Sandisk Is Up More Than 35-Fold in a Year and Sits Nearly a Third Below Its Peak. History Says What Comes Next.

Sandisk (SNDK) is still about 32% below its June 22 peak ($2,354.39) despite a 35x+ run up from its $43.20 52-week low; shares are near ~$1,600 and down ~10% on the day. The article highlights high earnings volatility from pricing—gross margin surged to 84.6% in fiscal Q4 2026 (vs 26.2% a year earlier) on revenue of $8.97B (+372% YoY). Management guided fiscal first-quarter revenue of $10.3B–$10.8B (from $8.97B) with gross margin expected at ~83%–85%, but the stock’s valuation implies the market may still be discounting near-term pricing risk.

Analysis

This is still a cyclical pricing story wearing a growth-stock costume. The important market mechanism is not the near-term revenue growth rate; it is how much of that growth was price-led versus structurally demanded, because price-led upside can unwind much faster than consensus models update. The forward multiple looks optically cheap only if memory pricing stays elevated; if pricing normalizes, the earnings base can fall off a cliff and the stock can de-rate even without a collapse in unit demand.

The real second-order winner from a softer memory tape is downstream OEM margin relief: PC, handset, storage, and AI server buyers get a BOM tailwind, which is incrementally positive for large platform names and negative for the memory suppliers most exposed to spot pricing. Within semis, MU is the cleaner proxy for a sector-wide read-through because it has enough scale to transmit cycle turns, while SNDK is more vulnerable to narrative volatility given the recent parabolic rerating. If memory stays tight, though, the supply discipline embedded in contract coverage should cushion the downside versus a pure spot name.

Near term, the tape can stay ugly for days to weeks if the group is de-risking, but the true catalyst is 1-3 months of pricing checks and the next margin guide. The thesis breaks if SNDK can hold margin in the low-80s while revenue keeps inflecting, or if channel data shows the feared inventory rebuild never happened. The consensus is probably overestimating how much historical analogies matter: this is less about whether the stock is down 30% from the peak and more about whether the memory cycle has already peaked.

Key Decisions for Investors

• Initiate a tactical short in SNDK on any intraday rebound, using a 1-3 month horizon; thesis is multiple compression if memory pricing rolls over. Risk/reward is attractive if the stock fails to reclaim the post-investor-day range, but cover immediately if gross margin guidance remains above 83% and revenue re-acceleration persists for a second quarter.
• Prefer a relative-value pair: short SNDK / long MU for 1-3 months, betting the market punishes the name most exposed to a perceived price peak while MU benefits from a less euphoric setup. This is a cleaner expression than outright short memory, but it should be abandoned if industry pricing data turns up again or MU management sounds more bullish than expected on supply discipline.
• Watch, do not chase, any weakness in NVDA and TSLA tied to memory-group fear; if the move spills into AI or EV platform names, it is likely an overreaction and potentially a buying opportunity on a 2-6 week horizon. The falsifier is sustained BOM inflation or supplier commentary that memory shortages are extending AI server lead times.
• If you need options, use SNDK put spreads rather than naked shorts; the stock can remain disconnected from fundamentals longer than expected, but downside can be sharp once pricing expectations reset. A close back above the prior peak would invalidate the bearish setup and argues for taking profits or hedging.