FY2025 (Jun 2025)
Period
- Revenue
- $7.4B
- Gross Profit
- $2.2B
- Gross Margin %
- 30.1%
- Op. Income
- -$1.4B
- Net Income
- -$1.6B
- Inventory
- $2.1B
How exposed is Sandisk to NAND contract pricing, and what does the latest 10-Q actually say about it?
This is a rich picture. Let me put the financials together and then break down the filing's actual language.
| FY2025 (Jun 2025) | $7.4B | $2.2B | 30.1% | -$1.4B | -$1.6B | $2.1B |
| Q1 FY2026 (Oct 2025) | $2.3B | $687.0M | 29.8% | $176.0M | $112.0M | $1.9B |
| Q2 FY2026 (Jan 2026) | $3.0B | $1.5B | 50.9% | $1.1B | $803.0M | $2.0B |
| Q3 FY2026 (Apr 2026) | $6.0B | $4.7B | 78.4% | $4.1B | $3.6B | $2.2B |
| Q4 FY2026 (Jul 2026) | $9.0B | $7.6B | 84.6% | $7.0B | $6.9B | $2.7B |
FY2025 (Jun 2025)
Period
Q1 FY2026 (Oct 2025)
Period
Q2 FY2026 (Jan 2026)
Period
Summary: Sandisk is among the most purely NAND-exposed large-cap equities in existence — its entire revenue base is NAND flash, cost structure is largely fixed via Flash Ventures, and the resulting gross margin has swung from ~30% to ~85% in four quarters purely on ASP movement.
Sandisk's cost base is quasi-fixed by design. The 10-Q is explicit: the company is obligated to pay half of Flash Ventures' fixed costs regardless of output purchased, and its three-month purchase commitments are binding and non-cancelable. This is classic operating leverage in a commodity business: costs don't move much, so every dollar of ASP swing falls almost entirely to the gross profit line.
The filing acknowledges this directly — it cannot estimate its total purchase commitment beyond three months because the price is determined by future production costs, not a locked contract rate.
The table above makes the pricing sensitivity undeniable:
The most interesting structural development is the dramatic jump in contract liabilities — from $25M at June 2025 to $511M at April 2026, driven by customer advances on long-term agreements. Furthermore, the company discloses $41.6B in remaining performance obligations from long-term agreements, with only ~15% expected in the next twelve months.
This $41.6B backlog is striking — it's roughly 2x trailing annual revenue — and suggests significant AI-driven enterprise customers are locking in supply. The 10-Q notes: "AI adoption is driving the need for NAND storage to support these workloads, leading to increased revenues."
However, the filing explicitly flags that these agreements cut both ways — if Sandisk fails to deliver on volume or spec, customers can reduce pricing commitments or terminate. And if a customer breaches, Sandisk may be unable to re-sell inventory at comparable prices.
The filing also discloses that price protection programs fell from 19% to 12% of gross revenue year-over-year. This is a secondary confirmation that pricing pressure has reversed — when the market tightens, Sandisk needs to offer fewer price concessions.
Sandisk has zero structural hedge against NAND pricing: costs are fixed (Flash Ventures obligations), revenue is 100% NAND, and there is no commodity hedging program disclosed. The $41.6B long-term backlog partially insulates the top line, but the terms are volume-and-spec-conditional, not price-locked. The business is currently operating at near-peak margin — 84.6% gross in Q4 FY2026 — which should be interpreted as the ceiling, not the floor. The FY2025 comps (-$1.6B net loss on a 30% gross margin) show exactly how fast that unwinds when NAND spot turns.
The single most important forward indicator to watch: whether the long-term agreement pricing is fixed or floating. The filing's silence on this point is telling.
Sources: SEC Form 10-Q (SNDK, filed 2026-05-01, period ending 2026-04-03); Annual and quarterly financial reports (SEC filings).
Figures as of the run date (Aug 11, 2026) · sources cited inline · information, not investment advice