$2,600 for Sandisk Stock? Why the Bull Case at Evercore Forecasts a Whopping 191% Upside Potential.

Sandisk reported a strong Q2 2026 beat, with revenue up 61% year over year to $3.0B and EPS up 404% to $6.20, ahead of the $3.62 consensus. Management guided Q3 revenue to $4.4B-$4.8B versus $4.66B expected and EPS to $12-$14, while Evercore initiated coverage with an Overweight rating, a $1,200 target, and a $2,600 bull case. The stock’s AI storage exposure, margin expansion to 51.1%, and stronger cash generation are driving a bullish re-rating narrative.

Analysis

The market is beginning to price NAND as an AI infrastructure bottleneck rather than a commodity sub-cycle. That matters because storage demand tied to training and, more importantly, inference tends to be stickier and more contractable than server memory demand, which should compress earnings volatility and raise the terminal multiple for the few vendors with credible high-density enterprise roadmaps. The second-order winner is not just SNDK; it is also the equipment and controller ecosystem that benefits from longer fab utilization and higher ASP discipline, while weaker consumer-exposed storage franchises get squeezed on mix.

The key surprise is that the earnings power is now being de-risked by balance sheet and contract structure, not just pricing. Upfront cash payments and multi-year supply commitments create a quasi-utility profile for a business that historically traded as a pure cycle name, which can keep downside contained even if spot NAND cools. That said, the current rerating assumes execution stays clean through multiple qualification cycles; any delay in ramping high-capacity SSDs or a slip in hyperscaler acceptance could hit sentiment quickly over the next 1-2 quarters.

Consensus may be underestimating how much of the upside is already in the stock if NAND supply discipline holds through 2026-2028. The bigger contrarian risk is that the market is treating AI storage demand as linear, when hyperscalers often over-order early and then digest inventory; if that happens, the forward multiple can compress faster than earnings rise. A second-order bear case is competitive response: consolidation talk and aggressive capital allocation by peers could eventually blunt pricing power, especially if the industry interprets current margins as an invitation to add capacity in late 2026/2027.

Key Decisions for Investors

• Long SNDK vs. short a basket of lower-quality memory exposure over 3-6 months; target is continued multiple expansion from contract visibility, but size tightly because the name is already rerated and sentiment is crowded.
• Buy SNDK on pullbacks only, ideally after a 7-10% retracement following any strong move; risk/reward is best if the market gives a cleaner entry before the next earnings print.
• Use call spreads rather than outright longs in SNDK for the next 1-2 quarters; upside remains meaningful, but implied optimism and already-high expectations argue for defined-risk exposure.
• Watch EVR as a weak read-through beneficiary; the initiation helps validate the narrative, but the name is more catalyst-dependent and should be treated as a trading signal rather than a core position.
• Avoid WDC as a proxy long for this theme until there is evidence of similar contract durability and mix shift; if NAND tightness is structural, the cleaner expression is SNDK, not the more diversified laggards.