Record Quarter, In-Line Guide, Peak Margin: The $59.8B Contract Book Now Carries the Thesis

Results vs. Consensus

Sandisk closed fiscal 2026 with the largest quarter in its standalone history by a wide margin, and with a print that cleared both the Street and its own guidance on every line that matters.

Consensus for the reported quarter spanned $8.30B to $8.48B on revenue and $34.24 to $34.96 on non-GAAP EPS across the providers we track; the table uses the mid of that range for revenue and the highest compiled figure for EPS, so the beat magnitudes shown are the conservative version of the story.

Year-over-Year Comparison

Sequential Comparison

Fiscal Year 2026 in Full

The four quarters of FY26 sum cleanly to the annual figures: Q4 revenue of $8,965M plus Q3 of $5,950M leaves $5,333M for the first half, which is exactly what the segment disclosures imply. Fiscal 2026 was a 53-week year ending July 3, 2026.

Revenue

Revenue of $8,965M grew 51% sequentially and 372% against a year-ago quarter in which Sandisk was still a freshly separated company printing 26.2% gross margins. The composition of the sequential growth is the number that matters for modeling: roughly one-third came from higher volumes and two-thirds from higher pricing. That is a meaningfully better mix than Q3, when the entire sequential gain came from price while bit shipments were deliberately held back. Volume is now contributing again, which is what you want to see if you believe the pricing curve is flattening.

Assessment: The revenue beat is high quality because it is no longer purely a price event. But two-thirds of a 51% sequential move is still price, and management guided Q1 growth to come from "both bit growth and modest price increases." The word doing the work in that sentence is "modest."

Margins

Non-GAAP gross margin of 84.6% is the fourth consecutive quarter of expansion and, in our view, the peak. Operating leverage below the gross line was equally striking: non-GAAP operating expenses of $484M represent 5.4% of revenue against 7.5% in Q3, so operating margin expanded 8.3 points to 79.2% on a 6.2-point gross margin gain. Roughly 65% of operating expenses is R&D, which is the right shape for a company whose supply growth comes from node transitions rather than new fabs.

Assessment: The operating leverage is structural and will persist; the gross margin is not going higher. Management guided Q1 to 83-85% with pricing up, and the reason is arithmetic rather than conservatism. NBMs price at "around 80%" and are scheduled to move from roughly a third of bits to more than half in FY27 and two-thirds in FY28. A book converging on 80% cannot sustain a blended 84.6%. Expect gross margin to grind toward the low 80s through FY27 and toward 80% in FY28, absent a spot market that keeps running.

Earnings and Cash

Non-GAAP EPS of $39.25 beat the guidance midpoint by 24.6%, on 157M diluted shares. Adjusted free cash flow of $5,035M is a 56% margin, and the company ended the year with $4,762M of cash, $1,777M of marketable equity securities, and no debt at all: both the current portion and the long-term balance are zero, against $1,849M a year ago. Gross capital expenditure was $562M in the quarter, 6.3% of revenue.

Assessment: Earnings quality is good and cash quality is merely fine. The $1,938M of NBM prepayments and deposits inside operating cash flow is the item to watch, because it grows the reported cash balance while creating a contractual obligation to deliver product. Two other balance-sheet lines moved in the same direction: refund liabilities went from $126M to $1,500M over the year, and contract liabilities from $25M to $1,242M including the non-current portion. The correct way to read the balance sheet is $4,762M of cash against roughly $2.5B of customer money that arrived early.

Segment Performance

Datacenter

Datacenter revenue more than doubled again, to $2,977M from $1,467M, and finished the fiscal year at $5,153M against $960M. The full-year growth rate of 437% is the single most important number in the release, because it is the transition from a company whose fortunes tracked handset and PC units to one levered to hyperscale capital budgets. The QLC Stargate platform began shipping for revenue in the quarter, which completes the portfolio: TLC for performance-intensive compute, QLC for high-capacity AI data lakes.

"A year ago, Datacenter represented roughly 12% of our bits. Exiting fiscal year '26, it represents 38% of our portfolio and is our fastest-growing end market."
— David Goeckeler, Chairman and CEO

There is a detail buried in that comparison worth pulling out. Datacenter was 33.2% of Q4 revenue but, on management's framing, 38% of the portfolio exiting the year. If the 38% is a bit-mix figure, as the preceding sentence about bits implies, then datacenter carries a below-average revenue per bit. That is what you would expect from high-capacity QLC drives sold against premium mobile and consumer NAND, and it has a consequence: growth in datacenter mix is dilutive to blended ASP per bit even while every individual price is rising.

Assessment: Datacenter is doing exactly what the thesis required, and faster. The mix caveat does not change the direction, but it does explain part of the gross margin guide-down that management attributed to "mix" without elaborating. As datacenter goes from 38% of bits toward half, the blended margin faces a second downward pull on top of the NBM pricing convergence.

Edge

Edge revenue of $5,432M grew 48% sequentially and 392% year over year, and at 60.6% of revenue it remains the largest end market by a distance. That is a striking result given that the underlying device markets are shrinking. Management expects smartphone and PC units down mid-teens this calendar year, with smartphone average capacity up significantly and PC capacity flat, then units flattening and capacities rising in calendar 2027.

"Near term, both PCs and smartphones are working through a period of adjustment as demand is shifting towards AI-enabled devices and premium configurations, driving higher storage content, particularly in smartphones."
— David Goeckeler, Chairman and CEO

The honest reading is that Edge grew 48% sequentially on price, not on units, and that the unit backdrop deteriorates before it improves. The offset is content per device, which is the durable part: an AI-capable phone needs more flash regardless of how many are sold.

Assessment: Edge is the segment most exposed to a pricing normalization, because it has the least NBM coverage relative to its size and its unit base is contracting. Management's own framing has these markets stabilizing in calendar 2027, which means FY27 Edge growth depends almost entirely on price holding. This is where a spot correction would show up first.

Consumer

Consumer revenue fell 32% sequentially to $556M and 5% year over year, the only line in the release that went backwards. This is not a demand collapse, it is a deliberate refusal to chase volume at prices the retail channel cannot absorb.

"So there's no doubt we're seeking at where is the right equilibrium point for pricing and the amount of products shipped and all these kinds of issues in consumer, and that's a little longer process than in some of the enterprise markets."
— David Goeckeler, Chairman and CEO

Retail pricing moves on a slower clock than enterprise contracts, so in a fast-rising market the consumer channel becomes the place where bits are worth the least. With every bit allocatable to datacenter or Edge at higher realized prices, shrinking consumer is the correct decision, and Consumer at 6.2% of revenue is now small enough that the decision costs little.

Assessment: Rational allocation, not weakness, and management was candid about it. The strategic question is what Consumer is for. Sandisk describes its retail brand and channel presence as a differentiator, and it is a genuine one, but a segment run at 6% of revenue and falling is being harvested rather than invested in. Worth watching for a strategic review if the mix keeps shifting.

The NBM Book: Now the Entire Thesis

The New Business Model framework was introduced last quarter with three signed agreements and $42B of RPO. One quarter later the book has roughly doubled on every measure the company discloses, from $42B of RPO to $59.8B at quarter end and $91.1B pro forma, and the disclosure has become detailed enough to underwrite.

"The total expected revenue from all our NBMs we have signed is a minimum of $93.9 billion, assuming floor pricing. We believe actual revenue will be above that minimum. The remaining performance obligation, or RPO, at the end of the quarter was $59.8 billion and would be $91.1 billion including the 2 NBMs signed after the quarter closed."
— Luis Visoso, CFO

Set against a $20.2B fiscal year, $91.1B of pro-forma RPO is roughly four and a half years of current revenue already contracted. The guarantee structure is what elevates this above an ordinary long-term agreement.

"Each one of our NBMs include financial guarantees through a combination of cash deposits and financial instruments totaling $16.5 billion, which are intended to protect Sandisk if a customer fails to satisfy its purchase obligations under these agreements."
— Luis Visoso, CFO

Key Topics & Management Commentary

Overall Management Tone: Assured and noticeably less promotional than last quarter, with the prepared remarks organized around durability rather than magnitude. Where the Q3 call sold the transformation, this one described it as finished and moved on to how the cash gets returned. The one place management was less convincing was the gross margin guide-down, which drew two separate analyst challenges and produced a mix-and-component-cost explanation that does not fully account for the NBM pricing convergence management had itself just described.

1. The Guide Is the Story, Not the Quarter

The Q1 FY27 outlook of $10.30-10.80B in revenue and $44-46 in non-GAAP EPS is a good absolute guide: revenue up 17.7% sequentially at the midpoint and EPS up 14.6%. It is also the first guide in four quarters that did not clear the Street by a wide margin. Compiled consensus sat at $10.62-10.80B on revenue, with buy-side expectations reaching $11.16B, so the guide's high end merely matches the published number and the midpoint is about 2% below it. On EPS the guide midpoint of $45.00 lands inside a $44.21-45.58 consensus range.

An analyst raised this directly on the call, noting the after-hours reaction in real time.

"If you look at Q1, we expect both -- the growth to come from both bit growth and modest price increases. So we expect both of them to be contributing to our revenue growth."
— Luis Visoso, CFO

Assessment: A company that beat its own guidance midpoint by 24.6% on EPS and then guided in line has told you something about the shape of the curve, whether or not it intended to. The beat cadence is decelerating in a specific way: Q3 cleared its EPS guide midpoint by roughly 80%, Q4 by 24.6%, and Q1 is guided to consensus. That is what the transition from a spot-priced book to a contracted one looks like, and it is a feature of the model rather than a warning about demand. It is nonetheless a real change in the return profile.

2. Gross Margin Guided Down While Pricing Is Guided Up

The most-pressed issue on the call, raised independently by two analysts, was why gross margin guides to 83-85% when the company also says pricing rises sequentially. Management gave three reasons: mix, prudent assumptions on component costs, and the observation that the high end of the guide is above what was just printed.

"I would not assume that our new business models are a drag on gross margin. They are good margins."
— Luis Visoso, CFO

Both things can be true. NBM margins near 80% are excellent against a business that earned 30.3% in FY25 and 26.4% in the year-ago quarter. They are also below 84.6%, and NBM share of bits rises from roughly a third to more than half this fiscal year. Mix toward a lower-margin book is dilutive to the blended rate by definition, however attractive that book is in absolute terms.

Assessment: The answer was framework-level where a numerical bridge would have settled it. We read 84.6% as the cycle peak and model gross margin drifting to the low 80s across FY27 and toward 80% in FY28 as NBM coverage reaches two-thirds. That is not a bear case. It is the price of the floors, and it is knowable now rather than discoverable later.

3. Capital Return Becomes the Dominant Use of Cash

Sandisk repurchased 2,836,000 shares for $4.5B in the quarter, consuming three-quarters of the $6B authorization announced only three months earlier, and the board added $14B on top. Remaining authorization is $15.5B, which reconciles exactly: $6.0B less the $4.52B deployed leaves $1.48B, plus $14B. Against a market capitalization near $188B, the outstanding authorization is roughly 8% of the company.

"We believe that at this point in time, and things may change as time goes on, we believe that the best way to do it is to return cash to investors via share buybacks. So we started that journey very quickly in last quarter, and we'll continue that journey, and we'll keep you updated. But that's the form we believe it's better for our shareholders, more tax efficient and probably the right thing for us to do at this point in time."
— Luis Visoso, CFO

The execution deserves scrutiny that it did not receive on the call. Dividing $4,524M of cash outlay by 2,836,000 shares gives an average repurchase price near $1,595. The stock closed at $1,258.58 on August 6. The first tranche is roughly 21% underwater, an unrealized shortfall in the neighborhood of $950M.

Assessment: The capital-return commitment is genuine and the balance sheet supports it. The price discipline is unproven, and buying aggressively into a stock that had risen more than 3,000% in twelve months was a poor entry regardless of the long-term view. "Very consistent in our execution" reads as a commitment to programmatic buying, which we would prefer to opportunistic timing given what the first tranche demonstrated. Investors should model the $15.5B as roughly 8% of shares at current prices and treat any accretion as a bonus rather than a plan.

4. Sellable Bit Growth Trimmed for FY27

The long-standing supply framework is mid-to-high teens bit growth achieved through node transitions rather than wafer additions. For FY27 that number comes down, because inventory is being built to service the NBM book.

"We plan to operate at higher inventory days, consistent with current levels to support our NBMs and account for higher component costs. The higher inventory levels reduce sellable bits to mid-teens for the full year fiscal year 2027."
— Luis Visoso, CFO

Capital expenditure moves the other way: dollars rise year over year to fund BiCS 8 and BiCS 10 transitions, while landing near 6% of revenue for the full year. Inventories already grew $619M over FY26 to $2,698M.

Assessment: A modest but real trim, and one that compounds with the margin convergence. If sellable bits grow mid-teens and blended pricing rises only modestly, FY27 revenue growth is far more dependent on the first-half run-rate holding than the headline TAM numbers imply. This is the quietest negative in the release and it went unchallenged in Q&A.

5. The TAM Framing and What It Does Not Reconcile With

Management put hard numbers on the industry for the first time: a NAND market exceeding $300B in calendar 2026, up threefold year over year, approaching $500B in calendar 2027, with datacenter share of that market moving from roughly 30% in 2025 to roughly 50% in 2026.

"We estimate the NAND market will exceed $300 billion in revenue in calendar year 2026, up 3x year-over-year. Looking further ahead, we estimate that the NAND market will approach $500 billion in revenue in calendar year 2027."
— Luis Visoso, CFO

An analyst did the arithmetic on air: if the industry grows roughly 67% in calendar 2027 and Sandisk's bit growth is not accelerating, the balance has to come from price. Management's answer was that it intends to grow with the market and expects to hold share. Annualizing the Q1 guide midpoint gives a $42B run-rate against a market approaching $500B, or under 9%, well below Sandisk's share of industry bits.

Assessment: The TAM figures are management's own and are not independently verifiable, and the share arithmetic does not obviously reconcile unless the $500B definition is broader than Sandisk's addressable slice, or unless Sandisk undergrows it. "Grow with the market" is the least quantitative answer on the call. Treat the TAM slide as directional evidence of tightness, not as a revenue bridge.

6. Bits on Allocation Beyond Calendar 2027

The supply-demand claim that underwrites everything else was stated plainly and without hedging.

"Demand from our customers is growing faster than our supply. We, therefore, expect bits to remain on allocation beyond calendar year 2027."
— Luis Visoso, CFO

Corroborating evidence arrived from an unusual direction. Customers who signed NBMs one quarter ago have already returned asking to increase committed volumes, which is why two of the five new agreements are expansions of existing deals rather than new logos.

Assessment: Customers raising multi-year commitments within a single quarter of signing is the most credible demand signal in the release, more so than any TAM forecast, because it costs those customers money to be wrong. Allocation extending beyond calendar 2027 is the single assumption on which the FY28 earnings power rests.

7. The Era of Inference and the Structural Demand Case

The demand narrative has tightened from last quarter's "only economically viable solution" framing into something more specific about workload architecture.

"Underlying our performance is the most important force in our market, the Era of Inference. AI is fundamentally a memory-centric storage-intensive problem."
— David Goeckeler, Chairman and CEO

On the KV cache opportunity specifically, management characterized its own view as still maturing and use-case dependent, while getting more bullish in aggregate as models grow and context lengths lengthen, with agentic workloads acting as a multiplier on top.

Assessment: The direction of travel is right and the customer engagement supports it. But management's framing has shifted subtly from asserting a fixed architectural requirement to saying customers will define the architecture and Sandisk will follow. That is a more honest position and a slightly weaker one. It concedes that the NAND intensity of inference is not yet a settled number.

8. High-Bandwidth Flash and BiCS 10

Sandisk announced a global standard for high-bandwidth flash with SK Hynix and has added members to its HBF advisory board, and BiCS 10 was announced within the last few weeks. Asked directly how far away HBF samples are, whether one, two or three years, management declined to answer and deferred to an investor day the following week. Asked whether BiCS 10 and HBF carry a price premium, the answer was that it is early.

"I think it's a little early to talk about pricing on some of those nodes just yet, right, and some of that technology, but we'll keep you updated on that."
— David Goeckeler, Chairman and CEO

Assessment: HBF is a genuine option on the AI memory hierarchy and a co-developed standard with a DRAM-scale partner is more than a science project. But it carries no revenue, no timeline, and no disclosed pricing, so it belongs in the optionality bucket rather than the model. The refusal to bracket sample timing even to the nearest year is a mild negative given the question was asked in the simplest possible form.

9. Balance Sheet: Debt-Free, With Customer Money Inside It

Sandisk ended the year with $4,762M of cash, $1,777M of marketable equity securities, and zero debt against $1,849M carried a year ago, with $1,900M of debt repaid during fiscal 2026. Total shareholders' equity is $15,736M against $9,216M a year ago, despite $4,537M of treasury stock.

The marketable equity securities line is entirely the Nanya position: $970M purchased in the quarter plus the $807M gain equals the $1,777M carrying value. Less discussed is the tax position. Income taxes payable rose to $1,286M current and $258M non-current, while cash paid for income taxes across all of FY26 was only $146M against $1,584M of GAAP expense.

Assessment: The balance sheet is genuinely strong and the debt paydown removes the last structural vulnerability from the spin. Two adjustments belong in any cash model. First, roughly $2.5B of the cash arrived as NBM prepayments and deposits and carries a delivery obligation. Second, a cash-tax catch-up of well over $1B is deferred and will land in FY27 or FY28, which is a free cash flow headwind nobody asked about.

10. Customer Concentration Arrives Without Disclosure

Eight customers now account for more than half of FY27 bits and roughly two-thirds of FY28. Management declined to name them or to characterize the concentration, calling specifics not in the company's interest to disclose, while confirming the deals span both datacenter and Edge.

Assessment: The customer quality is almost certainly high, given the counterparties posted $16.5B of guarantees through third-party financial institutions. But the risk profile has changed materially and the disclosure has not kept pace. A business where eight relationships determine two-thirds of FY28 output is a different security than one selling into a fragmented spot market, and the annual report's concentration table will be the first place investors can size it.

11. Investor Day as a Deferral Mechanism

An investor day in New York was scheduled for the week after the call, and it absorbed at least four questions: HBF timing, the inference architecture framework, BiCS 10 detail, and the long-term financial model. Management referred to a forthcoming "new model for everybody" last quarter and pointed to the event repeatedly on this one.

Assessment: Legitimate sequencing rather than evasion, but it does mean this print left the two most important forward questions unanswered: what steady-state gross margin looks like once NBM coverage matures, and what fiscal 2027 looks like in full. Both should arrive within days, which makes the investor day the more important event of the two.

Guidance & Outlook

Implied sequential ramp: the revenue midpoint of $10,550M requires 17.7% sequential growth, from both bit growth and what management called "modest price increases." In Q4 the split was one-third volume and two-thirds price. With full-year sellable bit growth guided to only mid-teens, a 17.7% step in the first quarter alone means either the volume contribution is heavily front-loaded against a flatter remainder of the year, or pricing is contributing more than the word "modest" implies. Both readings front-load FY27, and neither is consistent with a smooth four-quarter ramp.

Street at: consensus revenue of $10.62-10.80B against a $10.55B midpoint, and consensus EPS of $44.21-45.58 against a $45.00 midpoint. The revenue guide is the miss; EPS is in line. The gap between the two tells you the Street was modeling a lower gross margin than the company guided, which is a point in management's favor that got lost in the reaction.

Guidance style: conservative, and demonstrably so. The company cleared the EPS midpoint of its guide by 24.6% in the quarter just reported and by roughly 80% in the one before it. There is also conservatism inside this guide: running the high end of revenue ($10,800M) at the high end of gross margin (85.0%) against the low end of operating expenses ($520M) and the guided 15% tax rate produces roughly $47.50 of EPS on 155M shares, about $1.50 above the guided $46.00 ceiling. Management is not underwriting the corners coinciding. The same arithmetic at the low end (revenue $10,300M, margin 83.0%, operating expenses $540M) reproduces the $44.00 floor almost exactly, so the conservatism sits at the top of the range rather than the bottom.

What is missing: no full-year FY27 revenue or EPS frame, despite more than half of FY27 bits committed with purchase orders already in hand and management claiming over four years of visibility. That absence is conspicuous, and the investor day is the stated reason.

Analyst Q&A Highlights

NBM Economics and Whether 80% Still Holds

The opening question went straight to the number that determines the medium-term model: whether the roughly 80% gross margin attached to NBM contracts last quarter still applies, or whether the contracts are repricing toward the higher rates the business is currently printing. Management confirmed the 80% level and framed the variable component as the mechanism for capturing some upside if prices keep rising.

Q: "Last quarter, you talked about the NBMs were being signed in this 80% margin range. I know there's a lot of puts and takes with regard to pricing. I was wondering if you can talk about whether it's in that ballpark or if it's kind of trending actually more towards the guidance for the next quarter. I would think it's more in the 80 ballpark."
— Ben Reitzes, Melius Research

A: "Ben, we -- not to pick a specific number, but we do expect to be around 80% for the new business models. As you know, there is some upside as prices continue to go up, we'll capture some of that upside as well. So -- but we feel very good about the work we've done."
— Luis Visoso, CFO

Assessment: This is the most important exchange on the call and it was the first one. Confirming 80% while the reported quarter printed 84.6%, and while NBM coverage moves from a third of bits to more than half, establishes the direction of blended gross margin without management having to say the word. Everything in the guidance discussion that followed is downstream of this answer.

Why Margin Falls When Price Rises

Two separate analysts pressed the same apparent contradiction, making it the dominant technical topic of the call: the guide has sequential pricing up and gross margin down. The response identified mix and component-cost assumptions, and noted that the top of the guided range sits above the printed quarter.

Q: "I'm trying to understand the margins are down modestly, but pricing is up. I just want to understand, is it the impacts of these new business agreements that's kind of capping -- leading to the modest growth in ASP? And then even with that, is there any other factors that are contributing to gross margin in September?"
— Multiple analysts incl. Blayne Curtis, Jefferies; C.J. Muse, Cantor Fitzgerald

A: "Now if you look at some of the reasons on the math that you're questioning, right? I mean there is some mix. There is -- the high end of the guide is 85%, right? And so it's a little bit up versus what we just printed. And we're making some prudent assumptions, if you wish, on component costs and other things. And when we factor all of that in, we believe that the right guidance to give you is somewhere between 83% and 85%."
— Luis Visoso, CFO

Assessment: The answer is accurate as far as it goes and incomplete. Mix is doing more work than the phrasing suggests, and the mix in question is the shift toward an 80%-margin contract book that had just been described in the prepared remarks. Management would have been better served bridging it explicitly, because leaving it implicit invited the market to supply a darker explanation about component costs or spot softness.

What Makes the Contracts Durable

The most probing question of the call asked why investors should believe in the durability of this cycle rather than simply extrapolating margin and EPS revisions. The answer covered both the contractual architecture and the change in the customer relationship, and it produced the most candid line on the guarantees.

Q: "The challenge for semi investors is moving less of a focus on margin and EPS revisions, but rather buying into the durability of this cycle. So can you focus on why you're so confident in the securities of NBMs and that today's supply is limited and that will drive a durable and elongated cycle?"
— C.J. Muse, Cantor Fitzgerald

A: "Number one is we put a contractual structure in place where we align our incentives. They're making financial commitments to us that if they don't follow through on their commitments that I think $16.5 billion in aggregate would flow to us on our balance sheet. We don't ever expect to see that money, quite frankly. It's just to align interest."
— David Goeckeler, Chairman and CEO

Assessment: The right answer, and the honest framing of the guarantees as an alignment mechanism rather than an expected recovery is more credible than treating $16.5B as a recoverable asset. The same exchange produced the clearest statement of intent on margin: management characterized mid-80s gross margin as "a fair return" and said it wants to get the boom and bust out of the business. Read together with the 80% NBM answer, this is management telling you it will trade some peak margin for duration.

The Pace and Size of Buybacks

A question framed the buyback in terms of what a sustained $5B quarterly pace would mean against the market capitalization, roughly 10% over four quarters, and asked whether that is the right order of magnitude. Management did not commit to a pace but pointed to what it had already executed and to the consistency it intends.

Q: "If we just round up to $5 billion per quarter, I know the authorization isn't this big. But if you commit to buying $5 billion a quarter, that's -- and you do that over the next 4 quarters, that's about 10% of your market cap. I mean, is that the kind of buyback magnitude we should be thinking of? Or is it too early to kind of call the pace?"
— Ben Reitzes, Melius Research

A: "Our second authorization now, you saw we were -- we had a $6 billion authorization. We executed $4.5 billion in the first quarter of that. Now we've reauthorized another $14 billion. We expect to be very consistent in our execution of this program. We have a lot of confidence in the cash generation of the portfolio."
— David Goeckeler, Chairman and CEO

Assessment: A non-answer on pace that nonetheless implies a large one, since $15.5B of authorization at roughly $5B a quarter is three quarters of buying. Nobody asked what the first tranche cost, and at an implied average near $1,595 against a $1,258.58 close, somebody should have. The commitment is sound; the entry was not.

Whether the Guide Is Volume, Price, or Conservatism

An analyst noted that the Q1 guide looked light against expectations and that the stock was already falling in the post-market, then asked directly whether the shortfall was volume, the pricing lock-in from the long-term agreements, or simply conservatism. The response addressed the Q4 split and the Q1 composition without engaging the conservatism question.

Q: "Particularly for the Q1 guide, it seems a bit lighter compared to expectations. Share price has been down a bit last time I checked in the post market. I just wondered if that is volume related or if that is because the locking in of pricing and so less price upside or perhaps a bit of both? Or is it a bit of conservatism?"
— Mark Newman, Bernstein

A: "So on Q4, Mark, what we mentioned is that about 1/3 of the growth came from incremental bits and 2/3 came from pricing. So you have that split there. And obviously, as you compare us versus others, there is a different timing on when price increases were taken. So you have to look at things over several quarters to really understand where we are because, as you know, we took significant pricing in the prior quarter."
— Luis Visoso, CFO

Assessment: The question offered management three explanations and an easy exit through the conservatism door, and it took none of them. Declining to say "we are being conservative" when the stock is visibly falling is either unusual discipline or a signal that the guide is closer to a real forecast than the last several were. Given the beat history, we lean toward the former, but the refusal is worth noting rather than assuming.

Industry Growth Versus Company Growth

A two-part question tested the internal consistency of the TAM framework: if the industry is forecast to grow from $300B to $500B while Sandisk's bit growth does not accelerate, is price normalization assumed, and does the company expect to outgrow, match, or lag that market. The answer committed only to matching it.

Q: "You have an industry forecast here of the NAND industry growing to $500 billion in '27 from $300 billion in '26. And we know that your bit growth is really not meaningfully accelerating for the industry. So just is the assumption that there is going to be any price normalization that we expect in 2027 or not?"
— Wamsi Mohan, Bank of America

A: "We want to grow with the market, and we plan to grow with the market. Look, I mean, Wamsi, we're transitioning our business into a more predictable business. We're working on pricing and predictability and duration. And so we expect to grow with the market as it goes and increase our visibility, we continue to see very robust demand through the end of '27 and into '28."
— David Goeckeler, Chairman and CEO

Assessment: The price-normalization half of the question went unanswered entirely, which is the most notable omission in the Q&A. A company with floors and ceilings on more than half its FY27 bits is better positioned than most to describe what normalization would do to it, and chose not to. A separate exchange confirmed the intention to hold segment share against the $500B figure, so the company is at least internally consistent about matching rather than beating the market.

Capital Spending and the Temptation to Accelerate

With peers who also make DRAM apparently prioritizing that product, an analyst asked whether Sandisk would consider spending more to accelerate its node transitions and capture more of the tight market. Management declined and restated the framework, while flagging the FY27 trim.

Q: "You talked a little bit about gross CapEx coming up a little for the BiCS 8 and BiCS 10 transitions. Can you just talk about how you think about that and the fact that the guys who also make DRAM seem to be focused more there. Any thought of spending more to accelerate those technology transitions?"
— Joe Moore, Morgan Stanley

A: "We continue to be committed to grow bits mid- to high teens, right? And we're spending to that level. The reason why our dollar increases from 1 year to another is because obviously, as you transition from one node to another, you will first take the easiest, right, or the cheapest transitions and then you would go on and do the more expensive ones, particularly when we had underutilized the fabs in the prior year, we have some tools that could be reused."
— Luis Visoso, CFO

Assessment: Discipline held for a fourth consecutive quarter, which is the behavior that keeps the industry tight and is the strongest single argument that this cycle differs from prior ones. The embedded disclosure matters more than the answer: rising capital expenditure dollars now buy the more expensive node transitions, because the cheap ones and the reusable tools from the underutilized-fab era are spent. Capital intensity per incremental bit is going up even as the revenue percentage falls.

What They're NOT Saying

1. No fiscal 2027 annual guidance: more than half of FY27 bits are committed, purchase orders are in hand, and management claims over four years of visibility, yet the outlook stops at one quarter. Whatever the investor-day rationale, a company with this much contracted volume declining to frame a year is withholding the most useful thing it has.
2. No floor price, and no floor margin: "attractive margins even at floor pricing" is an assertion, not a number. Without the floor level, the downside case that justifies the entire de-cyclicalisation thesis cannot be modeled by anyone outside the company.
3. No NBM ceiling disclosure either: ceilings were acknowledged for the first time this quarter and never quantified. The ceiling determines how much of any further spot rally accrues to shareholders, and it is now the binding constraint on more than half the book.
4. No customer concentration figure: eight customers, more than half of FY27 bits, roughly two-thirds of FY28, and not one revenue-concentration percentage. Management explicitly declined, saying specifics are not in the company's interest to disclose.
5. Nothing on the buyback's average price: shares and dollars were both given, which lets anyone derive roughly $1,595, but no one on the call connected that to a stock trading materially lower. The silence is understandable and the omission is still an omission.
6. No explanation of the refund liability: it went from $126M to $1,500M over the fiscal year, and contract liabilities from $25M to $1,242M. The split between returnable NBM deposits and ordinary channel price protection is material to cash quality and was not addressed.
7. No mention of the coming cash-tax catch-up: $146M of cash taxes paid in FY26 against $1,584M of GAAP expense, with $1,544M now sitting in income taxes payable. That reverses into free cash flow at some point and nobody asked when.
8. No HBF timeline: asked whether samples are one, two or three years away, management deferred entirely. The question was constructed to accept the vaguest possible answer and still did not get one.
9. No steady-state margin framework: a second consecutive quarter of declining to name a target gross margin, despite the NBM mix schedule making the trajectory largely arithmetic at this point.
10. Nothing on a dividend: raised directly, and answered by restating a preference for buybacks as more tax-efficient. Reasonable, but a company generating $8.7B of adjusted free cash flow with no debt will face the question again.
11. No answer on price normalization in 2027: the explicit half of an analyst's two-part question, skipped in the response and not followed up.

Market Reaction

• Pre-print setup: the stock closed at $1,350.50 on August 5, itself down 5.4% that session. It entered the print up 468.9% year to date and up 3,120.8% over the trailing twelve months, but down 22.6% over the trailing thirty days from $1,744.43 on July 6. The 52-week closing range was $40.69 to $2,335.00, so the shares came into the print roughly 42% below their closing high.
• After hours: down roughly 8% to near $1,243 immediately on the release, extending to a 9% to 10% decline in the pre-market as the guide was digested.
• Reaction session, August 6: opened at $1,173.26, a 13.1% gap down, traded as low as $1,163.09 (down 13.9%), then recovered through the session to close at $1,258.58, down 6.8% or $91.92. The intraday recovery off the low was 8.2%.
• Volume: 19.2M shares against a 30-day average of 15.3M, or 1.3x. Elevated but not extreme, which is notable for a 13.1% opening gap.
• Market context: the S&P 500 was down 0.2% on the session and up 12.8% year to date, so essentially all of the move was company-specific.

The reaction has a clear proximate cause and a more important underlying one. The proximate cause is the revenue guide: $10.30-10.80B against a consensus of $10.62-10.80B and buy-side numbers reaching $11.16B. For a stock that had beaten and raised by enormous margins for four consecutive quarters, an in-line guide is a regime change, and momentum capital exits regime changes quickly.

The underlying cause is that the de-rating started weeks earlier. The shares were already down 22.6% over the thirty days into the print and 42% from their closing high, so the market had begun repricing the multiple before it saw a single number. That sequence matters for how the reaction should be read: this was not a shock that broke a rising stock, it was a confirmation delivered to a market that had already started asking whether the margin could hold.

The intraday pattern is the constructive detail. A 13.9% opening drawdown that closes down 6.8% on only 1.3x volume is a market absorbing supply rather than liquidating. Buyers stepped in below $1,200 with the print's actual numbers in hand: record revenue, 84.6% gross margin, a contract book at $91.1B pro forma, and $15.5B of buyback authorization. Whoever sold at the open sold the guide; whoever bought at the low bought the book.

Street Perspective

Debate: Is 84.6% the Peak, and Does It Matter?

Bull view: peak or not, the level is extraordinary and the floor underneath it is contractual. Even a blended margin converging to 80% leaves operating margin near 75% and earnings power far above anything the Street modeled a year ago. The bull camp argues the market is anchoring on the second derivative while ignoring the level.

Bear view: memory stocks de-rate when margins peak, not when they roll over, and the peak is now identifiable and dated. The bear camp points out that management guided margin down in a quarter it guided price up, and that the NBM ceilings mean no upside surprise can rescue the trend.

Our take: the bears have the mechanism right and the magnitude wrong. Margin has peaked and the multiple should compress, which is precisely what the last thirty days have been doing. But the compression from a 52-week closing high of $2,335 to $1,258.58 has already discounted a great deal of that, and a business earning 80% gross margins on a four-year contracted book does not deserve the trough multiple the market applies to unhedged memory. The de-rating is real, largely done, and not a reason to sell at 7x.

Debate: Is the RPO an Asset or an Accounting Artifact?

Bull view: $59.8B of RPO on the books, $91.1B including deals signed after quarter end, $16.5B of financial guarantees behind it through third-party institutions, purchase orders in hand for the year, and customers returning within a single quarter to raise their commitments. The bull camp treats this as the most de-risked revenue backlog in the history of the NAND industry.

Bear view: the $93.9B minimum assumes floor pricing that has never been disclosed, so the minimum could sit far below current revenue rates. Guarantees align incentives but do not compel purchase, and management said openly it never expects to collect them. Eight counterparties determine two-thirds of FY28, and hyperscale capital budgets have been revised down before.

Our take: closer to the bulls, with one correction. The book is a genuine asset and the guarantee structure is more than window dressing, because posting $16.5B through third-party institutions is expensive and firms do not do it casually. But the undisclosed floor is the weak point, and it is the reason we lowered conviction rather than raised it. Until Sandisk quantifies the floor, the downside case rests on management's word that the margins there are "attractive," and an unfalsifiable floor is not the same as a known one.

Debate: Was the $4.5B Buyback Good Capital Allocation?

Bull view: the company generated $5.0B of adjusted free cash flow in the same quarter it spent $4.5B on stock, so the buyback was funded entirely from the period's cash and left the balance sheet stronger, not weaker. Buying roughly 1.9% of the company at any price inside this earnings trajectory looks cheap in hindsight, and the $14B follow-on signals conviction.

Bear view: an average near $1,595 against a $1,258.58 close is roughly $950M of value destroyed in three months, and it happened at the most euphoric point in the stock's history. Capital would have been better held for the node transitions that management says are becoming more expensive, or simply deployed more slowly.

Our take: the bears win this one on the facts and the bulls win it on the framework. Spending 75% of an authorization within a single quarter, at an average price 27% above where the stock trades today, is poor execution and should be named as such. But the underlying decision to return cash is correct, the balance sheet carries no debt, and $15.5B of remaining authorization at current prices is materially more accretive than what was just spent. Judge the program on the next $15.5B, not the last $4.5B, while noting that management has given no evidence of price sensitivity.

Model Update Needed

Valuation: at the August 6 close of $1,258.58, roughly 149M shares outstanding puts the market capitalization near $188B. Annualizing the Q1 guide midpoint of $45.00 gives $180 of run-rate EPS, or 7.0x. Against our FY27 range of $170-200 the stock trades at 6.3x to 7.4x, and against a conservative FY28 of $140-180 it trades at 7.0x to 9.0x. Adjusted free cash flow of $22-25B against $188B is an 11.7% to 13.3% yield, before the cash-tax reversal.

Fair value range raised to $1,600-1,950, from $1,400-1,800. The range applies a 9x to 11x multiple to a conservative FY27 figure near $175, which sits below our base case deliberately. Against the $1,258.58 close that implies 27% to 55% upside. The multiple, not the earnings, is what we are least certain about: memory franchises at cyclical peak earnings have historically traded at mid-single-digit multiples, and the argument for 9x to 11x rests entirely on the NBM floors converting this into a partially contracted business. If the floors turn out to be far below current pricing, the correct multiple is lower and so is the fair value.

Catalysts: the investor day the week of August 10, where a long-term financial model and HBF detail are both expected; the FY26 Form 10-K, which will carry the first customer-concentration disclosure and the RPO recognition schedule; and the Q1 FY27 print, which is the first quarter in which NBM coverage exceeds half the book.

Thesis Scorecard Post-Earnings

Overall: the thesis is intact but its shape has changed, and this is the first quarter since initiation in which a bull pillar was downgraded. The de-cyclicalisation pillar strengthened decisively; the margin-expansion pillar is finished. Sandisk is no longer a story about how high margins can go. It is a story about how long they stay high, and the answer is now written into contracts that investors can reason about but cannot yet see the key terms of. Two new bear points enter the scorecard, both consequences of the same structural trade rather than new risks to the business.

Action: Maintaining Outperform, lowering conviction from High to Moderate. Hold existing positions. Add on weakness below $1,150, where the stock would trade under 6.5x an annualized run-rate with a contractually floored revenue book. Trim into $1,900 or above, where the multiple would price durability that the disclosure does not yet support. The investor day the week of August 10 is the near-term swing factor: a credible long-term margin framework and a disclosed NBM floor would justify raising conviction back to High.