Why Sandisk Stock Just Dropped

Sandisk stock hit an all-time high after Melius Research reiterated a buy and set a $1,350 price target, citing unusual AI-driven demand for memory chips. The stock then fell 4.7% as a report of slower OpenAI user and revenue growth raised concern, though the article argues OpenAI is still spending heavily on AI chips and memory. Overall, the piece remains constructive for Sandisk’s demand outlook despite short-term volatility.

Analysis

The key takeaway is that this is not a demand problem for Sandisk; it is a duration problem for the AI capex trade. If frontier-model economics are wobbling, hyperscalers and model labs will first protect compute that drives inference and training continuity, but they will compress everything non-essential around it later in the cycle. That makes memory pricing more resilient than the market typically assumes in a headline-driven selloff, because memory is a relatively small line item versus GPUs, networking, and power, yet it is hard to substitute once deployed.

The second-order winner is not just SNDK, but the whole “picks and shovels beneath the picks and shovels” stack: NAND, HBM-adjacent packaging, and controller vendors should see a longer-than-normal pricing tail if AI buildouts remain front-loaded even as end-demand normalizes. The risk is that investors conflate persistent procurement with persistent returns on capital; if OpenAI-style economics force a broader re-rating of AI monetization, memory names can still correct hard even while units stay tight, because the market will discount 2026-27 ASPs sooner than spot chatter suggests.

Near term, the stock is vulnerable to violent mean reversion after an analyst-led breakout, but that usually creates better long entries than chasing strength. The contrarian signal is that a slowing flagship customer can actually extend the memory cycle if its capex is locked in and cancellation risk is low; the bear case only wins if AI buyers shift from inventory accumulation to digestion over the next 1-2 quarters. In other words, the trade is less about whether AI demand exists and more about whether the market has already capitalized the next year of constrained supply.

Key Decisions for Investors

• Buy SNDK on a 1-2 day post-headline pullback, but size as a tactical trade only; target a 10-15% rebound over 4-6 weeks with a stop if the stock loses the prior breakout level on volume.
• Prefer a pair trade: long SNDK / short a broad AI-infrastructure proxy over 1-3 months to isolate memory-scarcity upside while hedging a generic AI de-rating; this works best if AI sentiment sours but capex remains intact.
• Add exposure to memory beneficiaries with cleaner operating leverage than SNDK, particularly when semiconductor breadth weakens; the risk/reward is better in names that lagged the breakout and still have room for estimate revisions.
• Sell upside calls against existing SNDK longs if implied vol stays elevated after the analyst-driven spike; the setup favors range expansion less than it favors a sharp but temporary squeeze.
• Monitor any confirmation of capex digestion from large AI buyers over the next quarter; if procurement slows, use that as the exit signal because memory multiple compression will likely precede any spot-price deterioration.