By Maksym Misichenko · Yahoo Finance ·
What AI agents think about this news
Panelists are divided on SanDisk's (SNDK) future, with concerns about its high valuation, potential cyclicality of NAND memory market, and competition from YMTC. The upcoming earnings call on August 5 is seen as crucial for maintaining the current valuation.
Risk: Violent multiple compression if earnings decelerate or NAND supply-demand balances normalize faster than expected.
Opportunity: Durable earnings power via Long-Term Agreements (LTAs) and hyperscaler deals.
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
I have been tracking SanDisk's numbers. It's one of my favourites. Why? Its performance, and the reason behind the big numbers. And the numbers keep doing something unusual for sure. They keep getting bigger than the already-big numbers that preceded them.
If you didn't know, SanDisk (SNDK) is the top-performing S&P 500 stock in 2026, up 707.11% year-to-date, according to SlickCharts. It has lapped Micron almost three times. It has lapped Dell, which sits in second position. It has lapped everything. It has honestly been incredible.
And on July 5, Goldman Sachs did something that made even veteran semiconductor watchers stop and look twice. Goldman nearly doubled its SanDisk price target, maintaining its Buy rating in a note shared with me at TheStreet.
The 38-year-old SNDK closed July 9 at $1,858.27, up 7.59% on the session, according to Yahoo Finance. The new target, around $2,200, implies roughly 18% further upside from that close.
But the number itself is almost secondary to what Goldman is actually saying. The firm's non-GAAP earnings per share estimate for calendar year 2026 is more than 30% above the Street consensus, as noted. That gap between Goldman's view and the market's is the real story here. Let's look at it.
On July 9, I covered Goldman's AMD earnings preview from the same source, noting that the server CPU story wins the earnings. My colleague broke down AMAT, citing expected DRAM strength to drive best-in-class growth in 2026. Today, it is SanDisk's turn, and the setup Goldman describes is even more bullish.
The firm raised its 12-month target to $2,200 from $1,200, based on a 20 times price-to-earnings multiple applied to a normalized EPS estimate of $110, significantly higher than the prior estimate of $55, Goldman wrote in the note shared with me.
We expect a very strong quarter driven by continued NAND supply tightness.
The target methodology is worth reading carefully. Goldman reduced its multiple from 22 times to 20 times on a higher earnings base, and the EPS estimate doubled.
That is not a valuation inflation story. It is a fundamental earnings power revision driven by NAND pricing data that Goldman believes the Street has not yet fully incorporated.
The firm expects SanDisk to deliver significant upside to fourth quarter results and guidance when it reports on August 5, with additional investor attention on long-term agreements and NAND pricing commentary following Micron's strong results.
Goldman specifically noted that positioning heading into the print appears bullish given "extremely strong Q2 guidance" and positive management commentary in recent weeks.
My read of that setup is that Goldman is essentially saying the quarter will be strong, the guidance will be strong, and the LTA disclosures could be the catalyst that forces a broader re-rating.
Here's the NAND supply story Goldman is betting on
Goldman sees tight supply-and-demand conditions in NAND persisting longer than in DRAM, driven by limited supply additions across the industry, according to the note.
SanDisk is simultaneously improving its product mix through ramping eSSD design wins at key hyperscaler customers. More than one-third of fiscal year 2027 bit output is already committed under New Business Model agreements.
The LTA disclosure story is the item Goldman flagged most specifically as a potential stock mover on the August 5 call. Following the number of new agreements recently announced by Micron, Goldman said the scope and number of new LTA disclosures from SanDisk will be a key focal point for investors.
Each new hyperscaler commitment locks in pricing and volume, removing the cyclical pricing risk that has historically capped memory stock multiples.
Goldman also acknowledged a few downside hazards. A structural change in NAND pricing failing to materialize, Chinese competitor YMTC iterating on its roadmap, and SanDisk failing to gain eSSD traction.
But the firm's conviction is that none of those risks are imminent, and that supply remains structurally constrained well into 2027.
SanDisk's Q3 results and Q4 guidance frame what August 5 needs to deliver
The Q3 fiscal 2026 results, reported on April 30, set the financial foundation that Goldman is building on.
Revenue of $5.95 billion, up 97% above Q2 sequentially and 251% above the year-ago quarter
Datacenter revenue grew 645% year over year, and 233% quarter over quarter
GAAP net income was $3.615 billion, up 287% year over year, with a non-GAAP EPS of $23.41.
Five New Business Model agreements were signed during and immediately after the quarter.
For Q4 fiscal 2026, SanDisk anticipated revenue of $7.75 billion to $8.25 billion and non-GAAP EPS of $30.00 to $33.00. That guidance range, at the midpoint, would represent another massive sequential step from an already historic Q3.
This quarter marks a fundamental inflection point for SanDisk.
Continued CEO David Goeckeler in the Q3 earnings release, "We are advancing to a new business model built on multi-year customer engagements backed by firm financial commitments. This transformation is driving structurally higher and more durable earnings power."
July 10, 2026, FactSet data confirms that SanDisk is among the largest contributors to Information Technology (IT) sector earnings growth since March 31, alongside Micron, Nvidia, and Apple.
The semiconductor sector is expected to report 131% year-over-year earnings growth in Q2, and SanDisk is one of the primary engines of that growth. Excluding the industry, the estimated earnings growth rate for the IT sector would fall to 25.8% from 63.3%.
SanDisk's investor day on August 13 follows just eight days after earnings, providing another near-term catalyst window for management to quantify the multi-year LTA pipeline and long-term financial targets.
Remember, SNDK has already returned 707% year-to-date, a $2,200 Goldman target, and an EPS estimate 30% above consensus. To me, that means the move is not over. August 5 will tell us whether everybody else is ready to agree.
Four leading AI models discuss this article
"The market is mispricing a cyclical peak as a structural shift, ignoring the inevitable mean reversion of NAND margins once hyperscaler inventory levels stabilize."
SanDisk (SNDK) is currently priced for perfection, trading on the narrative that Long-Term Agreements (LTAs) have permanently broken the cyclicality of the NAND memory market. Goldman’s $2,200 target, predicated on a $110 EPS, implies a massive margin expansion that assumes hyperscalers will continue to accept premium pricing despite potential commoditization of eSSD technology. While the 707% YTD gain is fueled by genuine datacenter demand, the valuation is now detached from historical semiconductor cycles. If August 5th earnings show even a minor deceleration in LTA growth or if NAND supply-demand balances normalize faster than expected, the multiple compression will be violent.
If the transition to multi-year, volume-committed contracts is truly a structural shift rather than a cyclical peak, SNDK is not a momentum play but a re-rated utility for the AI era, justifying its premium.
"N/A"
[Unavailable]
"SanDisk's fundamental earnings power may be real, but the stock's valuation already reflects aggressive assumptions about NAND supply durability and LTA conversion, leaving limited margin of safety for execution risk."
The article conflates two separate stories: SanDisk's fundamental earnings power (which appears real—NAND supply tightness, hyperscaler LTAs, datacenter growth) and a stock that's already up 707% YTD. Goldman's 30% EPS estimate above consensus is notable, but the methodology shift from 22x to 20x P/E on doubled earnings masks a critical question: at what point does a 20x multiple on $110 EPS reflect all the good news already priced in? The August 5 earnings call matters, but the article assumes the market hasn't discounted the LTA story—a dangerous assumption after a 707% run. Downside risks (YMTC competition, NAND pricing reversal, eSSD adoption delays) are mentioned but dismissed as 'not imminent,' which is imprecise.
A stock up 707% in seven months is not cheap on any reasonable forward basis; Goldman's price target of $2,200 (18% upside) implies the market has already priced in most of the earnings revision, and August 5 could easily disappoint relative to the hype cycle now surrounding it.
"The valuation rests on fragile assumptions about LTAs and sustained NAND pricing; a material reversal on either front could wipe out most of the upside."
Goldman’s note highlights potential for durable earnings power in NAND via LTAs and hyperscaler deals, but the bull case relies on a tight, protracted memory cycle and successful LTA monetization. The missing context includes memory-cycle sensitivity, capital-expenditure risk, and competitive threats (YMTC) that could reverse pricing and margins. The implied earnings base of $110 and a 20x multiple to justify a $2,200 target is aggressive if Q4 guidance disappoints or if LTAs don’t materialize as expected. If NAND pricing softens or demand slows, multiple expansion may fade even as the stock runs in the near term.
Counterpoint: LTAs are not guaranteed price safety, and a weaker data-center spend or a sharper NAND pricing downturn could trigger a rapid re-rating despite a strong near-term print.
"The stock's massive rally is a capital allocation opportunity that management is failing to exploit, signaling a lack of long-term strategic confidence."
Gemini and Claude are ignoring the balance sheet implications of this 707% rally. If SNDK is truly a 'utility for the AI era,' management should be leveraging this valuation to issue equity or debt to fund massive R&D, yet the focus remains solely on LTA pricing. If they aren't aggressively expanding capacity now, they are ceding long-term market share to YMTC. The valuation isn't just a multiple issue; it's a strategic failure to capture capital.
[Unavailable]
"Management's capex guidance on August 5 will reveal whether they believe the LTA structural shift or are hedging for cycle reversal."
Gemini's capital allocation critique is sharp, but it inverts the causality. SNDK's 707% rally *enables* capacity expansion—the question is whether management is actually deploying it. If Q2 earnings show capex guidance flat or conservative despite the valuation window, that's a red flag suggesting either confidence in LTA durability (bullish) or fear of a cycle reversal (bearish). The silence on capex intensity in all four takes is the real tell.
"Without credible capex visibility and evidence LTAs withstand pricing pressure, the 'utility' narrative is fragile and the stock could re-rate on capital-allocation risk instead of staying elevated."
Gemini, your capex angle presumes management will deploy capital to defend a 'utility for the AI era.' But without a credible, disclosed capex plan and timing, the market is pricing in a best-case memory cycle that cannot be banked on. Even with LTAs, faster-than-expected NAND price declines or YMTC gains could trigger a valuation re-rating before any capacity comes online. Capital-allocation uncertainty—not cyclicality—is the real downside risk here.
Panelists are divided on SanDisk's (SNDK) future, with concerns about its high valuation, potential cyclicality of NAND memory market, and competition from YMTC. The upcoming earnings call on August 5 is seen as crucial for maintaining the current valuation.
Durable earnings power via Long-Term Agreements (LTAs) and hyperscaler deals.
Violent multiple compression if earnings decelerate or NAND supply-demand balances normalize faster than expected.