SanDisk Corporation (NASDAQ: SNDK) — Profile
Status: HOLD post-Q3 FY26. $1,187 share price (May 2 2026), $176B market cap, fwd P/E 7.1× — a 35× move from $33 52-week low. 78% non-GAAP gross margin, 71% op margin, $6B buyback authorized. Five NBM agreements signed locking $42B+ minimum revenue from 3 contracts + $11B+ aggregate financial guarantees + >⅓ of FY27 bits. Central debate is whether NBMs prove structural NAND durability or just lock in a peak. Stock fell on Q3 FY26 print despite the numbers — durability skepticism is the bear case.
Quick Snapshot (2026-05-03)
| Metric | Value |
|---|---|
| Price | $1,187 |
| Market cap | $175.7B |
| EV | $171.9B |
| 52w range | $33.13 – $1,189.24 (at 52w high) |
| Shares out | 148M |
| Trailing P/E | 40.5× |
| Forward P/E | 7.1× (reflects massive FY27 earnings ramp embedded in consensus) |
| P/Sales (TTM) | 13.3× |
| P/Book | 17.2× |
| Op margin (TTM, GAAP) | 70.0% |
| Profit margin (TTM) | 34.2% |
| Revenue growth (TTM) | +251% |
| Total cash | $3.74B |
| Total debt | $0 (net cash position post Q3 FY26) |
| Analyst PT mean / high / low | $1,206 / $1,800 / $650 |
| Analyst count | 21 |
| Recommendation | Buy (1.62) |
| Industry | Computer Hardware |
| Employees | 11,000 |
| Next earnings | ~late July 2026 (Q4 FY26) |
Corporate Overview
Spun off from Western Digital (WDC) on February 21, 2025, becoming a pure-play NAND flash company. The brand traces back to the original SanDisk founded 1988 by Eli Harari (invented system flash technology), acquired by WDC for $19B in 2016, then spun back out in 2025. The 2025 spin separated WDC's HDD business (which retained the WDC ticker) from the NAND/SSD business (now SNDK).
Core asset: 50/50 BiCS NAND fabrication JV with Kioxia (TSE: 285A) at the Yokkaichi (Mie) Fab — largest NAND fab in the world — and the Kitakami (Iwate) Fab. JV extended through December 2034 (announced January 30, 2026); SanDisk pays Kioxia $1.165B in installments through 2029 to extend the JV:
| Date | Payment to Kioxia |
|---|---|
| 2026-04-15 | $175M |
| 2026-12-01 | $200M |
| 2027-12-01 | $230M |
| 2028-12-01 | $260M |
| 2029-12-01 | $300M |
Means SanDisk has guaranteed wafer access for the next 8+ years, paid for in advance — a major source of structural visibility on the supply side.
Business segments (Q3 FY26 disclosure):
- Data Center — $1.467B (+233% sequentially), enterprise SSDs now 25% of portfolio. Calendar 2026 growth revised to mid-70s% (from 60s)
- Edge — $3.163B (+118% sequentially); client SSDs, mobile, retail
- Total Q3 FY26 revenue — $5.95B
Q3 FY26 Earnings (April 30, 2026) — The NBM Disclosure
Headline numbers (non-GAAP unless noted)
| Metric | Q3 FY26 Actual | Q4 FY26 Guide |
|---|---|---|
| Revenue | $5.95B (+97% QoQ, +251% YoY) | $7.75B–$8.25B |
| Gross margin | 78.4% | 79–81% |
| Operating margin | 70.9% | — |
| EPS | $23.41 | $30–33 |
| FCF | $2.96B (49.7% margin) | — |
| Capex | $240M (4% of revenue) | — |
| OpEx | — | $480–500M |
Bit shipments: flat YoY, down high-teens QoQ; up 18% fiscal year-to-date. Top-line growth is ASP-driven, not bit-driven — important context for the NBM discussion.
NBM agreements — the structural shift
NBM = "New Business Model" in SanDisk framing. Multi-year customer supply agreements with prepayments and walkaway penalties.
| Disclosed Element | Value |
|---|---|
| Total agreements signed | 5 (3 in Q3, 2 more in Q4 FY26) |
| Minimum contractual revenue (3 Q3 contracts only) | ~$42 billion |
| Aggregate financial guarantees (all 5) | >$11 billion |
| Prepayments on Q3 BS | $400 million |
| FY27 bits locked under NBMs | >1/3 of bit shipments |
| Maximum contract duration | 5 years |
| Pricing structure | Fixed near-term, variable longer-term (upside capture on the variable portion) |
| Walkaway treatment | Prepayment retained + financial guarantees seized; on escrow portion, recognized at 100% gross margin |
Capital allocation post-Q3
- Repaid $650M TLB balance → $0 total debt
- $6B share buyback authorized (no expiration)
- Cash position $3.74B; net cash
- $1B invested in Nanya for DRAM supply security (separate from JV)
The Durability Debate
Why this stock fell on Q3 FY26 print despite the numbers: investors are debating whether NBMs prove structural multi-year NAND demand (bull) or simply cap upside on 1/3 of bits while exposing the rest to a coming cyclical reset (bear).
See themes/memory-sector-brief Section 1 for the full bull/bear framing and 5 settling signals to watch over the next 12-18 months.
Bull case (in 30 words): Customers signing 5-year contracts with 10-30% prepayments and walkaway penalties signal that AI infrastructure buildout makes memory shortage a multi-year structural problem, not a cyclical peak.
Bear case (in 30 words): Fixed-price components on 1/3 of FY27 bits cap upside if spot keeps rising. Bit growth flat YoY suggests ASP-driven peak, not unit-driven structural growth. 78% GM is unrepeatable.
Key fact pattern: SanDisk is not alone. Samsung, SK Hynix, and Micron have also disclosed multi-year customer supply agreements with prepayments (10-30% of contract value) and walkaway penalties. This is an industry-wide commercial-model shift.
Risk Factors
- Fixed-pricing trap — locked-in NBM prices cap upside on 1/3 of FY27 bits if spot NAND continues rising. Bear case for the stock.
- NBM realization risk — $42B is "minimum contractual revenue" but GAAP recognition path is opaque given variable price components and 5-year duration. Bears worry the headline number is theatrical.
- Bit-growth deceleration — Q3 FY26 bit shipments flat YoY, down high-teens QoQ. Top-line growth is ASP-driven; if NBM fixed prices anchor part of FY27 ASPs, ASP momentum decelerates.
- NAND oversupply 2027–2028 — YMTC (China) ramping 294L NAND; combined with Big Four capex ($20B+/yr at Micron alone), NAND oversupply could materialize 2H 2027.
- Customer concentration — Hyperscalers unnamed but presumed (Amazon, Google, Microsoft, possibly Meta + xAI). Counterparty credit risk on $11B+ guarantees if AI capex unwinds.
- Kioxia dependence — 50% of NAND output comes from JV with Kioxia. Any disruption to the JV relationship is existential.
Pink's Position Context
Pink holds SNDK as her primary NAND durability bet. Sizing logic:
- If durability thesis right: SNDK is the cleanest expression — purest NBM disclosure of any memory name. Add on dips.
- If durability thesis wrong: 78% GM is unsustainable; trim hard.
- Currently: active debate; hold and watch Q4 FY26 earnings (~late July 2026) + FY27 guide as the next major resolution.
Linked positions:
- 285A — JV partner; receives $1.165B from SNDK 2026-2029
- 2337 — Macronix; downstream beneficiary if NBMs accelerate Big Four eMMC exit
- DRAM ETF (Roundhill) — basket exposure to memory; SNDK is a top-5 holding
Watch List
| Date | Event | What to look for |
|---|---|---|
| ~2026-07-late | Q4 FY26 earnings + FY27 guide | First numerical anchor for FY27; NBM revenue recognition path |
| 2026-05-15 | Kioxia (285A.T) earnings | JV partner color on NAND demand environment |
| 2026-Q3 | Spot NAND price trajectory | If spot plateaus while NBMs hold = bull confirmed |
| 2026-12 | HBM4 contract finalizations across the complex | Read-through to NAND durability |
Cross-References
- Memory sector brief (full durability debate): themes/memory-sector-brief
- Memory industry primer: memory-industry-primer
- Kioxia (JV partner): 285A
- Macronix (read-through): 2337/2337.md
- Source — Q3 FY26 transcript: Motley Fool
- Source — Q3 FY26 results: Business Wire
- Source — stf Substack note: substack.com/@stfbutnou/note/c-252436046
- Source — Seeking Alpha NBM analysis: SA article
- Source — SanDisk IR: investor.sandisk.com
Briefings
- 2026-04-26 · Memory Sector Primer: HBM, NAND, and the 2026 Up-Cycle · vault