Sandisk Stock Is Up 770% This Year and Polymarket Just Called It "the Most Overbought Stock in History." Here's the Case for Buying Anyway.

Sandisk reported fiscal Q3 revenue of $5.95 billion, up 251% year over year, with data center sales surging 645% to $1.46 billion and net income swinging to $3.6 billion from a $1.9 billion loss. The stock has risen 726% this year and 5,370% since the Western Digital spinoff, though its RSI of 99 and forward P/E of 34.7 highlight stretched technical and valuation conditions. Overall, the article argues the AI-driven fundamentals remain strong despite overheating signals.

Analysis

The market is treating SNDK less like a cyclical storage vendor and more like a constrained AI infrastructure bottleneck, which is why valuation multiples that would normally cap upside are being ignored. The key second-order effect is that every incremental dollar of AI capex is not flowing uniformly across semis; it is concentrating into memory, packaging, and storage layers where supply is tighter and re-rating can persist longer than in the headline GPU names. That argues this move is not just sentiment-driven—it's a regime shift in how the market prices infrastructure scarcity.

The bigger risk is not that the stock is "overbought" in a technical sense, but that expectations have outrun the pace at which the data center mix can compound from here. When a stock is already discounting multi-quarter execution, the next failure point is usually not demand but cadence: any sign of gross margin normalization, slower backlog conversion, or a pause in hyperscaler ordering can hit the multiple first and the earnings later. In other words, the near-term trade is about flow and positioning; the medium-term risk is that the earnings power becomes less incremental than the chart implies.

The contrarian read is that the market may still be underestimating how structurally levered SNDK is to AI memory intensity, especially if inference workloads keep raising storage demand per deployed compute dollar. If that dynamic holds, the right comparison set is not legacy storage peers but AI picks-and-shovels with scarce capacity and operating leverage. That said, this is a classic name where upside can continue while the risk/reward becomes more asymmetric by the week, so sizing and timing matter more than directional conviction.

Key Decisions for Investors

• Stay long SNDK, but only as a tactical momentum position for 2-8 weeks; use a trailing stop or a close below the 20-day moving average as the exit trigger because the risk is a sharp de-rating once positioning gets crowded.
• For new money, prefer a call spread in SNDK over outright stock exposure: buy 1-2 month 5-10% out-of-the-money calls and sell the next strike up to monetize elevated implied volatility while keeping upside to continued squeeze.
• Pair trade: long SNDK / short a basket of slower-growth storage or legacy hardware names over the next quarter to isolate AI-storage scarcity premium; the thesis works best if the market keeps rewarding earnings revision momentum over multiple expansion.
• Add NVDA and GOOGL only on weakness, not strength, because SNDK strength may reflect a rising tide in AI capex, but their upside from here is more execution-dependent and less levered to a scarcity rerate.
• Watch for the first sign of hyperscaler digestion in the next earnings cycle; if SNDK guides to even a modest slowdown in data-center growth, consider flipping to a short-term put spread because crowded winners tend to unwind 15-25% very quickly once the growth inflection narrative stalls.