SanDisk +3,685% in 1 Year, Has the NAND Cycle Ended?

TL;DR

  • SNDK delivered Q3 FY26 revenue of $5.95B (+251% YoY, +97% QoQ), Non-GAAP EPS of $23.41 (+59.7% beat), and 78.4% gross margin, among the strongest quarters in memory history. Shares ran +46.8% over 6 sessions post-earnings to a fresh 52-week high
  • NBM (New Business Models) multi-year supply contracts are the core: 5 deals signed through Q3, $42B+ revenue locked in, $11B+ in financial guarantees, and one-third of FY27 bit supply already committed. Structurally dampens memory cyclicality
  • 12M PT $1,800 (Implied Upside +15.2%). But beta 5.04, MA200 +304% stretch, and a record-high 78% GM make chasing inadvisable. We recommend staggered entry in the $1,300–1,400 zone

SanDisk +3,685% in 1 Year, Has the NAND Cycle Ended?

"From cycle to contract, memory redefined"

ItemValue
RatingHOLD (Tactical) / BUY (12M)
12M Price Target$1,800 (probability-weighted)
Implied Upside+15.2% (vs $1,562 current)
Current Price$1,562.34 (2026-05-08, ~₩2.267M)
Market Cap~$247B / ~₩358T
52-Week High/Low$1,564.00 / $35.79
YTD Return+558.2%
1-Year Return+3,685.7%
Next Earnings2026-08-13 (Q4 FY2026)

⚠️ Caveat for this report: SNDK is not a SaaS name; it's a semiconductor / NAND memory business. Instead of NRR / ARR / Rule of 40, we use ASP, Bit Shipments, NAND ASP, and NBM contract backlog as core KPIs. The company is in its second listed year following its 2025-02-24 spin-off.

1. Executive Summary

SanDisk's Q3 FY2026 (reported 2026-04-30) was one of the strongest quarters in the history of the memory industry. Revenue $5.95B (+251% YoY, +97% QoQ), Non-GAAP EPS $23.41 (vs consensus $14.66, +59.7% beat), and GM 78.4% (vs 51.1% prior quarter), numbers a cyclical memory company simply cannot produce under normal conditions.

The more decisive change is the NBM (New Business Models) multi-year supply framework, five contracts through Q3, $42B locked in + $11B in financial guarantees + $400M in prepayments. Shares surged +46.8% over six sessions post-print, setting fresh 52-week highs.

We maintain 12M BUY ($1,800 PT, +15%) but rate the name HOLD on a tactical basis, chasing is not advised. The most recent 9 sessions delivered +46.8%, leaving RSI stretched. We recommend staggered accumulation in the $1,300–1,400 zone. Q4 print (8/13) needs to confirm additional NBM signings and the magnitude of Stargate revenue recognition.

2. Pre/Post-Earnings Timeline

DateCloseDaily %VolumeKey Event
2026-04-29 (D-1)$1,064.21+3.0%11.4MPre-print, consensus revenue $4.73B
2026-04-30 (D)$1,096.51+3.0%15.1MReports $5.95B revenue, $23.41 EPS
2026-05-01 (D+1)$1,187.00+8.3%23.1MMarket digests NBM lock-in
2026-05-04 (D+2)$1,255.86+5.8%17.9MSusquehanna PT $1,000 → $2,000
2026-05-05 (D+3)$1,406.32+12.0%20.7MCantor PT $1,800 + Bernstein upgrade
2026-05-08 (D+6)$1,562.34+16.6%20.6MFresh 52-week high, cumulative +46.8%

3. Decomposing the +46.8% Move

ComponentEstimated ContributionJustification
Revenue beat +25.8% ($5.95B vs $4.73B)+12–15 ppt
EPS beat +59.7% ($23.41 vs $14.66)+8–10 ppt
Q4 guide mid $8.0B (+75% vs normalized expectations)+10–12 ppt
NBM $42B lock-in + $11B financial guarantee+8–10 ppt✅ Cycle discount unwind
Sell-side PT doublings (Susquehanna $1k→$2k)+5–7 ppt🟡
Momentum / FOMO / Motley Fool "$4,000 in 1 yr"+3–5 ppt🟡

→ Fundamental contribution ~38–47 ppt, meaning 80%+ of the move is justifiable. PT doublings and FOMO account for the speculative tail.

4. Fundamental KPIs (NAND Sector)

KPIQ1 FY26Q2 FY26Q3 FY26Q4 GuideTrendSignal
Revenue ($B)1.692.315.957.75–8.25Accelerating🟢
Non-GAAP EPS ($)-0.421.2223.4130–33Accelerating🟢
Non-GAAP GM (%)22.427.878.4~80Accelerating🟢
FCF Margin (%)-81449.7n/aAccelerating🟢
Bit Shipments QoQ+5%+12%-high teensn/aSlowing🔴
NAND ASP QoQ+20%+35%+85–90%+5–10%Accelerating🟢
NBM Backlog ($B)0042+New🟢
CapEx / Revenue (%)~22~15"substantially down"further downImproving🟢

Verdict: 7🟢 / 1🔴 → consistent fundamental acceleration. The shrinking bit shipment number flags ASP dependence as a structural exposure.

5. Q4 Guidance, No Peak Recognition

ItemQ4 GuideConsensusMarket Read
Revenuemid $8.0B~$4.5BMassive Beat (+75%)
Non-GAAP GM80%60s%Further upside
Non-GAAP EPSmid $31.5 ($30–33)$14–16+100% above

Management tone:

  • CEO Goeckeler: "Highly confident caution", emphasizes that NAND price acceleration is not transitory
  • CFO Visoso: Repeats "durable, more predictable", NBMs are the tool reshaping the business model
  • Signal: This is more than a beat-and-raise, it's a narrative shift from "memory company → multi-year contract company"

6. Earnings Call, CEO Key Quotes

#QuoteOur Reading
1"NAND has become a critical component of the underlying infrastructure (for AI workloads)"NAND elevated to AI infrastructure essential
2"These partnerships support durable, structurally higher earnings and a significantly more predictable and less cyclical business."NBMs structurally dampen memory cyclicality
3"We are not interested in trading away that value for certainty."Refusing pure long-term fixed-price deals; preserving pricing power
4"The market is moving very quickly, literally every day, and that makes it difficult to forecast."Signal that Q4 guidance is conservative
5"Hyperscalers want everything you can make and then some."Underscores supply tightness
6Frames BiCS 8 as the "gold standard for NAND technology"Kioxia JV extended through 2034
7QLC Stargate "begins shipping for revenue in the next quarter"Data center revenue further accelerates from Q4
8"HBF: NAND die later this year, system-level early-to-mid next year. HBF complements, not replaces, enterprise SSD"Preempts cannibalization concerns

CFO Visoso highlights:

  • Revenue $5.95B = $1.15B above the high end of guidance ($4.8B)
  • GM 78.4% = +27.3 ppt improvement
  • FCF $2.955B (49.7% FCF margin), "Spectacular cash generator"
  • Three of the five NBM deals signed in Q3 alone secure $42B in revenue and $11B+ in financial guarantees

7. ⭐ Underappreciated Signals (5)

  1. 49.7% FCF margin, SaaS-grade cash conversion that's hard to attribute to a memory business. The market still applies a cyclical multiple
  2. One-third of FY27 bits already locked in + $11B financial guarantees, structurally removes memory's biggest weakness, "no ASP visibility"
  3. Bit shipments QoQ -high teens, yet revenue +97%, achieved via price × mix alone. Bit normalization later provides additional operating leverage
  4. Kioxia JV extended to 2034, joint development capex on BiCS9/10 shared, justifying continued declines in CapEx as a percentage of revenue
  5. HBF is a new category, not an HBM substitute, inference KV cache is a domain outside SK Hynix and Samsung's HBM stronghold, and SanDisk is the first mover

8. 🟡 Yellow Flags (5)

  1. GM 78.4% is the highest in NAND industry history, normalization implies sharp deleverage
  2. Bit shipment decline, high-teens QoQ. If ASPs roll over, revenue deleverages simultaneously
  3. Valuation stretch, Forward EV/S 5.15x (industry avg 3.06x), YTD +558%, beta 5.04
  4. NBM "minimum" is a floor, not a ceiling, but in a price collapse the floor can effectively cap upside
  5. If Samsung and SK Hynix resume CapEx in 2026, 2H 2027 supply normalization is a real risk

9. Business Model (3-Box)

SegmentQ3 RevenueQoQGrowthMargin
Data Center$1.467B (~25%)+233%🟢🟢Very high
Edge (Client+Mobile)$3.16B (~53%)+118%🟢High
Consumer (Cards/USB)~$0.82B (~14%)-10%🟡 SeasonalModerate

We expect data center to cross 30% of revenue from Q4 onwards.

10. Peer Comparison

TickerNAND ShareRevenue YoYNAND ASP QoQYTD ReturnFwd EV/Sales
SanDisk (SNDK)12.4%+251%+85–90%+558%5.15x
Samsung (NAND only)32.3%n/a+70%Separaten/a
SK Hynix19.3%+50s%+45%+120%4.2x
Kioxia15.3%(adj GM 66%)+90%+180%3.8x
Micron (MU)~12%+60%n/a+85%3.0x

SNDK's YTD +558% reflects pure NAND exposure plus NBM differentiation, 3x+ outperformance vs the peer average. Forward EV/S of 5.15x represents a +47% premium to the peer average of 3.5x, whether NBMs justify it is the central PT variable.

11. 12M Price Target Derivation

Diluted shares 158M, cash $3.7B.

ScenarioAssumptionFY27E RevenueEV/SalesPTProbability
BullNBMs end cycle, EV/S 7x premium sustained, $42B revenue$42B7.0x$1,88635%
BaseSupercycle +1 year, normalization from FY28, $32B revenue$32B5.0x$1,03845%
BearPeak recognized, FY28 GM in 50s, $22B revenue$22B3.5x$51320%

Probability-weighted 12M PT: 0.35×$1,886 + 0.45×$1,038 + 0.20×$513 = $1,830 → rounded to $1,800 (vs $1,562 current, +15%).

12. ₩1,000,000 Investment Scenarios

ScenarioAfter 1 Year
🟢 Bull (+21%)~₩1,210,000
🟡 Base (-34%)~₩660,000
🔴 Bear (-67%)~₩330,000

⚠️ Base/Bear scenarios both imply meaningful losses, entry price is the single largest variable. At $1,562, the price sits in territory that could fall 34% to the Base PT of $1,038.

13. Technical Context

ItemValue
Close$1,562.34
52-Week High$1,564.00 (set today)
1Y Return+3,685.7%
6M / 3M / 1M / 1W+621.6% / +161.3% / +83.5% / +31.6%
MA20 / MA50 / MA200$1,087 / $838 / $387
Current vs MA200+303.9% ⚠️ extreme stretch
Volume Ratio1.13x (normal range)
Golden CrossMA50 > MA200 (strong uptrend)

Volume at 1.13x indicates gradual absorption rather than blow-off-top frenzy, additional upside remains plausible. Near-term support $1,400 / $1,250; resistance $1,600 / $1,800 (Cantor PT).

14. Monitoring KPIs (Q4 FY26, 8/13)

PriorityKPIQ3 CurrentBull ThresholdBear Threshold
🥇Q4 Revenue actualGuide mid $8.0B$8.5B+<$7.5B
🥈Q4 Non-GAAP GM78.4%81%+<77%
🥉New NBM signings5 deals / $42B++3 deals / +$15B0
4Stargate Q4 rev recognition$0$400M+<$100M
5Bit shipments QoQ-high teensflat~+5%-20%

→ At Q4 print (8/13), 3+ 🟢 readings raise PT to $2,000, 3+ 🔴 readings cut PT to $1,200.

15. Recommendation Summary

SanDisk (NASDAQ: SNDK), Investment Summary

Rating
HOLD (Tactical) / BUY (12M)
12M Price Target
$1,800
Implied Upside
+15.2%
Confidence
Medium
Entry Zone
$1,300 – $1,400 (chasing not recommended)
Stop Loss
$1,200 (-23%)
Position Size
Aggressive 5–10%, Neutral 3–5% (not recommended for conservative; beta 5.04)
Time Horizon
6–12 months
Key Watch
Q4 Revenue / GM / NBMs (2026-08-13)

16. Korean Investor Perspective

ItemComment
FXKRW 1,450.8/USD stable, low FX impact
FX HedgingUnnecessary for medium-term holdings
Capital Gains Tax22% above ₩2.5M annual exemption; YTD +558% suggests proactive realization
Toss Securities$1,562 = ₩2.26M per share, fractional shares + staggered entry are essential
Risk ManagementBeta 5.04, 5x KOSPI volatility. Cap portfolio weight at 5–10%

KRW-based 12M expected return: ~+12–15% (probability-weighted).

Conclusion

SanDisk's +3,685% one-year run is built on two pillars: an AI infrastructure NAND supercycle and NBMs structurally dampening memory cyclicality. Fundamentals, guidance, and NBM lock-ins are all real, but the combination of MA200 +304% stretch + beta 5.04 + a record-high 78.4% GM makes short-term chasing unattractive. We rate the name 12M BUY with staggered entry in $1,300–1,400 as the right approach. If Q4 (8/13) delivers more NBM signings + Stargate revenue + 80% GM, we'll consider raising PT to $2,000.

For a plain-English version, see our SanDisk +3,685% beginner guide. For comparison with peers that surged or crashed in the same earnings season, DigitalOcean +40% / Datadog +31% / Cloudflare -24% / ServiceNow -17%, the AI capex cycle's winner/loser logic becomes much clearer.

This report is for market analysis purposes and does not constitute investment advice. SNDK is a 1.2-year-old listing with limited time-series data; NBMs and HBF are new models, and actual results may diverge meaningfully from our assumptions. All investment decisions remain the responsibility of the reader.

FAQ

Why is SanDisk up 3,685% in a year?

An AI infrastructure NAND supercycle plus a differentiated business model. (1) Data centers became NAND's largest demand segment in 2026, overtaking mobile for the first time in 15 years. (2) Q3 revenue +251% YoY with 78.4% GM, the strongest in memory history. (3) Five NBM contracts cumulatively, $42B locked in plus $11B in financial guarantees. (4) Forward EV/Sales of 5.15x vs the peer average of 3.06x, a +47% premium that the NBM model attempts to justify.

What are NBMs and why do they matter?

NBMs are SanDisk's new multi-year supply contracts: minimum revenue commitments + financial guarantees + prepayments. Five signed through Q3 lock in $42B+ in revenue, $11B+ in guarantees, $400M in prepayments received, with one-third of FY27 bit supply already committed. They structurally remove memory's biggest weakness, lack of ASP visibility.

Should I buy at $1,562?

Chasing here is not recommended. Beta 5.04 (5x KOSPI volatility), +304% above the 200-day MA, YTD +558%, short-term overheated. RSI is stretched after a 9-session +46.8% run. We recommend staggered buying in the $1,300–1,400 zone. If Q4 earnings (2026-08-13) deliver $8.5B+ revenue, 81%+ GM, and 3+ new NBM signings, we'd raise PT to $2,000; misses would push us toward $1,200.

Has the NAND cycle truly ended?

NBMs are the swing variable. Bull (45%): NBMs end the cycle, no normalization. Base (35%): supercycle extends one year, normalization from FY28. Bear (20%): peak recognition, GM reverts to the 50s. If Samsung and SK Hynix resume CapEx in 2027, supply normalization is a real risk. Their July 2026 Q2 guidance is the decisive signal.