---
title: "Sandisk's record quarter crashed anyway: the flash pricing engine that drove it just stalled"
description: "Sandisk beat its own guidance and still lost 18% in two sessions. Kioxia's flat December guide and Morgan Stanley's through-cycle model mark the turn: the flash price surge did the work last quarter, and it is stopping."
publisher: "The Inference"
section: "Chips"
published: 2026-11-02T15:38:21.745Z
modified: 2026-11-02T15:38:21.745Z
canonical: https://theinference.org/article/sandisk-s-record-quarter-crashed-anyway-the-flash-pricing-engine-that-drove-it-just-stalled
language: en
keywords: "Sandisk, Flash Memory, NAND, Kioxia, Morgan Stanley"
---

# Sandisk's record quarter crashed anyway: the flash pricing engine that drove it just stalled

> Sandisk beat its own guidance and still lost 18% in two sessions. Kioxia's flat December guide and Morgan Stanley's through-cycle model mark the turn: the flash price surge did the work last quarter, and it is stopping.

[Sandisk](https://theinference.org/markets/companies/sandisk) sold a record $11.03 billion of flash in its September quarter and [beat its own guidance](https://finance.yahoo.com/markets/stocks/articles/sandisk-sinks-13-last-friday-161406084.html) [1]. The stock still lost about 18% across two sessions [2]. What cracked was not the quarter; it was the engine underneath it, rising flash prices, the thing investors were actually buying. Kioxia, the Tokyo maker that fabs Sandisk's chips, put numbers on the stall on Monday and shed almost 16% of its own value saying so [3][4].

*Chart: **Sandisk's quarterly revenue went vertical over the past year** Line chart of Sandisk's quarterly revenue in billions of dollars across ten quarters, rising from $1.76 billion in the June 2024 quarter to $11.03 billion in the September 2026 quarter, with the steepest climb in the final three quarters.*

*Sandisk quarterly revenue in USD billions, from its SEC filings; labels are calendar quarters, the last being fiscal Q1 2026, ended October 2, 2026. [7]*

## Price did the work, and volume did not take the load

Sandisk's boom was mostly price. Management put pricing at about two-thirds of the June quarter's sequential growth [1]. By September, CFO Luis Visoso put pricing at roughly 40% of a much smaller gain and guided only modest increases from here [1].

Volume did not take the slack. Consumer revenue fell 29% as buyers balked at higher prices [1], and Edge, the phone-and-PC unit that is Sandisk's largest, shipped 20% less flash by volume than a year earlier while still collecting more money [2]. The selling ran through the memory chain with it: [Micron](https://theinference.org/markets/companies/micron) slid alongside, and [Intel](https://theinference.org/markets/companies/intel)'s weak PC guide blamed memory costs [1].

*Chart: **Sandisk's sequential growth halved in one quarter, with more slowing guided** Column chart of Sandisk's sequential revenue growth: 51 percent in the June quarter, 23 percent in the September quarter, and about 12 percent guided for the December quarter at the midpoint of the $12.2-12.6 billion range, with the December bar drawn as guidance.*

*Sandisk's sequential revenue growth in percent, as reported for the June and September quarters and as guided for the December quarter at the midpoint of the $12.2-12.6 billion range; the December figure is company guidance. [1]*

## Tokyo put the turn in writing

[Kioxia](https://theinference.org/markets/companies/kioxia-holdings-corporation) [guided December-quarter operating profit essentially flat](https://www.bloomberg.com/news/articles/2026-11-02/kioxia-shares-tumble-after-outlook-suggests-demand-leveling-off) with the quarter before, well short of the ¥2.35 trillion analysts expected [3]. Its average NAND selling price rose in the mid-20% range last quarter, down from a pace near 70% the quarter before [4]. CFO Yoshihiko Kawamura calls the gap timing: some orders were pulled forward at customers' request, others slip into March because enterprise storage takes longer to build, and he expects rapid growth to return then [3].

Investors did not buy the timing story. The deeper read is that the steepest part of the flash price cycle has passed [3].

## The contract book secures volumes and freezes prices

The long-term agreements Sandisk signed while prices ran give it visibility, and they also explain why price cannot re-accelerate. Sandisk's average selling price rose only about 9% last quarter, [Morgan Stanley](https://theinference.org/markets/companies/morgan-stanley)'s Joseph Moore wrote, because it locked its contracts early while spot prices ran [2]. The book covers about half of fiscal 2027 bits and two-thirds of fiscal 2028 [1]. Those volumes are secure, but bits priced under old terms do not reprice when the spot market does.

Kioxia's 2027 output is sold out and more than half of its 2028 production is allocated, so the supply that would carry higher prices is already committed [3]. CEO [David Goeckeler](https://theinference.org/markets/people/david-goeckeler) [sees customer commitments running through 2030](https://finance.yahoo.com/markets/stocks/articles/sandisk-ceo-david-goeckeler-sees-123005695.html), with some customers offering the deals themselves [5].

Morgan Stanley's own model says the same. Moore [kept his Overweight rating and $1,750 target](https://www.thestreet.com/investing/morgan-stanley-sandisk-stock-decline) on October 30 [2]. Yet the same note values Sandisk on through-cycle earnings of $62.50 a share, a fraction of his own 2027 estimate, and shows fiscal 2029 earnings falling about 21% [2]. Even the most bullish house on the name is underwriting the cycle's return.

## December is the verdict, and it lands split-adjusted

Sandisk guided December-quarter revenue to $12.2-12.6 billion [1][5]. The gross margin guide, 84.5-85.5%, is the number to watch [2]. It asks Sandisk to hold the 85.1% it just printed while pricing adds almost nothing and its largest unit ships less volume [2][6]. Datacenter revenue grew 61% sequentially [6], but wafer lead times limit how fast capacity shifts from consumer to datacenter, the same lead times Kioxia blames [3]. If low-end weakness spreads to premium phones, datacenter has to carry everything, and the guide leaves little room.

The calendar compresses it. The record date for the 8-for-1 split is November 9, with split-adjusted trading from November 20 [6], so December gets judged on a new screen days after it opens. Friday's close sat just above Morgan Stanley's $1,100 bear case, the level options traders gave near even odds of breaking within a year [2]. Morningstar had already [cut its fair value to $870 on October 30](https://stocktwits.com/news-articles/markets/equity/sndk-dips-premarket-despite-strong-q1-analyst-sees-memory-pricing-peak-on-the-horizon/cZUhFpWR9Td) and called a pricing peak near [8]; the market has now repriced that peak faster than the research houses did.

## What ended is the pricing phase, not the AI build-out

None of this is a verdict against the AI build-out. Sandisk's customers are signing supply commitments through 2030 [5], and [SK Hynix](https://theinference.org/markets/companies/sk-hynix) and Samsung both expect demand to outrun supply into 2027 [5]. What ended is the phase where price did the work. From here Sandisk is a volume and execution story, and the December print, judged split-adjusted, is where the market finds out whether the volume shows up.

## Takeaway

Sandisk beat guidance with a record $11.03 billion September quarter, yet fell 18% because the flash price surge that drove its growth is stalling: Kioxia guided December profit flat and Sandisk's own guide asks it to hold an 85% gross margin with pricing adding almost nothing. The December print, judged split-adjusted, tells whether volume takes over from price.

## Sources

1. [Yahoo Finance / TIKR, Sandisk Sinks 13% Last Friday After a Beat: Kioxia Says the Flash Price Surge Is Fading, 1 November 2026](https://finance.yahoo.com/markets/stocks/articles/sandisk-sinks-13-last-friday-161406084.html)
2. [TheStreet, Morgan Stanley stands by Sandisk as CFO flags a low-end pullback, 2 November 2026](https://www.thestreet.com/investing/morgan-stanley-sandisk-stock-decline)
3. [Bloomberg, Kioxia Shares Tumble After Outlook Suggests Demand Leveling Off, 2 November 2026](https://www.bloomberg.com/news/articles/2026-11-02/kioxia-shares-tumble-after-outlook-suggests-demand-leveling-off)
4. [ShiftDelete.net, Kioxia Stock Drops 16% as Flash Memory Profit Growth Stalls, 2 November 2026](https://shiftdelete.net/en/kioxia-stock)
5. [Yahoo Finance / Benzinga, SanDisk CEO David Goeckeler Sees 'Tremendous Pull' as Customers Lock in Supply and Come Back for More, 1 November 2026](https://finance.yahoo.com/markets/stocks/articles/sandisk-ceo-david-goeckeler-sees-123005695.html)
6. [FinanceFeeds, Sandisk Stock Split: 8-for-1 Effective 19 November, Split-Adjusted Trading From 20 November, 30 October 2026](https://financefeeds.com/sandisk-stock-split-8-for-1-19-november/)
7. [Sandisk, quarterly revenue from its SEC filings, retrieved 2 November 2026](https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0002023554)
8. [Stocktwits, SNDK Dips Premarket Despite Strong Q1: Analyst Sees Memory Pricing Peak On The Horizon, 30 October 2026](https://stocktwits.com/news-articles/markets/equity/sndk-dips-premarket-despite-strong-q1-analyst-sees-memory-pricing-peak-on-the-horizon/cZUhFpWR9Td)


---

Sandisk's record quarter crashed anyway: the flash pricing engine that drove it just stalled — The Inference. Canonical: https://theinference.org/article/sandisk-s-record-quarter-crashed-anyway-the-flash-pricing-engine-that-drove-it-just-stalled
