage, which could eventually weaken pricing power. Despite the recent drop, analysts remain constructive—77% of 30 tracked analysts rate SNDK a buy, with a median 1-year price target of $2,500, implying material upside from the ~$1,400 entry referenced on Aug. 5.

Analysis

The market is pricing the first real crack in the AI-scarcity narrative, but the setup is more nuanced than a simple de-rating of semis. SNDK’s earnings power is tied to pricing discipline in NAND-like storage, which tends to persist longer than investors expect because capex cuts only show up with a lag; the immediate air pocket is sentiment, not fundamentals. The bigger second-order risk is that a slowdown in hyperscaler capex would first hit NVDA multiple assumptions, then wash into storage names as customers defer large system builds.

Near term, the selloff can still overshoot because crowded AI winners tend to de-gross together when investors question the return on spend. Over the next 1-3 months, watch for commentary from MU and the hyperscalers on inventory, pricing, and enterprise SSD demand; if contract pricing holds and lead times stay extended, the drawdown is likely a trading event rather than the start of a cycle break. Over 6-18 months, the key falsifier is a visible capex reset or a rapid increase in NAND output that erodes pricing power faster than expected.

The contrarian view is that the market may be underestimating how long scarcity can last even if AI model efficiency improves. If storage remains the bottleneck while compute narratives cool, SNDK can still compound, but the stock likely trades more on multiple stability than on pure earnings growth from here. Relative value is cleaner than outright beta: the better risk is to own the most liquid beneficiaries of the tight supply backdrop and hedge the parts of the AI complex most exposed to capex disappointment.

More News