# Sandisk (SNDK) — InvestMoat Analysis

_Last analyzed: October 30, 2026_
_Asset class: equity · Canonical page: https://www.investmoat.com/stocks/sndk_

## Scores

| Dimension | Score (0–100) |
| --- | --- |
| Moat durability | 70 |
| Growth trajectory | 69 |
| Valuation | 64 |
| **Composite** | **67** |

Scores are computed deterministically from this asset’s data by the InvestMoat formula (see https://www.investmoat.com/llms.txt for methodology). Scores are not directly comparable across asset classes.

## Key stats

- **Ticker:** SNDK
- **Market Cap:** ~$181B

## Moat

One of the five NAND producers at scale, sharing wafer output with Kioxia through a joint venture extended to December 2034, with a growing book of take-or-pay supply contracts. The contracts are the moat news; the product itself is still a commodity bit.

### The Contracted Commodity

Sandisk's edge is **scale in a five-player NAND market plus multi-year contracts that turn spot exposure into committed volume**, not a product competitors cannot make:

- **New Business Model Agreements:** Eight customers have signed multi-year NBM agreements with committed volumes and floor-and-ceiling pricing. Management puts the minimum expected revenue at $93.9B at floor prices, backed by $16.5B of cash deposits and financial guarantees, covering about half of FY27 bits and two-thirds of FY28 bits. That is more than four years of FY26 revenue under contract, the same structure that moved Micron's transaction embedding to strong.
- **Kioxia Joint Venture Scale:** Sandisk and Kioxia share the Yokkaichi and Kitakami fabs and split the wafer output, and the JV was extended through December 2034. The JV is how a company with no fabs of its own sits among the scale producers, and it is why bit growth in the mid-to-high teens comes from BiCS node transitions rather than new clean rooms. Samsung, SK hynix with Solidigm, and Micron run the same playbook at similar or larger scale.
- **Datacenter Mix Shift:** Datacenter revenue went from $213M in Q4 FY25 to $2.98B in Q4 FY26 as enterprise SSDs qualified at hyperscalers for AI storage. Qualification is real friction, but the same sockets are dual- or triple-sourced against Samsung, Solidigm, Kioxia and Micron, so it rates as embedding, not a lock.

**Moat verdict:** Sandisk's moat is contracts on a commodity: transaction embedding is strong on $93.9B of NBM minimum revenue and $16.5B of guarantees, while process IP, scale through the Kioxia JV, talent and the retail brand are intact. None of it is AI-vulnerable software, and AI storage demand is the reason the contracts exist; the AI-era risk is the capex cycle, which sits in growth, not here.

### Top competitors

- **[Samsung Electronics (005930.KS)](https://www.investmoat.com/stocks/samsung):** The largest NAND producer.
- **[SK hynix (000660.KS)](https://www.investmoat.com/stocks/skhynix):** NAND and enterprise SSDs with Solidigm.
- **[Micron (MU)](https://www.investmoat.com/stocks/micron):** NAND and data-center SSDs alongside DRAM.
- **Kioxia (285A.T):** Joint-venture partner and NAND rival.

## Growth

Fiscal Q1 2027 (quarter to early October, reported October 29 after the close): revenue $11.03B, +23% QoQ and +378% YoY, above the $10.3–10.8B guide and ~$10.7B consensus. GAAP gross margin was 85.0%, non-GAAP EPS $49.51 (GAAP $52.60) against a $44–46 guide, and free cash flow $6.31B. Datacenter grew 61% QoQ to $4.80B; Edge rose 7% to $5.83B and Consumer fell 29% to $397M as bits went to higher-priced channels. Sandisk returned $5.0B through buybacks (~3.2M shares) and announced an eight-for-one split effective November 19, 2026. Fiscal Q2 is guided to revenue of $12.2–12.6B, gross margin 84.4–85.4% and non-GAAP EPS $55–57 (pre-split).

- **Revenue CAGR estimate:** 12–20%
- **Primary type:** both
- **Margin trend:** stable
- **Key risk (high):** NAND pricing peaks in 2027 as Samsung, SK hynix, Kioxia and YMTC add bits and hyperscalers digest AI storage, so the half of FY2027 bits outside the NBM book reprices down. Falsifiable: gross margin below 50% for two consecutive quarters, or FY2028 non-GAAP EPS below $60 (FY2026 was $70.88), would mean the contracts cushioned the cycle rather than rewrote it. Western Digital's October 22 warning on rising input costs is the first sign to watch.
- **Drivers:**
  - Datacenter (enterprise SSD) — Q1 FY27 $4.80B, +61% QoQ (Q4 $2.98B); now 44% of revenue (accelerating)
  - Edge (PC, mobile, auto) — Q1 FY27 $5.83B, +7% QoQ after +48% in Q4; mostly price (decelerating)
  - Consumer (cards, drives, retail) — Q1 FY27 $397M, −29% QoQ as bits were allocated to higher-priced channels (decelerating)
- **Score derivation:** Base 80.7 (12–20% CAGR, midpoint 16%, set off the cycle per G5) − 1.3 trajectory (Datacenter accelerating; Edge and Consumer decelerating) + 0 stable margin (85% gross margin is at the peak, not expanding) − 10 high key risk = 69. The October 30 refresh moves Edge to decelerating (+48% QoQ in Q4 → +7% in Q1).

## Valuation

Re-marked at the $1,330.05 close on October 29, 2026, before the fiscal Q1 print released after that close: 64 on the held $700 / $1,300 / $2,000 ladder (pre-split; the eight-for-one split is effective November 19), from 68 at the $1238.03 mark. The ladder, struck on October 26, is held: the quarter beat and the Q2 guide (EPS $55–57) runs above it, but the band is set off the cycle and the bear already prices a 2027 roll-over. The text below was written at the earlier mark. At $1,238.03 (October 23 close, ~$181B market cap) Sandisk is ~47% below its $2,354 52-week high and still up more than 400% this year. On the Q1 FY2027 EPS guide of $44–46 the run-rate multiple is ~7×; on FY2026's $70.88 it is ~17×. The ladder uses Micron's method so the two memory names can be read side by side: the base is ~12–13× a mid-cycle EPS near $100–105 (about 0.6× the guided run-rate, the same discount Micron's base takes), the bear is ~6–7× that mid-cycle figure with the NBM floor under it, and the bull is ~11× the run-rate. Spot is ~90% of the way from the $700 bear to the $1,300 base, for a valuation score of 68. The base is ~0.61× the ~$2,120 Street mean, the same far-below-Street shape as Micron (see the valuation-floor open call).

| Multiple | Value | Note |
| --- | --- | --- |
| Trailing P/E (FY26 non-GAAP) | ~17× | FY2026 non-GAAP EPS $70.88; GAAP $73.76 |
| Run-rate P/E (Q1 FY27 guide) | ~7× | $44–46 non-GAAP EPS guided for one quarter, ~155M diluted shares |
| Price / Sales (FY27 consensus) | ~4.3× | ~$181B / ~$42B FY2027 consensus revenue |
| FCF | $11.5B FY26 | Q4 FCF $7.08B; no debt; $15.5B buyback authorisation remaining |

Memory at peak earnings trades at single-digit multiples, and Sandisk is there on the run-rate. The question the price asks is how far FY2028–29 earnings fall from a $180 annualised run-rate. The NBM floor pricing on half to two-thirds of bits is why the bear sits at $700 rather than at a full-cycle trough; the uncontracted half is why the base does not capitalise the run-rate. _(as of October 23, 2026 (price $1,238.03))_

## Price scenarios

### Bear — $700

NAND pricing rolls over in 2027 as supply catches up, the uncontracted half of bits reprices, and the stock trades at ~6–7× a mid-cycle EPS with the NBM floor holding the rest.

- Q1 FY27 lands at the low end of $10.3–10.8B and Q2 guides gross margin down as contract-price increases stall
- Samsung, SK hynix and YMTC bit additions push enterprise SSD prices down through 2027
- FY2028 consensus EPS falls toward $60, with NBM floors and guarantees preventing a deeper trough

### Base — $1,300

FY2027 revenue near the ~$42B consensus, a digestion year in FY2028 cushioned by the NBM book, and the market capitalising a mid-cycle EPS of ~$100–105 at ~12–13×.

- Q1 FY27 delivers $10.3–10.8B and $44–46 EPS, and the NBM book keeps growing past eight customers
- Datacenter keeps gaining share of revenue as enterprise SSD qualifications broaden
- Gross margin settles well above prior-cycle peaks as contracted volume reduces spot exposure

### Bull — $2,000

AI storage demand keeps NAND tight into 2028, more of the book moves under NBM contracts at higher floors, and the market pays ~11× the guided run-rate.

- NBM coverage rises above two-thirds of FY28 bits at higher floor prices
- FY2028 EPS holds near the FY2027 run-rate instead of digesting
- Buybacks from the $15.5B authorisation shrink the share count materially

---

InvestMoat is a research and education framework. Nothing here is financial advice. Past performance does not guarantee future results.