SanDisk SNDK stock prediction: $3,000 bull vs $1,000 bear

The best-performing stock in the S&P 500 just fell 37% in a month, and the reason was not in its order book — it was in someone else’s margin account. SanDisk (SNDK) closed July at $1,214.83, down from a June 25 record of $2,335, after printing $1,016 on July 29 — the day Leopold Aschenbrenner’s Situational Awareness fund was forced to liquidate a SanDisk position worth roughly $1.1 billion in stock and calls, its second-largest disclosed bet. The stock that had gained 707% year-to-date at its peak became the largest casualty of the largest forced unwind since Archegos — and that distinction anchors this SanDisk stock prediction: $2,218 base, $3,000 bull, $1,000 bear.

Here is what the drawdown headlines bury: nothing in the NAND market broke in July. The supply shortage that carried SanDisk from $275 in January to $2,335 in June — capacity discipline after the spin-off from Western Digital, AI storage demand, and the high-bandwidth flash (HBF) roadmap — is the same market it operates in at $1,215. What changed is that the single most aggressive levered holder of the equity ceased to exist in six trading days, and his $1.1 billion position moved through the tape on its way to Citadel’s book. When the street’s price targets span $1,000 to $3,169 around a spot of $1,215, the market is not debating SanDisk’s flash cycle; it is debating how much of the crash was information and how much was inventory. On Wednesday 5 August, for the first time since the unwind, the company itself gets to weigh in.

Key facts

• SNDK closed $1,214.83 on 31 July 2026, down 5.09% (−$65.13) on the day — down 37% across July from the June 25 high of $2,335, after a July 29 low of $1,015.89, per StockAnalysis
• Q4 FY26 earnings: Wednesday 5 August 2026, with an Investor Day on 13 August. Wall Street consensus is EPS of $34.67 for the quarter against $0.29 in the year-ago quarter, per earnings-preview coverage
• Options-implied move: ~25.08% in either direction on the print, against an 8.75% average across the past four quarters, per TipRanks — a band of roughly $910 to $1,520 from Friday’s close
• Consensus 12-month target: $2,217.77 across 23 analysts rated Buy; range $1,000–$3,169, per StockAnalysis consensus data. Market cap $179.9 billion; 52-week range $40.10–$2,354.39
• The July upgrade wave preceded the crash: Bernstein lifted its target from $1,700 to $3,000 (Outperform) and Wedbush from $1,200 to $2,000 in early July
• Situational Awareness held roughly $1.1 billion of SNDK in combined equity ($724.4m) and call options per its May 18 13F — sold to Citadel in the July 30 pre-open block, per BitMEX Research
• Core Situational Awareness holdings fell 27–54% in July as the 4x-levered book unwound, per SpotGamma’s post-mortem

The 5 August print: what the options market is charging for

This is the first quarter SanDisk reports as a post-crash stock, and the options market is treating it accordingly. An implied move of roughly 25% is close to three times the 8.75% the stock has actually averaged after its last four prints — the same pattern of over-priced event risk that showed up across the AI complex this season, except here it is layered on top of a liquidation that is barely a week old. Applied to Friday’s $1,214.83 close, that band runs from about $910 — below the July 29 liquidation low — to about $1,520, back above the pre-flush shelf.

What the sell side wants from the release is narrow and specific. Goldman Sachs analyst James Schneider has said he expects a very strong quarter on continued NAND supply tightness, and has raised estimates on improving pricing. The consensus EPS of $34.67 against $0.29 a year ago is not a typo — it is what a full flash upcycle does to a business with operating leverage this violent. So the headline number is not really the question. The three things that decide the reaction are the depth of the industry undersupply management describes, whether the NAND price increases are holding into the current quarter, and any further detail on long-term agreements (LTAs) that lock pricing forward. The 13 August Investor Day sits eight days later as a second, larger catalyst — which gives management an obvious incentive to hold its best disclosure back, and gives the market an obvious reason not to fully resolve the debate on Wednesday night.

The honest risk to the bull framing is already visible in the tape: NAND prices have shown signs of slipping from the summer peak, and Chinese suppliers are expanding capacity. Neither is fatal to a multi-year undersupply thesis, but both are exactly the sort of second-derivative wobble that a stock priced for perfection cannot absorb quietly.

What actually happened in July

Rewind to the first week of July and SanDisk was the market’s momentum queen: Bernstein had just lifted its target to $3,000, Wedbush to $2,000, and the stock was leading the S&P 500 on a supply-shortage thesis with years of visibility. Three forces then hit in sequence. The July 17 Meta Compute announcement de-rated the whole AI-infrastructure complex, and SanDisk — a storage supplier priced as an AI pure-play — fell with its customer base. The de-rating detonated the leverage: Situational Awareness, holding SanDisk as its second-largest position at 4x gross leverage, absorbed what SpotGamma calculates as “a ~120% hit to equity” from a ~30% long-book decline, and by July 24 Aschenbrenner was telling investors the fund had “not been immune” to the turbulence. Margin calls from its three prime brokers followed; on July 29 SNDK flushed to $1,015.89 — within 2% of the street’s lowest target — and before the next open the entire book, calls included, was Citadel’s.

Note the calls specifically: $1.1 billion of combined exposure meant dealers who had sold the fund call options were hedged with long stock, and as the position was unwound those hedges unwound with it — a mechanical accelerant that turned selling into cascading selling. The 26% single-day recovery off the July 29 low, the moment the block cleared, is the measure of how much of the final leg was flow rather than fundamentals. That the stock then gave back 5.09% on the final session of the month is the reminder that the overhang has not finished clearing.

From spin-off orphan to supercycle flagship

SanDisk’s 2026 arc only makes sense against its February 2025 birth. Spun out of Western Digital as the flash business nobody on the sell side wanted to model, it listed into apathy — the stock spent its first months in the low hundreds (and traded as low as the $40s adjusted in its earliest sessions) while the market treated NAND as the memory industry’s perpetual problem child. Two things broke the frame. First, the industry’s post-spin capacity discipline held: with fewer independent players and no one racing to add wafers, the customary NAND oversupply never arrived. Second, AI demand landed on the storage layer with force — training clusters and inference farms consume enterprise SSDs in volumes the old smartphone-and-PC models never contemplated, and pricing inflected upward through every 2026 quarter. A stock priced for cyclical mediocrity re-rated toward structural shortage, and momentum capital — Situational Awareness the most aggressive among it — chased the re-rating from three digits to four.

The fund’s structure around the position deserves its own paragraph, because it explains July’s ferocity. Situational Awareness was not simply long SanDisk: it paired the long against put exposure on Micron inside its $8.47 billion semiconductor put book — a relative-value expression that flash pure-play beats diversified memory. In a rising tape that pair compounds beautifully. In July’s unwind it failed on both legs simultaneously: the SanDisk long collapsed with the AI complex while the put wing bled premium on names that fell less than the things the fund owned. A hedge that correlates with your longs at exactly the wrong moment is not a hedge; it is double exposure wearing a disguise.

The business the margin call could not touch

SanDisk’s 2026 is a supply-side story, and supply did not change in July. Since the February 2025 spin-off from Western Digital, the NAND industry has run with unprecedented capacity discipline, and AI has added a structurally new demand layer: training clusters consume enterprise SSDs at densities that have repriced the entire flash complex, with SanDisk’s high-bandwidth flash roadmap positioning it in the memory-adjacent tier the AI build-out is short of. That is the thesis Bernstein’s $3,000 target prices, and nothing in the July tape contradicted it — the upgrade wave’s core argument (multi-year undersupply meeting inelastic AI demand) survives the month intact. Wednesday’s guidance is where it either gets confirmed in the company’s own numbers or starts to fray.

The honest bear case is cyclical memory history: flash is a commodity with a hundred percent drawdown pedigree, +707% YTD at the peak had priced several years of the upcycle in advance, and the $1,000 low target on the street is a reminder that when NAND turns, it turns fast. Add the new overhang — Citadel holding a distressed $1.1 billion position it did not choose — and the case for patience writes itself. What the bear case cannot claim is July as evidence: a forced seller’s exit price is not a demand signal.

The numbers: price vs targets

Reference	Level	vs $1,214.83 close
Earnings band low (25.08% implied)	$910	−25%
Street-low target / July 29 low (bear line)	$1,000 / $1,015.89	−18% / −16%
Earnings band high (25.08% implied)	$1,520	+25%
Wedbush target	$2,000	+65%
Consensus (23 analysts)	$2,217.77	+83%
June 25 record	$2,335	+92%
Bernstein target (bull line)	$3,000	+147%
Street-high target	$3,169	+161%

Sources: StockAnalysis quote and consensus data (31 July 2026 close); TipRanks for the options-implied move; July upgrade coverage for individual targets. The 2026 open was $275.24, so $1,214.83 is still roughly +341% year-to-date.

The FinanceFeeds call: $2,218 base, $3,000 bull, $1,000 bear

Base case $2,218 over 12 months — the consensus, which effectively assumes the NAND upcycle runs as modelled while the stock rebuilds the multiple the unwind destroyed. Note what that implies: the street has not marked its targets down after the liquidation, so the consensus now sits 83% above spot. Either the analysts are slow, or the crash really was inventory. Bull case $3,000 — Bernstein’s number — requires Wednesday’s print to confirm flash pricing momentum and the HBF roadmap converting into named AI design wins; in that world July becomes a historic entry printed by someone else’s leverage. Bear case $1,000 — the street’s lowest target, which the July 29 tape validated to within two percent — is the invalidation line: a weekly close below it would say the NAND cycle itself is rolling over, converting a flow event into a fundamental one. The technical map between: $1,600 is the pre-flush shelf where the last uncoerced buyers sat, and repeated high-volume absorption days at these levels would be the tell that Citadel’s inventory is finding real hands.

How to read the post-block tape: the mechanics matter more than usual because of the options component. The dealers who bought back their hedges as the fund’s calls were unwound removed a structural bid from the stock; conversely, Citadel’s inherited inventory caps rallies until distributed. The practical tells are volume texture over direction. Heavy-volume days that close flat around $1,200–1,300 are absorption; a high-volume reclaim of $1,600, the pre-flush shelf, would signal the overhang cleared; light-volume drift back toward $1,000 would say real buyers are waiting for the second test of the low that post-liquidation tapes often deliver.

What would change this call

Four triggers. NAND contract pricing rolling over — the entire base case is a supply-discipline bet, and Wednesday’s commentary plus the 13 August Investor Day are the next two reads on it. A second forced holder emerging (July taught the market to ask who else owns concentrated storage books at leverage). A weekly close below $1,000, per above. And to the upside: the first HBF volume-shipment announcement with a named AI customer, which would do for SanDisk what HBM qualification did for the DRAM winners. The macro sits underneath it all — a Fed held 9–3 with three hike dissents is a headwind for every momentum multiple, this one included — and the sector-wide question of hyperscaler self-supply runs through the same complex-wide repricing covered in our IREN analysis, the other large position caught in the same unwind.

The regulatory and macro seam

Two external currents complete the picture. The Federal Reserve’s hawkish posture — a 9–3 July hold with all three dissents wanting a hike — is the discount-rate headwind against every four-digit momentum multiple in the market, and a September hike would test the base case’s assumption that SanDisk rebuilds its rating within the year. Storage also sits inside the export-control perimeter: enterprise SSD and advanced NAND flows into China remain a live policy file, and any widening of controls cuts both ways for a US-aligned supplier — lost China revenue against tighter effective global supply. Neither current changes the shortage arithmetic; both set the multiple the market will pay for it. The July episode itself may draw the more consequential regulatory eye: a single fund’s leverage moving a top-ten S&P component 22% below every analyst’s fair value for a day is exactly the market-structure evidence that revives prime-brokerage disclosure debates.

FAQ

When does SanDisk report Q4 FY26 earnings?

SanDisk reports fiscal fourth-quarter and full-year results on Wednesday 5 August 2026, followed by an Investor Day on 13 August. Consensus is for EPS of $34.67 against $0.29 in the year-ago quarter, and the options market is pricing a move of roughly 25% in either direction — far above the 8.75% the stock has averaged after its last four prints.

Why did SanDisk stock drop in July 2026?

A sector de-rating (Meta Compute, July 17) collided with the forced liquidation of the stock’s biggest levered holder: Situational Awareness held ~$1.1 billion of SNDK in stock and calls at 4x gross leverage, was margin-called by its three prime brokers, and sold everything to Citadel — SNDK printed $1,015.89 on liquidation day and finished July at $1,214.83, down 37% for the month.

What is the SanDisk stock price prediction for 2026?

Our call: $2,218 base (the 23-analyst consensus), $3,000 bull (Bernstein’s July target, on confirmed NAND pricing and HBF design wins), $1,000 bear (the street-low target the July 29 flush nearly touched). Spot reference: $1,214.83 at the 31 July close.

Is SanDisk still the best-performing S&P 500 stock of 2026?

Even after a 37% July, the stock trades at $1,214.83 against a $275.24 January open — up roughly 341% year-to-date, still among the index’s very best. At the June 25 record of $2,335 the YTD gain touched 707%.

How much SanDisk did Aschenbrenner’s fund hold?

Roughly $1.1 billion combined — $724.4 million of equity plus call options — per the May 18, 2026 13F, making SNDK the fund’s second-largest disclosed position. The stake went to Citadel in the July 30 pre-open block trade.

What happens to the SanDisk shares Citadel bought?

Citadel acquired the position as distressed inventory, not conviction, and will distribute it as market liquidity allows — which typically caps rallies for several weeks after a block of this size. Watching volume texture around $1,200–1,300 tells you when the inventory clears; a high-volume reclaim of the $1,600 pre-flush shelf would be the confirmation signal.

What is high-bandwidth flash (HBF)?

An architecture positioning NAND flash closer to AI accelerators as a capacity tier below high-bandwidth memory — the roadmap that has SanDisk priced as an AI-infrastructure supplier rather than a commodity storage vendor. Its first named-customer volume win is the bull case’s key pending catalyst.

One closing benchmark for sizing the opportunity against its risk: at $1,214.83, SanDisk trades 45% below the consensus target and 21% above the street’s most bearish number — a risk-reward geometry of roughly two-to-one toward the consensus that existed nowhere in this stock during its untouchable spring run. Forced sellers create those geometries; they do not guarantee them. Wednesday night is when the company gets to say whether the geometry is real.

This article is informational analysis only and is not financial or investment advice. Equities are volatile and can lose substantial value rapidly. Past performance and analyst targets do not guarantee outcomes. Do your own research and consult a regulated financial adviser before making any investment decision.