Sandisk Stock Is Up 1,500% in the Past Year Due to AI -- Is It Still a Buy? Wall Street Has a Surprising Answer for Investors.

SanDisk reported fiscal Q2 2026 revenue of $3.0 billion (up 61% YoY) and non-GAAP EPS of $6.20 (up 404% YoY), and guided Q3 to $4.6 billion in revenue with non-GAAP EPS of $13.00 at the midpoint, driven by a severe NAND flash supply shortage and strong data-center demand. The company gained 2 percentage points of NAND market share through September 2025, Wall Street projects adjusted earnings to grow ~156% annually to June 2027 and values the stock at roughly 80x earnings; analyst targets range from $235 to $1,000 with a median of $690. The combination of outsized near-term earnings growth and elevated memory pricing supports the rally, but the cyclical nature of semiconductors poses meaningful downside risk if supply/demand dynamics normalize.

Analysis

Market structure: The immediate winners are SNDK, its Kioxia JV, and semiconductor equipment suppliers (e.g., LRCX, AMAT) as NAND prices rose ~50% in Q4 2025 and are projected +40–50% into Q1 2026. Losers include incumbent memory leaders (Samsung, SK Hynix) losing share and OEMs with fixed BOMs; higher SSD/DRAM prices will compress margins for PC/phone OEMs within 1–2 quarters. Cross-asset effects include higher equity volatility in semis, modest upward pressure on capex-led corporate borrowing (may widen IG spreads short-term), and potential JPY flows into Kioxia-linked assets if Japan-centric capex ramps.

Risk assessment: Key tail risks are a rapid industry capex response producing a supply glut within 12–24 months, US-China export restrictions disrupting customer mix, or a Kioxia JV operational failure; any of these could swing SNDK ±40–60%. Near-term (days–weeks) earnings momentum could re-rate the stock; medium-term (3–12 months) analysts will reprice estimates; long-term (12–36 months) cyclicality likely compresses multiples from ~80x to 15–25x normalized EPS. Hidden dependency: SNDK’s growth is contingent on Kioxia wafer supply and multi-quarter hyperscaler contracts that can be renegotiated.

Trade implications: Favor a size-limited long in SNDK to capture upside to consensus $690–$1,000 targets but hedge for cyclicality via defined-risk options and pair trades. Specific tactics: buy call spreads to limit premium, or pair SNDK long vs. short WDC or Samsung to isolate NAND share gains. Entry should be staggered: initial buy now, add on pullback >20% or after confirmed Q3 guidance delivery; trim at +20–40% or on any memory-price reversal >10% QoQ.

Contrarian angles: The consensus underestimates time-to-supply relief — fab ramps take 12–24 months, which supports SNDK for now, but the market may overpay for transitory earnings (1–2 year window). The 1,500% rally since Feb 2025 suggests momentum risk; historical DRAM/NAND cycles (2017–2019) show sharp reversals once capex shifts. Unintended consequence: sky-high prices incentivize competitors to accelerate capacity, creating a 12–18 month mean-reversion risk that can produce -50%+ drawdowns in semis.

Key Decisions for Investors

• Establish a size-limited long in SNDK equal to 2–3% of portfolio NAV: initiate 50% of allocation at market (~$576) and add remaining 50% on any pullback to ≤$460 (20% drop). Take profits to reduce to 1% at $690 (median target) and fully trim to 0% if SNDK reaches $1,000 or memory spot prices roll over >10% QoQ.
• Implement a defined-risk options trade for leveraged upside: buy SNDK Jul 16 2026 600C / 900C call debit spread (max loss = premium) sized so max loss = 0.5% portfolio NAV; roll or sell if spread value >3x or underlying falls >25% in 30 days.
• Run a pair trade: long SNDK (1–2% NAV) vs short WDC or SSNLF (equal dollar) to capture relative NAND share gains; set pair stop-loss if SNDK underperforms by 15% relative over 60 days or if Samsung/Kioxia announce ≥20k wafer starts expansion.
• Rotate 1–2% from consumer OEMs (PC/phone vendors) into semiconductor equipment names (LRCX, AMAT) to play sustained capex; reduce OEM exposure further if memory spot prices increase inventory-day count >30% over 2 consecutive quarters.
• Monitor two specific catalysts for active management: (A) official announcements of new fab capacity from Samsung/Hynix/Kioxia within next 90–360 days (trim/hedge if capacity adds ≥15% industry supply), and (B) memory spot-price decline >10% QoQ (initiate protective hedges or trim longs immediately).