Prediction: Sandisk Stock Will Soar to $5,000 in 2 Years

Sandisk’s stock has already surged more than 7x in 2026, and the article argues the rally can continue as AI-driven data-center demand and tight NAND supply support further pricing gains. Mizuho and BofA raised price targets to $2,200 and $2,100, respectively, while the article cites fiscal 2028 EPS of $188.78 and a possible share price of $5,098 if the stock trades near the Nasdaq-100’s 27x forward P/E. The piece is bullish on earnings momentum and analyst upside, but it is opinion-driven rather than new company-reported news.

Analysis

The market is still treating this as a cyclical pricing story, but the second-order setup is more structural: AI is not just lifting near-term NAND pricing, it is reallocating industry capex away from NAND capacity toward HBM, which mechanically tightens future flash supply. That means the margin expansion can persist longer than a normal demand spike because supply is being constrained by rational competitor behavior, not just temporary underinvestment. In that regime, the pure-play exposure matters more than the broader storage names: SNDK has cleaner operating leverage than diversified peers, while STX and WDC are more exposed to any eventual normalization in HDD replacement demand.

Consensus likely underestimates the convexity in earnings because it is extrapolating from revenue growth rather than modeling the spread between spot pricing, mix, and utilization. If enterprise SSD adoption keeps displacing HDDs in AI infrastructure over the next 12-24 months, the earnings inflection can outpace even aggressive analyst revisions, because each incremental unit ships into a much richer pricing environment. The key nuance is that the trade is less about end-demand “surprise” and more about the lagged response of the supply base; that lag can keep quarterly numbers ahead of Street estimates for multiple prints.

The main risk is timing, not thesis: if HDD capacity comes back sooner than expected or if memory producers reverse course and prioritize NAND over HBM, the scarcity premium compresses quickly. The other risk is sentiment overshoot — after a >7x move, any earnings miss or guide conservatism could trigger a violent multiple reset even if fundamentals remain strong. That makes the next 1-2 earnings cycles the critical window; beyond that, the stock needs continued upward estimate revisions rather than just stable fundamentals.

Contrarian angle: the market may be paying for a 2028 earnings outcome today, but not enough for the possibility that the cycle peak arrives earlier and at a higher margin than expected. If that happens, SNDK can keep rerating even without further estimate beats, because the numerator and the multiple can expand simultaneously. Relative to NVDA, which is already widely owned for AI capex, SNDK is a less crowded way to express the same infrastructure buildout with a different supply-demand bottleneck.

Key Decisions for Investors

• Long SNDK on 1-2 post-earnings dips with a 6-12 month horizon; favor adding only if management confirms supply tightness and forward ASPs remain firm. Risk/reward remains attractive as long as estimate revisions stay positive, but trim if the stock stops outperforming after a beat.
• Pair trade: long SNDK / short STX or WDC for 3-6 months. Thesis is that pure-play NAND leverage and tighter pricing should outperform diversified storage exposure if the AI storage cycle persists; stop out if HDD supply relief accelerates or NAND pricing rolls over.
• Use call spreads instead of outright equity for SNDK into the next earnings cycle, e.g. 9-12 month bull call spreads, to capture upside while limiting multiple-compression risk after the recent run-up.
• Take profits incrementally on any 20-30% further rally unless estimates are revised materially higher. After a 7x YTD move, the stock is vulnerable to a sharp de-rating on even modest guidance conservatism.
• Watch NVDA as a proxy for AI infrastructure capex strength, but prefer SNDK as the higher-beta, less consensus-traded expression of storage bottlenecks; rotate part of AI exposure from crowded semiconductor winners into SNDK if you want operating leverage to the same theme.