52.7% Gross Margin, a September Guide 25% Above the Street, and a Stock That Fell 18% Into the Print — Maintaining Outperform

Results vs. Consensus

FQ4 FY2026 Scorecard

Year-Over-Year Comparisons

Quarter-Over-Quarter Comparisons

Fiscal Year 2026 in Full

One housekeeping note that neither the release nor the call raised: fiscal 2026 was a 53-week year, with the extra week falling in the September 2025 quarter. The fourth quarter itself was a clean 13-week period against a 13-week year-ago comparison, so every quarterly figure above is week-for-week comparable; only the full-year growth rates carry the extra week, worth roughly two to three points of the 34%.

Quality of Beat

Revenue. The $3,629M print cleared the guide's own ceiling by $79M, the fifth consecutive quarter above the top of the range, and the composition is the most interesting of the streak. Data center revenue of $2,932M grew 57% year-over-year against 43% exabyte growth, an unusually wide 14-point spread that is entirely price and mix. Edge IoT, long the drag, grew 14% sequentially and 20% year-over-year to $697M on tight supply and rising NAND pricing pushing marginal demand toward hard drives; enterprise nearline revenue rose for a fifth consecutive quarter. Under build-to-order, upside of this size arrives through two channels: qualification pull-ins running ahead of plan, and incremental exabytes released by better yields clearing at spot rather than contract. Management was unusually explicit that the second channel was live this quarter, which matters for how durable the print is.

Margins. The 570 basis points of sequential gross margin expansion is the largest of the thirteen-quarter streak and arrived roughly 300 basis points above the guide's own arithmetic. Operating leverage did the rest: opex fell to 8.1% of revenue from 9.5%, so 44.6% operating margin was 710 basis points better sequentially on 570 of gross margin. The composition question is now the central analytical issue for this name, because the CFO's own decomposition has changed. Utilization was exhausted as a driver two quarters ago; mix (the HAMR transition, and specifically Mozaic 4+ delivering more terabytes on an unchanged bill of materials) is structural and running ahead of plan; pricing is doing more work than at any prior point, and part of that pricing is above-contract. A margin built on mix compounds. A margin built on spot clearing does not.

EPS. The $5.71 beat the guide's ceiling by $0.51 and consensus by $0.62, with essentially no help from below the line or from the tax rate: other income and expense of $58M was in the expected range, the GAAP effective tax rate was 14.0% ($211M on $1,505M of pretax income), and the non-GAAP diluted share count of 231M was exactly as guided. Convertible dilution continues to shrink, from roughly 3M shares guided for June to 2M guided for September. This is as clean an EPS beat as this company has produced: it is operating income, not financial engineering.

Segment Performance

Revenue by Segment — FQ4 FY2026

Key Volume & Mix KPIs

Data Center — The Price Spread Widens Again

Data center exabytes grew 43% year-over-year while data center revenue grew 57%, a 14-point spread against the 8-point spread the same segment posted in March. Translated into the metric management actually tracks, revenue per exabyte rose roughly 10% year-over-year and 5.4% sequentially. That is not a rounding difference from the "mid-single-digit" strategy management has described for twelve consecutive quarters; it is roughly double it. Cloud remains the dominant buyer with three years of consecutive sequential exabyte growth, but the composition keeps broadening: enterprise OEM delivered strong double-digit year-over-year growth in both revenue and exabytes, enterprise nearline revenue rose for a fifth straight quarter, and management described active engagement with neocloud operators and model developers adopting tiered architectures.

"We shipped 195 EB into the data center market, up 11% sequentially and 43% year-over-year, with data center revenue coming in at $2.9 billion, up 17% sequentially and 57% year-over-year."
— Gianluca Romano, CFO

Assessment: The segment is now doing something HDD segments have not done in the industry's history: growing price and volume simultaneously, for a fifth consecutive quarter, with the price component accelerating. The analytical caution is that acceleration of this size is partly a scarcity rent rather than a contractual escalator, and management said as much. The segment test for the next two quarters is whether revenue per exabyte holds its new level when the incremental supply that cleared at spot is absorbed into the calendar 2027 contracted book.

Edge IoT — The Forgotten Segment Compounds

Edge IoT grew 14% sequentially and 20% year-over-year to $697M, its strongest showing of the fiscal year and a segment that spent most of the last three years as a source of drag and seasonality. The mechanism is not a demand recovery in client and consumer markets so much as a relative-price effect: NAND contract pricing has been rising far faster than hard drive pricing, and with HDD supply itself tight, marginal buyers are being pushed toward drives. The strategic point management continues to defer is the down-market HAMR play, using 4TB and 5TB-per-disk platforms to build cost-efficient lower-capacity products for enterprise and edge, which stays on the shelf while cloud absorbs every available head and platter.

Assessment: A segment growing 20% year-over-year on scarcity economics rather than product cycle is pleasant but not thesis-relevant on its own. What is thesis-relevant is that it is not consuming HAMR capacity while doing so, and that the deferred down-market platform remains a multi-billion-dollar revenue pool that converts precisely when cloud demand pauses. That is the closest thing to cycle insurance this business has, and it is still unspent.

Key Topics & Management Commentary

Overall Management Tone: Confident and notably less rhetorical than last quarter, which had carried the weight of a formal framework raise. This call substituted arithmetic for declaration: a fiscal 2027 growth statement that speaks for itself, a contracted horizon extended by a year, and a September guide that management let stand without embellishment. The one place the posture softened was capital returns, where the buyback commitment was restated in the same qualitative terms for a fourth consecutive quarter. Management was also more willing than usual to concede that current pricing exceeds the stated strategy, which is a form of candour that cuts both ways.

1. Fiscal 2027 Growth Will Outpace a 34% Year

The single most consequential sentence of the call was delivered in the CEO's second paragraph, without preamble, and it supersedes the "minimum of 20%" long-term target management raised only three months ago. Fiscal 2026 grew 34%. Fiscal 2027 is expected to grow faster than that. On a $12,195M base, the floor implied is above $16.3B, and the September quarter alone at $4.1B annualizes to $16.4B before any of the sequential growth management has separately committed to for each remaining quarter of the year.

"Given our momentum and the improved visibility we have into demand, we expect fiscal 2027 revenue growth to outpace our performance in fiscal 2026."
— Dave Mosley, Chair & CEO

Management did not formalize this as guidance, did not attach a dollar range, and did not repeat it in the CFO's prepared remarks. It was stated once, plainly, and left there. The three-month-old "minimum 20%" framework was not withdrawn and was not reconciled with it either.

Assessment: A company that raised its long-term growth floor to 20% in April and in July described the very next year as growing above 34% has either found a great deal more demand than it expected or has been deliberately conservative with its public framework. Both readings favor the equity; the second favors it more, because it implies the framework has further to travel. The reason this is not simply guidance is that management is protecting negotiating leverage into the calendar 2028 volume round, which is the same reason it declined to quantify fiscal 2027 pricing in April.

2. The Contracted Book Extends a Full Year, Into Calendar 2028

Every quarter of this coverage has added roughly a quarter of forward visibility. This one added a year. In April, nearline capacity was "almost fully allocated through calendar 2027" with fiscal 2027 build-to-order contracts being finalized. In July, the vast majority of nearline exabytes are allocated into calendar 2028, configuration and pricing are signed for the entirety of calendar 2027, and the conversations have moved out again.

"Based on the long-term supply agreements in place today, the vast majority of our nearline exabytes are now allocated into calendar 2028. Importantly, we are not seeing customers pull back on planning horizons. As our strategic relationships deepen, many are actively seeking to extend planning horizons through 2029 and beyond."
— Dave Mosley, Chair & CEO

The "not seeing customers pull back on planning horizons" clause is doing deliberate work. It is a direct response to the AI-capex-digestion fear that took 18% out of the stock in the five sessions before this print, and it is the only place on the call where management addressed the tape at all.

Assessment: The April report identified the calendar 2028 negotiation as the first genuine cyclical test of this cycle, and it has now moved from a future event to a book in hand. That does not eliminate the test; it relocates it to calendar 2029, which is far enough out to be unmodellable. The practical consequence is that the fiscal 2027 P&L is contracted and the fiscal 2028 P&L is substantially allocated, which is a visibility profile no storage company has ever presented, and it is the reason the gap between the fundamentals and the tape is currently as wide as it is.

3. Pricing Is Running Above the Stated Strategy, and Management Said Why

The most analytically important exchange of the call concerned the source of the pricing acceleration. Management has for twelve quarters described a value-based pricing strategy delivering mid-single-digit revenue-per-terabyte growth. The realized figure this quarter was roughly 10% year-over-year, and the September guide implies materially more. The CFO's answer was that the strategy is unchanged and the supply-demand gap is not.

"As I said before, it's not that we are changing our strategy, but for sure, the gap between supply and demand is now a little bit bigger than a few quarters ago. Our volume was little bit higher in fiscal Q4. We think can be maybe a little bit of output available in fiscal Q1. Of course, we are pricing that increased output at a very good price right now."
— Gianluca Romano, CFO

The CEO extended it into a mechanism, and in doing so gave the clearest description yet of how the above-contract layer actually forms: better-than-planned yields release exabytes two or three quarters after the contract is set, and those exabytes clear above the contracted price.

"We execute a little bit better. We have more exabytes to give, and we determine how hungry the market really is for those exabytes, and usually they'll pay more than that contract price, if you will, for those exabytes. That's why you see these step functions."
— Dave Mosley, Chair & CEO

Assessment: This is the quarter's most important disclosure and the one the bull case should be most careful with. Two layers now drive revenue per exabyte: a contracted escalator (durable, signed through calendar 2027) and a spot layer on yield-released supply (cyclical, priced to scarcity). The contracted layer justifies the multiple; the spot layer inflates the current print. We size the spot contribution at a few points of the 570 basis-point sequential margin step, which is not enough to change the direction but is enough that a demand pause would show up in margin before it showed up in revenue. That asymmetry is new this quarter and belongs in every model.

4. The September Guide: $4.1B, $7.30, and a 50% Operating Margin

The guide is the largest single-quarter step this company has ever put on the table: revenue of $4.1B plus or minus $100M (+13.0% sequentially, +56% year-over-year), non-GAAP opex of roughly $300M, non-GAAP operating margin of around 50%, and non-GAAP EPS of $7.30 plus or minus $0.20 on a 16% tax rate and 231M shares. The arithmetic those figures imply is a gross margin near 57.3%, a fourteenth consecutive record and 460 basis points above a quarter that itself beat by 300.

"Non-GAAP operating expenses are expected to be approximately $300 million. Based on the midpoint of our revenue guidance, non-GAAP operating margin is expected to be around 50%."
— Gianluca Romano, CFO

Against a Street sitting near $3.79B and $5.85 going into the print, the guide lands roughly 8% above on revenue and roughly 25% above on earnings. For scale on how fast this has moved: four quarters ago this company guided $2.5B and $2.30.

Assessment: A 50% operating margin on a hard drive company is a category error against every historical frame for this industry, and it is next quarter's midpoint. The guide also does something the April guide did not: it front-loads. Guiding 13% sequential growth into what has historically been a transitional quarter, on top of a quarter that already cleared its ceiling, leaves less room for the beat cadence that has driven four consecutive re-ratings. That is not a criticism of the number; it is a note that the bar management has set for itself is now genuinely high.

5. HAMR Hits Its First Stated Milestone, and Mozaic 3+ Finishes

Two roadmap items closed this quarter. Mozaic 3+ is now qualified and running in production across all major cloud customers, completing the qualification campaign that began five quarters ago and closing the final two customers management committed to in April. And HAMR reached roughly 40% of the nearline exabyte shipment run rate exiting the fiscal year, the first quantified milestone management had set.

"Now we have actually just achieved our first milestone, that was to achieve 40% of nearline exabyte sold on HAMR drive by June. We just did that. I'll say we are on track to achieve the future goals."
— Gianluca Romano, CFO

Mozaic 4+ shipped "pretty low" volume in the March quarter, ramped meaningfully in June, and is expected to contribute more in September. The next milestone is 50% of HAMR exabytes on Mozaic 4+ exiting calendar 2026, with Mozaic 5+ (5TB per disk, 50TB drives) still targeted for qualification shipments in late calendar 2027. Asked directly whether the 70%-of-nearline-exabytes-on-HAMR-by-June-2027 target still holds, the CEO said the company is on target for all of it.

Assessment: The platform risk that dominated this thesis two years ago is now essentially retired, and the milestone language is worth reading precisely. In April, Mozaic 4+ was expected to be "a majority of HAMR exabyte shipments exiting calendar 2026"; in July the same milestone is stated as "50%." Those are the same intent expressed at slightly different precision, and we flag it only because it is the fourth consecutive quarter in which HAMR yields have gone unquantified, which leaves the ramp's governor a matter of inference from the margin line rather than disclosure. A 57% implied gross margin is a strong indirect yield signal. It is still indirect.

6. The Balance Sheet Campaign Finishes, and Equity Turns Positive

Fiscal 2026 was the year the balance sheet stopped being a constraint. Gross debt fell $1,430M to $3,565M, net leverage reached 0.4x on $1,684M of quarterly adjusted EBITDA, cash reached $1,704M with $3B of total liquidity, and total shareholders' equity swung from a $(453)M deficit to $2,167M positive, a $2,620M move in twelve months. The pace accelerates from here: $1B of high-yield senior notes was already extinguished in July, the remaining convertible balance retires in September, and the CFO expects to exit the September quarter at $2.4B of debt.

"We ended our fiscal 2026 with $3.6 billion in debt, which is already a huge reduction from about $5 billion that we had at the beginning of the fiscal year. We will reduce debt even more during the quarter... but we will probably end fiscal Q1 at $2.4 billion in debt."
— Gianluca Romano, CFO

One note at "a fairly high interest rate" remains, to be addressed "maybe next quarter, maybe the following quarter." Other income and expense of $58M in June is guided down to roughly $45M in September purely on the lower balance, which is worth about $0.05 of quarterly EPS on its own.

Assessment: A company generating $1.1B of quarterly free cash flow against $2.4B of remaining debt has run out of balance sheet to fix. That is unambiguously good and it also removes the last excuse for the capital-returns question below. From fiscal 2027 the free cash flow has nowhere to go but shareholders, and how management deploys it becomes a first-order valuation input rather than a footnote.

7. Capital Returns: The Buyback That Keeps Not Arriving

This is the fourth consecutive quarter in which management has described a coming pivot to open-market repurchases in qualitative terms, and the fourth in which the realized number has been immaterial. Fiscal 2026 returned $810M to shareholders, of which $634M was the dividend and $176M was buybacks: 5.7% of the year's $3,105M of free cash flow spent on repurchase. The quarterly dividend was declared at $0.74, unchanged sequentially, against 0.4x net leverage and record cash generation.

"We are doing already more share buybacks than what we have done in the prior quarter. This quarter we are executing a higher level of share buyback, and we will continue in the next several quarters."
— Gianluca Romano, CFO

In April the sequencing was explicit: retire the remaining converts "this quarter or next," then "the majority will probably go to share buybacks." The converts have now slipped to September, and the buyback commitment remains a comparative rather than a number.

Assessment: This is the one place where management's execution has consistently lagged its language, and it is not a small one. At $769 the stock trades at roughly 24x our fiscal 2027 estimate after a 31.7% drawdown from the high, which is the most attractive repurchase window this company has seen in eighteen months, and the disclosed run-rate is a rounding error against the cash available. We are prepared to accept the debt-first sequencing through September. If the December quarter passes without a repurchase figure with a "B" in front of it, the revealed preference becomes information about how management views its own valuation, and it belongs in the bear column.

8. Unit Discipline Holds, and the Investment Line Gets Explained

The CFO's prepared remarks contained a new phrase, that the company is "strategically investing in additional tools and technology to support the manufacturing of our HAMR products," which drew an immediate question about whether unit discipline was loosening. The answer was the most technically specific version of the capacity-discipline argument management has given.

"If you look our last year, and if you look at the number of disk and the number of heads inside the box, they probably grew between 15% and 20%, and the units were absolutely flat."
— Gianluca Romano, CFO

The CEO added the part that actually explains the capex: the driver of factory complexity is not drive count or component count but the product transitions themselves, with 4TB-to-5TB-per-disk platforms requiring more time in the tools and sometimes multiple passes through them. Capital expenditure came in at 4.7% of revenue for fiscal 2026 and is guided "well within" the 4% to 6% range for fiscal 2027, which is the operative constraint.

Assessment: Component intensity rising 15% to 20% on flat units is exactly what an areal-density strategy looks like from the supply chain's vantage point, and it will keep generating the recurring bear datapoint that head and media suppliers are growing faster than Seagate says it is. The capex envelope is the tell that matters, and it did not move. A company adding capacity into a demand claim raises capex; this one is holding it at under 5% of a revenue base growing 34%.

9. The Demand Base Broadens: KV Cache, Neoclouds, and the Fifth Straight Enterprise Quarter

Each quarter of this coverage has added a demand layer: video in October, agentic workloads in January, physical AI and sovereign clouds in April. This quarter added technical substantiation rather than a new category. Management published a white paper with a memory manufacturer arguing that extending key-value cache across memory, SSD and hard drive tiers lets organizations retain more context and avoid recomputing generated data, which frees GPU capacity for revenue-generating work.

"When you set up these agents, you really need to give them context, and sometimes that's a very broad set of rules across your business or your problem set or whatever. As you do that, then you don't want to have to redo that context every time. You don't want to have to recompute all that context every time. That's what's driving storage, but still at very early days."
— Dave Mosley, Chair & CEO

Alongside it: enterprise nearline revenue up for a fifth consecutive quarter, strong double-digit year-over-year growth in enterprise OEM revenue and exabytes, and increasing engagement with neocloud operators and model developers that management was careful to frame as additive to hyperscaler demand rather than competitive with it.

Assessment: The KV-cache argument is the first demand mechanism in this cycle with a quantified economic payoff to the buyer rather than a storage-volume assertion: it frames hard drives as a way to reduce GPU consumption, which is the only budget line a hyperscaler cares more about than storage. Management was appropriately restrained ("still at very early days") and did not raise the exabyte growth framework on the back of it. The broadening matters most for cycle duration: a demand base spread across cloud, enterprise, neocloud and eventually physical AI is harder to pause simultaneously than one that rests on hyperscaler training capex.

10. The Gross Margin Question Nobody Would Answer

Two analysts, at different points in the call, tried to establish where gross margin tops out. One asked, half in jest, when the company reaches 80%. Another asked whether 60% is the right frame given the incremental margins being posted. The CFO declined both, twice, in nearly identical language.

"Our incremental gross margin has been very strong for the last several quarters. Overall gross margin is improving sequentially very well. We don't have a specific target. We will continue to improve based on the business situation. We know already that for the rest of the fiscal year, we will have sequential improvement every quarter."
— Gianluca Romano, CFO

What he did confirm, when pressed on the incremental figure, is that it runs "well above" 60%. Our calculation puts the realized sequential incremental gross margin at 87% this quarter, and the incremental operating margin at 87% as well, since opex was flat in dollars.

Assessment: The refusal to name a ceiling is the correct posture for a company still setting prices into calendar 2028, and the "sequential improvement every quarter for the rest of the fiscal year" commitment is the substantive content: it is a four-quarter margin commitment given without a number attached. The modelling consequence is that fiscal 2027 gross margin exits somewhere near 60%, a figure that would have been dismissed as absurd in every year of this company's public history before 2025.

Guidance & Outlook

The guide continues the shape of the four before it at a much larger scale: double-digit sequential revenue growth, several hundred basis points of margin expansion, and earnings growth (+27.8% sequentially) compounding on a record base. What is new is the year-level commitment sitting behind it. Fiscal 2027 revenue growth is expected to exceed fiscal 2026's 34%, sequential revenue growth and margin expansion are committed through the full fiscal year, and free cash flow is expected to improve throughout. Those three statements together constrain the model far more tightly than the September numbers alone.

Implied Q-over-Q ramp: The September guide alone annualizes to $16.4B. Layering the committed sequential growth across the remaining three quarters at a deliberately conservative 4% to 5% per quarter produces fiscal 2027 revenue near $17.5B, or roughly +43%, comfortably inside the "faster than 34%" commitment. On the earnings line, $7.30 in September with sequential margin expansion through the year and a declining share count builds to a fiscal 2027 figure near $32, against $15.58 just reported.

Street at: pre-print consensus had September near $3.79B and $5.85. The guide midpoints exceed both by roughly 8% and 25%, which forces another full-Street model rebuild, the fifth consecutive quarter in which that has been true. The three published post-print target raises we can see moved 5% to 8%, which is a fraction of the estimate move and leaves the sell-side deck stale in the other direction for once.

Guidance style: five consecutive quarters of results above the top of the guided range makes the pattern a property of the model rather than a run of luck: these are contracted floors on a sold-out book. The variance that matters is qualification timing and yield on the Mozaic 4+ ramp, and management's willingness to guide a 57% implied gross margin while that ramp scales is the strongest yield disclosure it has made without using the word.

Analyst Q&A Highlights

Whether the Implied 57% Gross Margin Is Really the Guide

The call opened with the question the guide invited: management guided operating margin and opex but not gross margin, so the first analyst simply did the arithmetic out loud and asked for confirmation, then pressed on whether the mid-teens annual cost-down per terabyte is sustainable through the Mozaic transitions. Management did not dispute the calculation and answered the cost question through the mechanics of the product transition rather than with a number.

Q: "I want to dig a little bit deeper into the gross margin. I guess given the guidance that you've outlined, it looks like your guide is implying a mid-57% or so gross margin into this next quarter. I guess my question is, one, is that kind of the guidance that you're providing? Two, how do you think about the cost down execution as we move through Mozaic 3+ to Mozaic 4+, you've been operating at a mid-teens kind of cost down per year on a per terabyte basis. Do you think that's sustainable, or how should we think about modeling that over the longer term?"
— Aaron Rakers, Wells Fargo

A: "The way we're thinking about these product transitions is, and I think you know this well, we have to actually put our factories on pause to go through the product transition a little bit. As we're moving product from 3 TB to 4 TB to 5 TB. Then there's yield issues, and as we out-execute our plan, what happens is we have opportunity for costs to drive better costs than we thought."
— Dave Mosley, Chair & CEO

Assessment: The 57% implied figure was allowed to stand unchallenged, which is as close to confirmation as an unguided metric gets. More useful is the cost framing: transitions cost yield and factory time up front, then over-deliver on cost when execution beats plan. That is the same asymmetry driving the above-contract pricing layer, and it means the margin surprises in this business are structurally skewed to the upside for as long as the transitions keep landing on schedule.

Why Pricing Should Not Accelerate Further From Here

The sharpest question of the call established that realized pricing has broken out of the framework management has described for three years, put the September implication at 20% or higher, and asked directly why the trend should not continue or steepen. The answer conceded the premise, which is the notable part.

Q: "For a number of quarters, you've been quite steadfast that price per exabyte growth would be this mid to high single digits year-over-year. You just reported 10% year-over-year price per exabyte growth in June. I think the September quarter guide implies pricing growth closer to maybe 20% year-over-year or even above that. Can you maybe just provide an update for us on how we should be thinking about pricing looking forward? Why this trend we're seeing in the September quarter shouldn't sustain or maybe even accelerate, just given supply-demand imbalance, customer demand strength, delivering more value to customers, et cetera."
— Erik Woodring, Morgan Stanley

A: "You are correct. As I said before, it's not that we are changing our strategy, but for sure, the gap between supply and demand is now a little bit bigger than a few quarters ago... Of course, we are pricing that increased output at a very good price right now. I'll say not really a change in our strategy, but a very good execution. Now with demand being particularly strong right now, we take a little bit more pricing benefit. Of course, every quarter is different."
— Gianluca Romano, CFO

Assessment: "You are correct" is a concession that realized pricing exceeds the stated strategy, and "every quarter is different" is a hedge placed carefully around it. Management is signalling that the current pricing level is a function of a supply-demand gap rather than a permanent repricing, while declining to say what happens when the gap narrows. That is the honest answer and it is also the single largest unhedged assumption in every bull model, including ours.

What Happens as Contracts Roll Off and Get Renegotiated

A follow-on line of questioning tried to separate like-for-like price increases from the benefit of newer, higher-capacity products, and to understand how renegotiation timing feeds the September step. The answer produced the clearest description of the above-contract mechanism management has given, and effectively disclosed that customers are paying above their own signed prices for incremental supply.

Q: "A follow-up on pricing. I'm curious if you could speak to like-for-like versus the benefit of newer products. Moreover, if you could speak to how we should be thinking about contracts rolling off, renegotiating of existing contracts, and how that is impacting the relative year-over-year pricing, particularly as it relates to that strong 20+ % number embedded in the September guide, and how to think about the moving parts into December and beyond."
— C.J. Muse, Cantor Fitzgerald

A: "What we're seeing over time is not only what we lock in for that period of a year, is we also see our ability to execute a little bit better, and that's usually two or three quarters out... We execute a little bit better. We have more exabytes to give, and we determine how hungry the market really is for those exabytes, and usually they'll pay more than that contract price, if you will, for those exabytes. That's why you see these step functions."
— Dave Mosley, Chair & CEO

Assessment: The most valuable disclosure of the call, and it cuts in two directions. It explains why prints keep clearing the guide ceiling (yield upside monetizes at spot two to three quarters after the contract is signed), and it establishes that a component of current margin is not contracted. Customers signing above their own contract prices is about as clean a scarcity signal as exists. It is also the first thing to disappear if hyperscaler ordering pauses, and it would take margin with it before it took revenue.

Whether Exabyte Growth Above 30% Can Persist Into Fiscal 2027

With realized exabyte growth running above 30% against a mid-20% framework, the question was whether the framework is simply conservative. Management declined to raise it, and the reasoning was about manufacturing physics rather than demand.

Q: "On the exabyte CAGR growth, I think you guys reiterated sort of this mid-20%. You guys obviously have been executing too much greater than that... could that sustain as you enter fiscal 2027, especially as you're migrating more towards your second-generation HAMR and then you're ramping further out into your Mozaic 5+? If you could just help us understand why exabyte CAGR could or could not sustain at this 30% as we look into fiscal 2027."
— Asiya Merchant, Citigroup

A: "We said before, we're not really increasing the box count. We are working really hard to get the heads and media inside the boxes to be able to go up and the technology capability to get exabytes out... We are going through product transitions... As you do that, there's a little bit of inefficiency in your factories. Long term, you actually get many, many more exabytes out as we go from Mozaic 3+ to Mozaic 4+ to Mozaic 5+... All these dynamics are how we actually have to predict the next few years, and that's one of the reasons we say mid-20s. Could we execute a little bit better? Conceivably, there's a lot of invention required still."
— Dave Mosley, Chair & CEO

Assessment: Holding the volume framework at mid-20s while raising the revenue growth expectation above 34% is an implicit statement that pricing carries the difference, and management has now made that gap explicit two quarters running without ever quantifying the pricing half. "There's a lot of invention required still" is also the most candid line about execution risk anyone offered on this call, and it belongs in the risk column rather than being read past.

How Much of Fiscal 2027 and 2028 Is Actually Locked, and the Preferential-Pricing Roll-Off

A two-part question sought a number for the locked-in book and probed whether the September pricing step is simply an artifact of an early HAMR customer's favorable pricing expiring. Management confirmed the roll-off is real but sized it as minor, which redirects the credit for the step back to underlying demand.

Q: "I was wondering if you could maybe just clarify on how much of your fiscal 2027, fiscal 2028 exabyte view is locked in via build to order versus maybe not under LTAs?... The pricing uplift in September, part of that is coming from one of your initial HAMR customers that had more favorable pricing rolling off. Should we still expect the price momentum to continue at those levels for the rest of the fiscal year?"
— Wamsi Mohan, Bank of America

A: "Relatively to that particular customer, the volume that was sold at the preferential price in June was minimal. September, we will not have any. There is a little bit of a positive impact from that, it is not the major reason why pricing is a little bit better in September than June. Of course, there's more overall demand and those are customers that are chasing a little bit more volume right now."
— Gianluca Romano, CFO

Assessment: A useful answer that closes a specific bear hypothesis: the September pricing step is not a one-time optical benefit from an expiring discount. Note what was not answered, though. The question asked for the percentage of fiscal 2027 and 2028 exabytes locked in, and the response addressed only the pricing sub-question. "Vast majority" has now stood in for a number across three consecutive calls, and the first cancellation test will be the only thing that reveals what it meant.

When the Buybacks Actually Start

The free-cash-flow question of the call paired a request for the manufacturing drivers of a higher FCF margin with a direct ask on the debt target and buyback timing. The debt half of the answer was precise and numeric; the buyback half was comparative.

Q: "I just wanted to ask a free cash flow question, if I could. Off of a 10-year high, can you talk about the dynamics that drive from just a manufacturing standpoint, a higher free cash flow margin in the future?... Just as a follow-up to that, just can you maybe talk about where, remind us where you want to get debt levels to and when buybacks could start?"
— Steven Fox, Fox Advisors

A: "We will probably end fiscal Q1 at $2.4 billion in debt. We still have one note that has a fairly high interest rate that I would like to address in the near future, maybe next quarter, maybe the following quarter. We are doing already more share buybacks than what we have done in the prior quarter. This quarter we are executing a higher level of share buyback, and we will continue in the next several quarters."
— Gianluca Romano, CFO

Assessment: The contrast in specificity is the finding. Debt gets a target, a timeline and a dollar figure; repurchases get "more than the prior quarter," which against a $176M full-year base is a low bar. Management has now had four quarters of record free cash flow and has chosen debt retirement every time, which was defensible at 1.8x leverage and is harder to defend at 0.4x with the stock 32% off its high. The December quarter is the deadline we set for this to become a number.

Where Neoclouds and Model Developers Fit in the Demand Stack

A question on the composition of demand beyond traditional hyperscalers asked management to parse the September outlook across hyperscale, neocloud and on-premise. The answer was qualitative and declined the parsing, but contained a genuinely new observation about why neocloud storage demand exists at all.

Q: "You spoke about qualifications on Mozaic among hyperscalers, how should we think about Seagate's growing exposure to NeoClouds and foundational model companies? I was hoping you could parse between demand from traditional hyperscale, NeoCloud, and maybe on-prem implicit in your September outlook."
— Karl Ackerman, BNP Paribas

A: "Two years ago, I would've said NeoCloud is probably largely compute based. We're starting to see that even some of the largest NeoClouds, they need a lot of data coming at them, and where did they get that data in the past? They might have got that from traditional hyperscalers, but there are some places where NeoClouds are saying, 'I need instances close to me.' By the way, I do not think that's necessarily competitive with the hyperscalers."
— Dave Mosley, Chair & CEO

Assessment: The observation that neoclouds are repatriating data locality rather than renting it from hyperscalers describes additive demand rather than a reallocation of the same exabytes, which is the distinction that determines whether the layer matters. The refusal to size it is consistent with every other customer-composition question this quarter. For now it stays a qualitative option in the thesis rather than a line in the model.

What They're NOT Saying

1. HAMR yields, for a fourth consecutive quarter: the ramp's governor has now gone unquantified across four calls while becoming more consequential each time. The 57% implied September gross margin is a powerful indirect signal and it remains indirect. This stays our top fundamental watch item precisely because inference has had to substitute for disclosure for a full year.
2. Any number attached to the locked-in book: "the vast majority of our nearline exabytes are now allocated into calendar 2028" was asked about directly and answered without a percentage, a dollar figure or an exabyte count, for the third consecutive quarter. The customers' commitments are the load-bearing element of the entire structural-growth case, and they are described exclusively in adjectives.
3. A gross margin ceiling, twice declined: "We don't have a specific target" is the correct answer for a company still setting calendar 2028 prices, and it also means the Street has no anchor for where the margin cycle tops. In a business whose historical gross margin was 20% to 25%, the absence of a ceiling is what makes the terminal-multiple debate unresolvable.
4. Customer concentration or composition in any quantified form: asked to parse September demand across hyperscale, neocloud and on-premise, management answered qualitatively. Cloud is "the vast majority" of a segment that is 81% of revenue, and the actual concentration has not been disclosed in any quarter of this coverage.
5. A dividend increase: declared flat at $0.74 for a third consecutive quarter against $3,105M of fiscal 2026 free cash flow, 0.4x net leverage, and a dividend consuming 20% of free cash flow. Not raising a dividend is a choice, and this management has now made it three times in the strongest cash-generation year in company history.
6. What happens to pricing when the supply-demand gap narrows: the CFO conceded that current pricing reflects a gap "a little bit bigger than a few quarters ago" and that "every quarter is different," then declined to describe the other side of that sentence. The above-contract layer has a size and management knows it; the market does not.
7. Any formal fiscal 2027 guidance despite an explicit growth commitment: management said fiscal 2027 revenue growth will exceed 34% and then offered no range, no dollar figure and no reconciliation to the "minimum 20%" framework raised three months earlier. Given five consecutive above-ceiling quarters the conservatism is tactical, and it means consensus spends another year chasing.
8. The 53rd week: fiscal 2026 contained an extra week, worth roughly two to three points of the headline 34% growth rate, and neither the release nor the call mentioned it. It does not change the quarter and it does slightly flatter the year that the fiscal 2027 commitment is being measured against.

Market Reaction

• Pre-print setup: STX closed July 28 at $747.30, up 171.4% year-to-date (S&P 500: +8.5%) and up 396.7% over the trailing twelve months, but down 17.0% over the trailing 30 days and 31.7% below its 52-week closing high of $1,094.04. The five sessions into the print took the stock from $913.36 to $747.30, a decline of 18.2%, including an 8.53% drop on print day itself on volume of 8.9M shares, roughly double the preceding sessions' pace. The stock entered the print 15.7% below its 50-day average of $886.31.
• After-hours move (July 28): the stock reversed immediately on the release and traded up as much as 10% during the evening session, settling in the mid-$780s as the 52.7% gross margin, the $7.30 September guide and the calendar 2028 allocation disclosure landed in sequence.
• Reaction session (July 29, in progress at publication): gapped open at $786.85 (+5.29%), printed $806.99 (+7.98%) intraday, then faded to $769.10 (+2.92%) by midday on 4.6M shares. The session has not closed and no closing level or full-day volume comparison is available.

The pre-print slide is the more informative half of this reaction. Nothing company-specific happened between July 23 and July 28; what moved was the market's willingness to pay for AI-infrastructure exposure, and Seagate was sold as a proxy alongside it. A stock that had compounded at 397% over twelve months gave back 18% in five sessions on a sentiment shift, then reported a quarter that beat its own guide ceiling on both lines and guided the next one 25% above consensus. That sequence is the cleanest illustration this coverage has produced of the difference between the business and the security.

The intraday fade from +8.0% to +2.9% is the third consecutive quarter in which an opening gap has been sold into, and it should be read the same way as the previous two: as a measurement of positioning rather than a verdict on the print. What is different this time is the starting point. The April fade happened from a record high after a 52% thirty-day run, which made it a straightforward profit-taking event. This one is happening 30% below the high, after a de-rating, into a guide that raised forward estimates by roughly a quarter. A fade from those conditions says the marginal seller is reducing sector exposure rather than taking a profit in this name, and that is a macro trade rather than a fundamental one.

The sell-side response so far has been modest relative to the estimate change: the target raises we can see moved 5% to 8%, against a September guide that moved forward earnings roughly 25%. The published objectives now sit 58% to 87% above the pre-print close, and the stale pre-print median sits roughly 34% above. For the first time in this coverage, the Street's numbers are the lagging indicator on the upside rather than the downside.

Street Perspective

Debate: Is the AI-Capex Digestion Fear Real, and Does Seagate's Contracted Book Actually Immunize It?

Bull view: The bull case is that the fear is generic and the exposure is specific. Seagate has configuration and pricing signed through calendar 2027, the vast majority of nearline exabytes allocated into calendar 2028, and customers asking to plan into 2029. Management stated explicitly that it is not seeing planning horizons pull back. A stock that fell 18% in five sessions on a fear its own order book has already contracted away is a mispricing, not a warning.

Bear view: The bear case is that build-to-order contracts have never been tested by a genuine hyperscaler pause, and "allocated" is a word, not a take-or-pay clause. The market is not questioning fiscal 2027; it is questioning what a storage supplier trades at when the AI capex cycle enters its first digestion phase, and the answer has historically been a much lower multiple regardless of what the backlog said. Every element of the last twelve months' return came from multiple expansion on top of estimates, and multiples do not honor contracts.

Our take: The bear has the better of the second half and the worse of the first. The contracted book genuinely does protect the fiscal 2027 P&L, and we do not think the earnings power is at risk on a twelve-month view. But the contracted book protects earnings, not the multiple, and the multiple is where all of the drawdown came from. That distinction is the reason this quarter's answer is to buy the estimate revision rather than to argue the stock cannot fall further. The point at which we would concede the bear case is not a price level; it is a hyperscaler capex guide that comes down, because that is the datapoint the contracts cannot argue with.

Debate: How Much of the 52.7% Gross Margin Is Contracted, and How Much Is Scarcity Rent?

Bull view: The bull view holds that the composition question is answered by the mix: Mozaic 4+ delivers more than 30% additional capacity on an unchanged bill of materials, opex is flat in dollars while revenue grows 17% sequentially, and the incremental margin of 87% is what a structurally repriced franchise produces. Pricing above contract is upside on top of that, not the basis of it, and the calendar 2027 book locks the base.

Bear view: The bear view is that the CFO conceded the supply-demand gap is doing the work, that customers are paying above their signed prices for incremental exabytes, and that this describes a shortage rent rather than pricing power. Shortage rents in storage have a perfect historical record of disappearing, usually within two quarters of the first supply response, and a margin that went from 42% to 53% in three quarters can retrace on the same timetable.

Our take: Both are partly right and the split is roughly quantifiable. The mix and cost story explains the trend line, which has run thirteen consecutive quarters through demand conditions that were not always this tight. The scarcity layer explains the slope of the last two quarters, and we size it at a few points of the 570 basis-point sequential step. The practical consequence for the model is that we underwrite the contracted base and treat the spot layer as unreliable: our fiscal 2027 estimate assumes margin expansion continues but at roughly half the recent pace. A model that extrapolates the last two quarters' slope is not being bullish, it is being careless.

Debate: What Is the Right Multiple for a Business With Two Years of Contracted Visibility?

Bull view: The bull argument is that the market is still applying a hardware-cycle frame to a business that no longer behaves like one: two years of signed configuration and pricing, 30%+ free cash flow margins, a net-cash balance sheet within two quarters, flat unit capacity, and a competitor a technology generation behind. Contracted infrastructure businesses with those characteristics trade at 30x and above, and at 24x the stock is being valued on the fear rather than the book.

Bear view: The bear argument is that the multiple is doing exactly what it should. A 24x forward multiple on peak-cycle margins is not cheap; it is a normal multiple on an abnormal earnings base, and the correct comparison is to normalized rather than guided earnings. Gross margins that expanded from 36% to 53% in eight quarters define the numerator's fragility, and no amount of contract duration changes what happens to a cyclical's multiple when the cycle is questioned.

Our take: The honest position is that both the multiple and the earnings base are higher-quality than they were a quarter ago, and the stock is 20% lower. We hold our multiple at 26x to 30x, deliberately below the 29x to 32x we applied in April, for two reasons: the estimate we are capitalizing is now the current fiscal year rather than the forward one, and the digestion risk the tape is pricing deserves a discount even though we do not think it hits fiscal 2027 earnings. That combination of a lower multiple on a much higher number is what produces a wider fair-value spread than we have had at any point in this coverage, and it is the entire basis for keeping the rating.

Model Update Needed

Valuation impact: Our fair-value range moves from $640–700 to $830–960, on a fiscal 2027 estimate raised from roughly $22 to roughly $32 and a multiple cut from 29–32x to 26–30x. Against the $769.10 midday price, the range midpoint of roughly $895 implies about +16%, and the low end still implies about +8%. That is the widest spread this coverage has carried since the January 2026 report, and it was produced by the estimate and the price moving in opposite directions in the same quarter rather than by any change in how generously we are willing to value the business. We note plainly that our range sits well below the $1,180 to $1,400 objectives published since the print; we are capitalizing the contracted fiscal 2027 path and declining to capitalize the calendar 2028 allocation until its pricing is signed.

Thesis Scorecard Post-Earnings

Overall: Thesis strengthened, and for the first time in this coverage the strengthening came with the stock lower rather than higher. Five of the six pillars moved in the right direction on evidence; the sixth, valuation and positioning, improved by being partially resolved rather than by being argued away. The bull case now rests on a book that extends into calendar 2028 and a fiscal 2027 P&L that is contracted; the residual risk has migrated from the business to the multiple, which is a better place for it to sit but not a harmless one.

Action: Maintain Outperform; fair-value range $830–960 against $769.10. Signposts for the September quarter: the roughly 57% implied gross margin print, the first quarter with a repurchase figure large enough to matter against $1.1B of quarterly free cash flow, the $2.4B exiting debt target, Mozaic 4+ progress toward 50% of HAMR exabytes exiting calendar 2026, and any first quantification of HAMR yields. The rating's exit conditions are revised to fit the new setup: a hyperscaler capital expenditure guide that comes down, a retreat from the fiscal 2027 sequential commitment, evidence that the above-contract pricing layer is reversing faster than the contracted base is growing, or a fourth consecutive quarter without a material buyback. Any one of those moves us to Hold. A price recovery toward the upper end of the fair-value range does the same, on discipline rather than doubt.