SanDisk Stock Down 45%; SNDK Valuation, Earnings, CXMT, and NAND Risks Explained

SanDisk stock (SNDK) fell 24% in three trading sessions, wiping out almost a month of gains just eight trading days before its August 5 earnings report. The headlines all point to the same trigger: Chinese memory maker CXMT went public in Shanghai on July 27 and surged 466% on its debut. There is just one problem with that story. CXMT makes DRAM. SanDisk does not sell a single chip of DRAM. It is a pure-play NAND flash company, and yet its stock fell nearly as hard as Micron's and SK Hynix's, the two companies that actually compete with CXMT.

Let's break down why a NAND company got dragged into a DRAM selloff, what SanDisk's own numbers say about whether Wall Street's price targets make sense, and what happens to SanDisk stock's bull case if the company's own Q4 guidance, already calling for its largest quarter in history, turns out not to be large enough for a market that has priced in even more.

Why SanDisk Stock (SNDK) Fell 24% in Three Trading Sessions

SanDisk lost roughly 24% of its value in three sessions, and sits about 45%-46% below its June 22 all-time high of $2,354.39.

Even after this massive drop, SanDisk is still up nearly 3,000% in twelve months, having started independent life at a $48.60 close on February 24, 2025, after its spin-off from Western Digital (WDC). A stock that has run 30 times over does not need a bad reason to fall 24%. It just needs investors to remember that gravity still applies. The CXMT IPO was the trigger. Whether it was the real reason is a separate question.

Think of SanDisk's business right now in two buckets, the way you might think of your own investments. One bucket is locked in, like a fixed deposit: a fixed rate, guaranteed for a set term, protected even if markets turn. The other bucket rides the market, like an equity SIP: it benefits fully when NAND prices rise, and it is exposed fully when they fall. The entire investment debate on SanDisk right now is about how big each bucket actually is. We will size that precisely further down.

SanDisk Q4 FY2026 Earnings Preview: Revenue, EPS and Guidance

SanDisk's fiscal third quarter, reported April 30, was genuinely extraordinary. Revenue came in at $5.95 billion, up 97% from the prior quarter and 251% year over year, blowing past the company's own guidance of $4.4 billion to $4.8 billion. Non-GAAP gross margin hit 78.4%, versus a guided 65 to 67%. Non-GAAP EPS was $23.41, against guidance of roughly $12 to $14.

Notice that Edge, not Datacenter, is the largest segment by dollars, and it grew 295% YoY too. The popular narrative that this is purely an "AI datacenter demand" story is incomplete. Broad-based NAND pricing gains lifted every channel, not just the servers going into AI clusters. That is a company-reported fact, not an interpretation, and it is worth knowing before you accept any single-factor explanation for why the stock has moved the way it has.

Now for the harder part. For fiscal Q4, SanDisk guided revenue of $7.75 billion to $8.25 billion and non-GAAP EPS of $30 to $33. In percentage terms, that is actually a slower sequential pace, 30 to 39%, versus Q3's 97% jump. In dollar terms it would still be the largest quarter in company history by a wide margin. Here is what makes clearing it harder than the guidance alone suggests: according to TipRanks, Wall Street's own consensus already sits at roughly $8.42 billion in revenue and $34.67 in EPS, both above the top end of the company's own guidance range. We ran the numbers on that gap ourselves.

This is not how it usually works. Consensus typically sits inside or near the low end of a company's own guidance range, especially right after a blowout quarter, because guidance is meant to be conservative. Here, the Street has already priced in a beat on top of guidance that was itself a record. 

Hitting the August 5 numbers will not be enough on its own. SanDisk has to beat a number that already assumes it beats its own forecast, a structurally harder bar than the one it cleared in April, and a specific, checkable reason the stock could react badly even to genuinely strong results.

SanDisk Stock Price Targets: Why Analysts Range From $1,620 to $3,100

Rarely does sell-side coverage disagree this violently on a single name. Within the same week, three separate banks reset their targets, sometimes twice.

Source: TipRanks, Benzinga

The average 12-month target lands anywhere from about $1,810 to $2,220, and one aggregator tracking 29 analysts puts the median at $2,500. Majority analysts rate it a Buy. Only Wells Fargo, among the major shops, holds a non-Buy rating despite raising its target price on SNDK stock. 

That is not a stock where the Street disagrees at the margins. Wells Fargo's own target implies the stock could be roughly fairly valued near current levels, while Evercore's implies it could nearly triple. Both cannot be right, and that gap is a more honest signal of genuine uncertainty than any single average.

This is a description of how wide the range of informed opinion actually is right now, which itself tells you something about how much this trades on assumptions rather than settled facts.

SanDisk NBM Contracts: How Much FY2027 NAND Revenue Is Protected?

Back to the FD-versus-SIP framing from the opening. SanDisk has spent the past two quarters signing what it calls New Business Models, or NBMs: multi-year supply agreements backed by real financial guarantees rather than the old-style take-or-pay contracts that memory companies have used and broken for two decades. 

As of the April 30 fiscal Q3 report, five such deals were signed, backed by more than $11 billion in enforceable financial guarantees, and the three contracts signed in the third quarter alone carry a minimum $41.6 billion in remaining performance obligations, the term for contractually committed future revenue. Management says these NBMs will cover just over a third of fiscal 2027's total bit output. Run the fixed-deposit analogy against that number.

• Roughly a third of next year's volume sits in the FD bucket: locked in, guaranteed, insulated from a NAND price crash the way a bank deposit is insulated from a stock market crash. 
• The other two-thirds, by the company's own disclosure, is still riding the open market, exposed the way an equity SIP is exposed. 
• A third locked in is a genuine improvement over SanDisk's history as a pure commodity producer with, in Morningstar's words, no economic moat. It is also, mathematically, not a majority. 

Two-thirds of the business you would be buying still depends on NAND prices behaving the way this cycle's boosters expect. The company's own cash position adds a separate layer of protection, not tied to pricing at all. 

As of the April quarter, SanDisk held $3.74 billion in cash, had fully repaid its $2 billion term loan, and authorized a $6 billion buyback. That is a genuinely stronger balance sheet than the SanDisk that limped through 2023 and 2024 with negative free cash flow. It buys time. It does not, by itself, control NAND pricing.

Is SNDK Stock Cheap? SanDisk vs Micron, SK Hynix and Samsung

SanDisk’s NBM transition is only two quarters old, and management has not provided a multi-year free cash flow forecast. The company will also not outline its detailed fiscal 2027 outlook until its Investor Day on August 13. Until then, the most practical way to assess SanDisk’s near-term earnings trajectory is to add its actual non-GAAP EPS for the first three quarters of fiscal 2026 and model different outcomes for Q4.

Across that entire range, once Q4 actually reports, SanDisk's trailing multiple compresses to somewhere in the high teens, a long way from a bubble reading, even after a 3,000% run. Push a step further using analyst fiscal 2027 estimates and the current price implies a forward multiple of roughly 5.7 to 6.0 times fiscal 2027 earnings. 

For comparison, Micron trades near 10x forward earnings, SK Hynix near 5.2x, and Samsung's blended forward multiple sits around 5 to 7x, per a March 2026 industry note.

What this shows: on a forward-earnings basis, SanDisk is not obviously more expensive than its DRAM-exposed peers, and by some of these estimates it is cheaper. What this does not show: whether those FY27 estimates will actually hold up. They assume the NAND upcycle will behave differently from previous cycles. The 2018 and 2022–2023 upcycles lasted four to seven quarters before being followed by downturns of four to eight quarters.

During those corrections, industry stock prices fell 50% to 60%, while SK Hynix’s net margin dropped to roughly negative 28% at the trough of the latest cycle. A cheap-looking forward multiple is only cheap if the forward earnings actually arrive. That is an assumption, not a fact, and it is the single biggest variable in this entire thesis.

Five Risks That Could Derail the SanDisk Stock Bull Case

CXMT vs SanDisk: Why a DRAM IPO Hit a NAND Stock

Here is the part most coverage has skipped. Investors sold SanDisk, a NAND-only business, almost as hard as they sold Micron and SK Hynix, both of which face CXMT head-on in DRAM. Several outlets covering the CXMT IPO explicitly flagged this as looking like an overreaction on SanDisk specifically, since the direct read-through from a DRAM IPO to a NAND producer's earnings is genuinely indirect. 

That gap between what actually happened (a DRAM company's IPO) and how a NAND company's stock reacted (nearly identically to its DRAM-exposed peers) is the most useful, checkable thing in this entire story. It suggests part of this week's move was a basket trade, chip and memory names sold as one block, rather than a repricing of SanDisk's specific fundamentals. 

Baskets unwind. Fundamentals take longer to change. Whether Tuesday's continued weakness is the market correcting that overreaction, or the start of a slower, separate repricing of the NAND-specific risks in the table above, is exactly what August 5 and August 13 should start to clarify.

SanDisk Stock Analysis: What Investors Should Watch Next

SanDisk went from a commodity NAND maker with no pricing power and a broken balance sheet in 2024, to one of the best-performing large-cap stocks in market history in under 18 months, on the back of a genuine, well-documented NAND supply shortage and a real (if partial) shift toward contracted revenue. 

This week's drop looks driven more by a sector-wide, partly misattributed reaction to a DRAM IPO and by ordinary pre-earnings de-risking than by any new fact about SanDisk's own business.

The bull case: A debt-free balance sheet, $41.6 billion in contractually guaranteed revenue, a trailing multiple that compresses into the high teens once Q4 actually reports, and a forward multiple that looks cheap next to DRAM-exposed peers if FY2027 estimates hold.

The bear case: Two-thirds of next year's volume still has no contractual floor, Wall Street's own consensus already sits above the ceiling of the company's own record guidance, the last two NAND cycles each ended in 50 to 60% industry-wide drawdowns, and NAND pricing growth is already decelerating in real time according to TrendForce's own numbers, right as SanDisk needs it to accelerate.

The debate to track going into August 5 and the August 13 Investor Day is not whether SanDisk beats its guidance. History says it usually does. It is whether SanDisk beats a Street number that has already priced in a beat, and whether management can show that meaningfully more than a third of fiscal 2027 volume is protected the way this week's selloff implicitly assumed it was not.

None of the above amounts to investment advice or a buy, hold, or sell call on SanDisk stock. Treat it as a breakdown of the verifiable numbers behind a genuinely noisy week, so that whichever way you decide, you are working from the same data the professionals are looking at.