Price Carried a Record Quarter While Exabytes Stalled, and a Supply Scare Had Already Taken 29% Off the Stock: Maintaining Outperform

Results vs. Consensus

FQ1 FY2027 Scorecard

Year-Over-Year Comparisons

A calendar note that neither the release nor the call raised: the year-ago quarter was a 14-week period (the extra week of the 53-week fiscal 2026), while this quarter was 13 weeks. In July 2025 management said its build-to-order customers did not treat that quarter as 14 weeks, and supply under those contracts is set by agreement rather than by days, so we do not adjust the growth rates above. If the extra week did add volume, the per-week rates would be higher, at most about 73% for revenue and about 30% for exabytes. The sequential comparisons below are week-for-week, and so is the December guide.

Quarter-Over-Quarter Comparisons

Quality of Beat

Revenue. The $4,233M print cleared the guide ceiling by $33M, the fourth consecutive quarter above the top of the range, but it got there by a different route from the last three. Volume contributed almost nothing: total exabytes rose 0.9% and data center exabytes 1.5%. Data center revenue still rose 17.6% sequentially, to $3,449M, and edge IoT rose 12.5%, to $784M. Management said in its prepared remarks that it was staging Mozaic 4 drives ahead of customer qualifications and that quarterly exabyte growth "may vary," so the pause was signalled rather than hidden. The quality question is not whether the revenue is real (it is contracted and shipped) but whether a quarter in which price carries almost all of the growth is repeatable. Management's answer is that December brings both: a meaningful exabyte step-up and further pricing.

Margins. The 510 basis points of sequential gross margin expansion compares with 570 in June, and the composition changed. In June, cost per exabyte fell and price rose. In September, our calculation shows non-GAAP cost per exabyte rising about 3.1% sequentially (from $7.88M to $8.12M) and falling only 6.2% year-over-year, against roughly 15% annual declines over the prior four quarters, as an analyst described them on the call without challenge. Management gave two reasons. First, externally sourced components cost much more than a year ago. Second, the Mozaic transitions temporarily raise scrap and lower yields. Price rose about 16% per exabyte, which more than covered both. The margin record is genuine, but it now rests entirely on pricing, and part of that pricing is a pass-through of component inflation rather than value capture.

EPS. The $7.68 beat the guide midpoint by $0.38, and about $0.35 of that came from operating income: non-GAAP operating income of $2,147M compared with roughly $2,050M implied by a 50% margin on $4.1B. Below the line helped slightly. Other income and expense was $38M against roughly $45M guided, and the effective non-GAAP tax rate was 15.6% against 16%. A diluted share count of 232M, 1M above the guide, took back about $0.03. The $0.09 gap between GAAP EPS of $7.59 and non-GAAP EPS is a net figure: share-based compensation ($0.33) and the $44M loss on early debt redemption ($0.19) are added back, and a $0.46 tax adjustment offsets most of them. The beat is operating, not financial.

Segment Performance

Revenue by Market — FQ1 FY2027

Data center revenue is the 8-K total less the $784M edge IoT figure given on the call; management rounded it to "$3.4 billion." Year-over-year rates compare a 13-week quarter with a 14-week quarter.

Key Volume, Price & Cost KPIs

Data Center — A Price Quarter

Data center revenue rose 17.6% sequentially on 1.5% more exabytes, so revenue per data center exabyte rose 15.9% in one quarter, to $17.42M, and 31.0% year-over-year. For scale, in July we described a roughly 10% year-over-year rise (5.4% sequentially) as about double the company's stated "mid-single-digit" strategy. This quarter's sequential step alone is about three times July's. The breadth beneath the cloud business also kept improving: cloud nearline posted a twelfth consecutive quarter of revenue growth, and enterprise and OEM data center revenue rose for a sixth consecutive quarter with double-digit growth both sequentially and year-over-year, reaching its second-highest level in more than a decade.

"Data center revenue was $3.4 billion, up 18% sequentially and 63% year-over-year and representing 81% of total revenue. Data center shipments totaled 198 exabytes, up 24% year-over-year and represented 90% of the company's overall exabyte volume."
— Gianluca Romano, CFO

Assessment: The segment delivered the quarter's whole result on price, and management described the mechanism plainly: contracts renewing at higher prices, plus external component costs passed through. That is a better-quality explanation than July's spot-clearing mechanism, because it is contracted. It is also a reminder of how much now rests on price, which supplied about nine-tenths of the segment's 17.6% sequential growth. If December's exabyte step-up arrives with flat pricing, revenue still grows, but if pricing pauses while volume also lags, the beat cadence ends. The December quarter should show volume doing more of the work.

Edge IoT — Still Growing on Scarcity

Edge IoT revenue rose 12.5% sequentially and 52.2% year-over-year (on a 14-week comparison) to $784M, the second consecutive double-digit sequential quarter for a segment that spent most of the last three years as a source of drag. Management did not explain the driver this quarter. The mechanism we described in July is the likely one: hard drive supply is tight and flash prices have been rising faster than drive prices, which pushes marginal client, consumer and network storage demand toward drives. Management did not discuss moving HAMR platforms into this market this quarter, and it expects cloud customers to keep taking the vast majority of output as Mozaic 4 ramps.

"Complementing the strength in our data center business, our edge IoT markets also delivered solid revenue growth. Revenue totaled $784 million representing 19% of total revenue. This was an increase of 13% sequentially and 52% year-over-year."
— Gianluca Romano, CFO

Assessment: A $3B annualized segment growing double digits sequentially without consuming HAMR capacity is a useful cushion, and it shows the pricing power is not confined to hyperscale contracts. It is also the part of the business most exposed to the memory cycle reversing if, as we believe, its growth depends on flash staying expensive. We treat it as a scarcity beneficiary, not a growth pillar.

Key Topics & Management Commentary

Overall Management Tone: Confident on the bottom line and on demand, but more defensive on the composition of the quarter than at any point in the past year. Most of the Q&A pressed on why exabytes stalled, why cost per terabyte rose and how much of the price step is durable. Management answered with process and mechanism rather than numbers, and when pressed on cost it steered back to operating income and free cash flow. The capital-return message was the most concrete it has been in five quarters. On competitor capacity, management declined to engage.

1. Price Did Almost All of the Work: Revenue per Exabyte Rose About 16% in One Quarter

The quarter's defining number is not the margin record but the split of the revenue step. Revenue rose $604M, or 16.6%, on 0.9% more exabytes. Blended revenue per exabyte rose 15.6% sequentially, to $19.24M, and data center revenue per exabyte rose 15.9%. One analyst on the call put it as an acceleration from "mid-single digits Q-on-Q to something like mid-teens," and management did not dispute the figure.

"Pricing continues to improve, reflecting both the execution of our value-based pricing framework and higher external component costs. We expect a stronger demand environment evident in customers' multiyear commitments to continue to support these healthy pricing dynamics."
— Dave Mosley, Chair & CEO

The new element is the second half of the first sentence. On the earlier calls of this coverage, management attributed pricing to its value-based framework and to the balance of supply and demand. This quarter it named component inflation as a co-driver, and the CFO made the pass-through explicit: when the company renegotiates a contract, it now builds current external supply-chain costs into the new price.

Assessment: Pricing of this size from a hard drive company is without precedent in this coverage, and the explanation is now two-part. The value-based part compounds and is what the multiple should capitalize. The pass-through part protects margin but does not add to it in a lasting way, and it would reverse if component costs fall. We cannot split the two, and management would not, so we underwrite a slowing price curve rather than extrapolate this quarter.

2. The Exabyte Pause: Staging, Not Shortfall

Exabytes shipped rose 0.9% sequentially and 20.9% year-over-year, a sharp-looking deceleration from 33.7% in June that several post-print notes cited as the reason for cutting price targets. Two things matter. First, management flagged the pause in its prepared remarks and tied it to staging Mozaic 4 inventory ahead of customer qualifications, with a "meaningful step-up" promised for December. Second, the year-ago quarter had an extra week. Management said in July 2025 that its build-to-order customers did not treat that quarter as 14 weeks, so the extra week may have added little supply. If it did add volume, per-week growth would be as high as about 30%, but we do not rely on that adjustment.

"In the September quarter, we shipped a record 198 exabytes nearline capacity into data center customers, and we expect a meaningful step-up in the December quarter, supported by the continued ramp of Mozaic 4. As we stage new generations of Mozaic products and ship to a broader customer base, quarterly exabyte growth may vary."
— Dave Mosley, Chair & CEO

The balance sheet supports the staging account: inventory rose $129M to $1,700M. Asked directly whether yield problems sat behind the pause, the CEO answered: "Yes, there's none." The long-run framework of mid-20% compound exabyte supply growth was restated without change.

Assessment: We accept the staging explanation for one quarter, because it was disclosed before it was questioned and the inventory build is consistent with it. It becomes a problem only if December does not deliver. A December quarter that ships fewer than about 235 EB (roughly 7% sequential growth) would turn a timing story into a capacity story, and capacity is the one variable this thesis cannot absorb.

3. Cost per Exabyte Rose: External Components and Transition Scrap

Our calculation puts non-GAAP cost per exabyte at $8.12M, up 3.1% from $7.88M in June and down only 6.2% year-over-year. An analyst on the call described the annual cost decline as slowing from about 15% to "mid-single digits," and management did not dispute it. Management gave two causes. The first is external: components the company buys rather than makes cost much more than a year ago. The second is internal and temporary: each Mozaic transition brings early scrap and lower yields that take "a quarter or 2" to recover.

"The cost of the external component, as you know, compared to a year ago, is now much higher. And therefore, there is at least a partial offset of our internal cost reduction from a higher external component cost. So we have to see where the external component costs will go in the future."
— Gianluca Romano, CFO

Management was firm that its own cost curve is intact: media cost is "continuing to be very, very strong," and moving from 4TB to 5TB per disk carries no significant cost increase in the HAMR technology itself. Asked to quantify the external component impact on gross margin, the CFO said only that it is included in the December guidance.

Assessment: This is the most important new disclosure of the quarter for the margin debate. Cost per exabyte rising in the same quarter that price per exabyte jumps 16% means the margin is now a pure pricing story in the near term. That is fine while the supply-demand gap is widening, which management said it probably is. It also means a quarter in which pricing merely holds would show margin flat or down, which the Street's models do not allow for.

4. Pricing Re-Attributed: Contract Renewals and Pass-Through, Not Spot

In July, the CEO described part of the pricing step as yield-released exabytes clearing above contract prices, and we flagged that spot layer as the least durable part of the margin. This quarter, asked how much of the price was contracted versus floating, he put the emphasis elsewhere: most of the step came from renewing contracts at higher prices, not from extra exabytes sold at spot.

"So when we negotiate a new contract, of course, we also consider the current situation for the external supply chain costs and adjust the pricing according to that. But apart from the change, the strategy is exactly the same that we have followed for the last 3 years at this point."
— Gianluca Romano, CFO

The hyperscale contracts roll on roughly four-quarter cycles, and other customer contracts on shorter ones, so a steady stream of renewals is re-pricing the book each quarter. Management also described price per terabyte as "fairly well blended across the entire portfolio" rather than concentrated in one customer type.

Assessment: This improves the quality of the price step relative to our July concern, and we revise the caveat on our margin pillar accordingly. A price set in a renewed contract persists for that contract's term. What it does not do is answer the forward question: the calendar 2028 renewals are now the price that matters, and they will be negotiated while competitors are adding capacity.

5. The December Guide: $4.8B, $9.20 and an Implied ~60% Gross Margin

The December guide calls for revenue of $4.8B plus or minus $100M (+13.4% sequentially and about +70% year-over-year on comparable 13-week quarters), non-GAAP opex of about $300M, operating margin expanding "towards the mid-50% range," and EPS of $9.20 plus or minus $0.20 on a 16% tax rate and 233M shares. Working back from the EPS midpoint gives an operating margin near 53.7% and a gross margin near 59.9%, a fifteenth consecutive record.

"Non-GAAP operating expenses are expected to remain flat at approximately $300 million. Based on the midpoint of our revenue guidance, non-GAAP operating margin is expected to expand towards the mid-50% range."
— Gianluca Romano, CFO

Against pre-print consensus near $4.55B and $8.60, the guide sits 5.6% above on revenue and 7.0% above on EPS. In July the same comparison was roughly 8% and 25%. Management also tightened its fiscal-year language: revenue growth is now expected to "meaningfully outpace" fiscal 2026's 34%, where July said "outpace."

Assessment: The guide is excellent in absolute terms and modest relative to expectations, because the Street had already moved most of the way. That gap between absolute quality and relative surprise is exactly how a stock de-rates through a beat-and-raise. The mechanics also matter: a 13% sequential step built on an exabyte recovery plus continued pricing gives December two engines rather than one, which is a better setup than September had.

6. Visibility Extends to 2030; Calendar 2028 Pricing Is the Next Negotiation

The contracted horizon lengthened again. In July, the vast majority of nearline exabytes were allocated into calendar 2028. In October, the same is true, and some cloud agreements now run through calendar 2030. Configuration and pricing are set through calendar 2027, and the company expects to begin finalizing calendar 2028 terms over the next few quarters.

"The vast majority of our nearline exabyte supply is now allocated under long-term supply agreements through calendar 2028 with some cloud customer agreements extending through calendar 2030. Our strategic planning discussions extend even further."
— Dave Mosley, Chair & CEO

Asked what distinguishes the 2030 agreements, the CEO said the process is unchanged: visibility on exabytes first, then configuration and pricing locked about 12 months out. He added that the company could "pull back" quickly if a customer were not taking supply according to plan, and that it sees no evidence of that.

Assessment: The 2030 agreements extend the planning horizon, not the pricing horizon, and that distinction should temper how much weight they carry. The commercially important event is the calendar 2028 price-setting round that begins in the next few quarters. It is the first renewal round since reports that a competitor plans to nearly double a drive plant's capacity, and the outcome will tell us more about the durability of this cycle than any allocation statement.

7. Mozaic 4 Breadth Unchanged, Upside to Mozaic 5 Timing, 8TB per Disk in the Lab

The roadmap news was mixed in timing and strong in direction. Mozaic 4 (up to 44TB) is still ramping at two leading cloud customers, the same count as in July, with additional qualifications across multiple global CSPs and enterprise OEMs expected "over the next couple of months." The HAMR milestone stands: roughly 50% of the nearline exabyte run rate exiting calendar 2026, with Mozaic 4 about half of that, and the fiscal 2027 target (70% of nearline exabytes on HAMR, per July) was restated without change.

"The 5-terabyte is that we're actually more optimistic than we were last quarter, I'd say we see the ability to make this transition a little bit faster than what we had originally planned, and we'll get to talking to that when we get into fiscal '28."
— Dave Mosley, Chair & CEO

Mozaic 5 (50TB-plus on the same 10-disk platform) remains targeted for customer qualification shipments toward the end of calendar 2027, now with upside to that timing, and the CEO sees 5TB per platter "extending fairly rapidly into 6 terabyte per platter." The lab demonstrated 8TB per disk, up from 7TB in January. Microsoft has also qualified a joint Seagate and Broadcom solution that pairs Mozaic HAMR drives with shingled recording, which adds capacity on top of HAMR for customers willing to change their software.

Assessment: The direction of the roadmap is better than it was 90 days ago, and the possible Mozaic 5 acceleration matters more than any other technical item because it is the next step down the cost curve. The timing of Mozaic 4 breadth is a mild disappointment: no new qualification closed in the quarter, and those qualifications are now what the December volume step depends on. HAMR yields went unquantified for a fifth consecutive quarter. Management has discussed transition yields and scrap on every call since October 2025, but this was the first quarter in which they showed up as a sequential rise in cost per exabyte.

8. Capital Returns: The Buyback Arrives, With a 75% Framework

This was the quarter's clearest delivery against a commitment we had graded harshly. Seagate repurchased $400M of stock in the September quarter, against $176M in the whole of fiscal 2026, and returned $568M in total with the dividend. The dividend rises about 3% to $0.76. The CEO restated the company's model of returning 75% of free cash flow, a long-standing target it last restated a year ago, and the CFO said repurchases will step up from the March quarter now that debt work is done.

"We exited the September quarter with $4.4 billion remaining on our existing share repurchase authorization and expect to significantly increase share repurchases starting calendar 2027."
— Gianluca Romano, CFO

At 75% of a free-cash-flow run rate near $1.5B and rising, the framework implies more than $1B per quarter of total returns from March, or roughly $900M to $1B of repurchases after the dividend. The diluted share count is guided up 1M to 233M for December, so the program has not yet started to shrink the count.

Assessment: In July we said a December quarter without a repurchase figure "with a B in front of it" would put capital allocation in the bear column. September delivered $400M, and management itself named the March quarter for the step-up. We move the deadline to March and hold management to the 75% figure. With the stock 36% below its high, this is the right time to buy, and with the debt work finished, the 75% model now has a start date that can be graded.

9. The Balance Sheet Is Finished

Seagate retired $1.2B of debt in the quarter, including the last convertible notes and about $1B of high-yield bonds, and a further $750M of high-yield notes after quarter-end. Principal debt is now about $1.6B, down from about $5B at the start of fiscal 2026. Net leverage was 0.2x on $6,094M of trailing adjusted EBITDA, and the company expects to be net cash positive exiting calendar 2026. Management said it has "regained investment-grade status."

"Following the close of the September quarter, we retired an additional $750 million of high-yield senior notes reducing principal debt to approximately $1.6 billion. As a result of our actions, we expect to be net cash positive exiting this calendar year."
— Gianluca Romano, CFO

Other income and expense falls to about $25M in December from $38M, worth about $0.05 of quarterly EPS. No further debt matures until June 2029, according to an analyst on the call, a point management did not contest.

Assessment: The CFO said the October redemption was the last debt activity planned. With no debt left to fix, every dollar of free cash flow now has to go to shareholders or sit as cash, which is why the 75% framework is credible.

10. The Supply Response: Toshiba, Peer Capex and a 4% to 6% Answer

The stock's October decline began on October 2, when reports said Toshiba would invest about $380M to nearly double the capacity of its Philippine drive plant by its fiscal 2027, as part of a goal to lift its share from about 11% toward 30%. On the call, one analyst noted that peers had talked about capital intensity rising toward 8% of revenue and asked whether Seagate's 4% to 6% risks share loss. Management declined to discuss competitors and defended its own envelope.

"From our perspective, demand remains really strong and from a supply side, we continue to believe that bringing meaningful HDD capacity to market is a really complex process that requires substantial investment in R&D and technology. It's an intricate supply chain balance. And the idea of just plugging in a few machines, I mean it's orders of magnitude more difficult than that."
— Dave Mosley, Chair & CEO

Management also said it cannot acquire HAMR capability and must build it, that the latest generation of tools is more capable than the tools it replaces, and that 4% to 6% of revenue is "plenty" for its plan. It said it has no technical reason to avoid adding platters beyond 10, but prefers one platform for simpler customer qualifications. Western Digital, reporting five days earlier, printed a 58.1% non-GAAP gross margin for its quarter ended the same day, above Seagate's 57.8%, and said its capex would run somewhat above its 4% to 6% long-run range for the next two years.

Assessment: The supply response is no longer hypothetical, and we move our competitive bear point to Emerging. Two facts limit the damage. First, even if Toshiba doubled all of its capacity, it would add about 11 points of the 2025 industry base over two years, against roughly 56% of demand growth over the same period at a mid-20s compound rate. Second, Toshiba's added capacity is not HAMR. The risk is not that prices fall; it is that the gap between supply and demand stops widening, which slows the price curve that is now carrying the whole margin story. The peer's gross margin matching Seagate's also says that much of today's margin is an industry shortage outcome rather than a HAMR premium.

11. The Demand Narrative: Persistent Data, Agents and an Enterprise Upturn

Management added one framing this quarter: AI is moving from training toward inference and agents, which makes persistent data more valuable and extends storage demand beyond the initial build-out. It cited a hyperscale customer's observation that its own customers keep finding new uses for stored data, and described KV-cache tiering as a mechanism that moves data between compute and storage layers more efficiently than ever. Physical AI is "not yet directly" a driver.

"As AI adoption expands beyond model training and toward inference and agentic applications, persistent data becomes increasingly important. These systems rely on vast amounts of historical data to make informed decisions and take effective actions while constantly generating new output that must be governed, retained and reused."
— Dave Mosley, Chair & CEO

On demand quality, management said it sees no product swaps or deferrals of the kind seen in past cycles, and that output above plan is spoken for.

"We're seeing none of that right now. Absolutely. Everything we can make above and beyond our plans is being spoken for."
— Dave Mosley, Chair & CEO

Asked separately whether hyperscaler power, land or shell constraints were delaying purchases, the CEO said no.

Assessment: The demand evidence is the strongest part of the call and the least contested: sixth consecutive quarter of enterprise growth, twelfth of cloud nearline growth, and a CFO statement that the supply-demand gap is "probably growing." The market is not doubting current demand; it is doubting the supply side and the durability of price. The demand narrative answers neither.

Guidance & Outlook

The guide keeps the shape of the last two: double-digit sequential revenue growth, a gross margin record and operating expenses roughly flat in dollars. What changes is the mix of drivers. September was almost all price. December, on management's account, adds a meaningful exabyte step-up from Mozaic 4 shipments to newly qualified customers, plus further pricing as contracts renew. The fiscal-year statement also hardened, from growth that will "outpace" fiscal 2026's 34% to growth that will "meaningfully outpace" it.

Implied Q-over-Q ramp: September plus the December midpoint is $9.0B for the first half. With the committed sequential growth continuing at a deliberately slowing 8% in March and 7% in June, fiscal 2027 revenue reaches about $19.8B, or about +62%, and EPS about $38.80. A flat second half at the December rate would still give $18.6B (+53%), so the "meaningfully outpace" language is safe on almost any path.

Street at: pre-print consensus for December sat near $4.55B and $8.60; the guide sits 5.6% and 7.0% above. The post-print fiscal 2027 consensus has moved to about $20.1B and $40, slightly above our build, and fiscal 2028 consensus implies roughly 40% further revenue growth. We are below the Street on both years by design: we do not underwrite pricing on calendar 2028 terms that have not been set.

Guidance style: Four consecutive quarters above the ceiling of both ranges make the conservatism systematic. This quarter's beat was the smallest of the streak in relative terms, though, and the December guide's premium to consensus is much narrower than July's. The guide is no longer a source of large upside surprises because the Street now models close to it. Upside from here has to come from realized results, not from the guide.

Analyst Q&A Highlights

Whether the Inventory Build Was Staging for December

The call opened on the volume pause. The question linked flat exabytes to a rise in inventory days and asked whether the company had pre-built product for December, then asked whether calendar 2027 exabyte growth would run above the mid-20s framework. Management confirmed the staging and tied it to specific qualifications due in the current quarter, and held the 25% long-run number.

Q: "I guess first question regarding exabyte deceleration in the September quarter. I'm curious, given your inventory days moved up to 97 from 83, was there kind of staging involved prebuilding that's supporting your outlook for the accelerated exabytes into December?"
— C.J. Muse, Cantor Fitzgerald

A: "Yes, C.J., we have a few customers that will qualify Mozaic 4 in the next quarter. And therefore, we are preparing manufacturing for those quals and the volume that we can ship to them in the December quarter and later."
— Gianluca Romano, CFO

Assessment: The answer converts the pause into a dated, checkable claim: several Mozaic 4 qualifications in the December quarter, with shipments behind them. It also concentrates risk. The December volume step now depends on customer qualification timing that Seagate does not fully control, and the inventory is built ahead of it.

Whether a Yield Problem Sits Behind the Pause

A follow-up asked management to rule out yield issues or customer data-center readiness as the cause of the slower exabyte growth. The answer was an unqualified no. Later in the call, however, a question on cost drew a more nuanced account in which transition yields were below target and recovering.

Q: "I just want to make sure there's no kind of right tail risk here in terms of any yield issues that we need to be thinking about anything in terms of data center readiness."
— Erik Woodring, Morgan Stanley

A: "Yes, there's none. The big step up in the June quarter, it tapered off a little bit in the September quarter, just timing and then it will be another big step-up in the December quarter."
— Dave Mosley, Chair & CEO

Assessment: The two answers are reconcilable: yields did not cause the volume pause, but they did raise cost. Read together, they say the Mozaic 4 transition is running below target yields for now, and that this quarter the shortfall reached the cost line. That makes the long-requested yield disclosure more important, not less.

Why Cost per Terabyte Went Up

The sharpest question on composition set out the apparent acceleration of price from mid-single digits to mid-teens sequentially, noted that cost appeared to rise in the same quarter, and asked whether anything temporary in mix or cost was distorting the picture. Management first pointed to operating income and free cash flow, then described scrap and lost capacity during transitions.

Q: "But when I look at the price, it seems to be accelerating significantly from what was something like mid-single digits Q-on-Q to something like mid-teens Q-on-Q in the September quarter? And meanwhile, your cost, still slightly confused about that, appears to be actually up Q-on-Q, you did point out about some of the component costs coming up."
— Mark Newman, Bernstein

A: "As we drive these hard product transitions, we have to incur some of that scrap. And we actually incur a little bit of manufacturing, I'll call it, [lossiness] when that happens as well. The yields on the product transition may not be where we want. We go take a quarter or 2, and we go beat those back into what the old yields were from the -- so there's a little lost capacity in that."
— Dave Mosley, Chair & CEO

Assessment: The most candid answer of the call on manufacturing, and the one that best explains both the cost line and the flat exabytes: lost capacity during a transition means fewer exabytes from the same tools. Management framed it as a recurring, recoverable feature of each transition. The test is whether cost per exabyte falls again in December as Mozaic 4 matures.

How Much of the Price Step Was Contracted

A question asked for the split between price already contracted and price floating, by customer type or capacity point. Management gave no percentages but gave a clear direction: most of the step came from contract renewals, not from extra supply sold above contract. The CFO added that new contracts now include current external component costs.

Q: "How much of this would you say was achieved with already contracted pricing versus what's floating, any relative percentages on there?"
— Wamsi Mohan, Bank of America

A: "Yes. I think most of it is on the, I'll call it, the rolling re-upping of the contracts, Wamsi. It's not because we're getting more exabytes out or we're pushing more exabytes out. We're running the business very predictably on that front."
— Dave Mosley, Chair & CEO

Assessment: This shifts the emphasis from July's answer, which highlighted yield-released supply sold above contract. A contracted price is more durable than a spot price, so the answer improves the quality of the margin. It also means each quarter's pricing depends on which contracts renew, which makes the price curve lumpier than a smooth escalator.

Whether Mid-Teens Quarterly Price Growth Can Continue

The follow-up asked directly whether the market should model mid-teens sequential price growth or treat September as a one-off. The CEO said he does not see the trend changing much. The CFO added that every quarter is different and that December's guide implies more gross margin expansion, now driven mainly by price.

Q: "Given all that and the huge price increase you've been getting recently, I mean, should we expect this type of quarterly ASP growth to continue in the mid-teens range? Or is that more of a one-off?"
— Mark Newman, Bernstein

A: "As we talked about on supply, there will be step-ups as new contracts come up all the time. I said in the prepared remarks, that we don't see any demand destruction at all. If anything, we see a lot more demand creation kind of quarter-over-quarter... And I don't see the trend changing too much right now."
— Dave Mosley, Chair & CEO

Assessment: "Step-ups as new contracts come up" describes a staircase, not a slope, and it is the right mental model for the next four quarters. We do not model mid-teens sequential pricing as a run rate, even though the CEO said he does not see the trend changing much. The CFO's remark that further margin expansion now comes "mainly from pricing" confirms that volume and cost will not help margins until the transitions mature.

Whether Peers' Capacity Plans Threaten Share

The only competitive question of the call set Seagate's flat 4% to 6% capital intensity against peers talking about 8% and against the third producer's reported capacity addition, and asked whether share loss was a risk. Management declined to discuss competitors and argued that adding drive capacity is far harder than adding tools.

Q: "I understand you're maintaining your capital intensity 4% to 6%. Obviously, your peers spoke about going into like 8% in the next couple of years. So obviously, the third player talking about adding capacity in Japan. So I'm just wondering, is the 4% to 6% still a good bogey to use for the next few years? Or should we worry about any potential share loss?"
— Krish Sankar, TD Cowen

A: "Well, I don't really speculate on what our competitors' plans are. I'd just say relative to ours, I think 4% to 6% is just fine."
— Dave Mosley, Chair & CEO

Assessment: The refusal to engage was predictable and correct for a company about to negotiate calendar 2028 prices. The substance sits in the rest of the answer: supply is allocated through 2028 and management expects it to "remain fairly tight." Holding capital intensity flat while peers raise theirs is either discipline or a share give-up, and the calendar 2028 renewals will show which.

How Much Free Cash Flow Goes Back to Shareholders

A question noted that no debt matures until 2029 and asked what share of free cash flow the March-quarter step-up implies. Management restated its model of returning 75% of free cash flow, with anything above that opportunistic. The CFO confirmed that the step-up starts in the March quarter, with some repurchases already under way.

Q: "Are you going to do 100% of the free cash flow that we start with in that March quarter of '27, and can that continue up until you have to pay down that debt? What kind of percentages of the free cash flow are we looking at here?"
— Ben Reitzes, Melius Research

A: "We're very happy to be through the period where we dealt with the debt, like you said, and we're back to being ourselves, our model is 75%, and we should be able to get back to that model over some period of time."
— Dave Mosley, Chair & CEO

Assessment: The 75% figure is not new, but for the first time it has no debt programme in front of it, which makes it gradeable. "Over some period of time" is a soft edge on it. With net cash by year-end and $4.4B of authorization, the March quarter should show repurchases near $1B; anything well below that would mean the framework is aspirational.

What the 2030 Agreements Actually Lock

A question asked how the newer agreements running to 2030 differ from the nearer-term ones on exabytes and pricing, and how they shape capacity plans. Management said the process is the same, that only the 12-month window is locked on configuration and price, and that the company retains the ability to pull back if a customer is not taking supply.

Q: "Maybe just touch on like -- is there any changes or differences between the agreements going out all the way through calendar '30 versus maybe the ones that are more near term in terms of calendar '28."
— Joseph Cardoso, JPMorgan

A: "Our process hasn't changed. So we're giving visibility to people about how -- what our exabyte growth will be and then we're saying, basically, when you get within, say, 12 months, we lock in the configuration that we're going to be building in our factories, and then we lock in pricing, and that's what I would call the build-to-order process that we have."
— Dave Mosley, Chair & CEO

Assessment: The clearest description yet of what the long agreements are: a shared exabyte plan, with binding price and configuration only inside a 12-month window. That makes the 2030 horizon a planning tool rather than a revenue guarantee, and it confirms that both sides can adjust. It is the most important caveat on the "visibility" bull argument, and management stated it plainly.

What They're NOT Saying

1. HAMR yields, for a fifth consecutive quarter: this quarter management said transition yields "may not be where we want" and take a quarter or two to recover, and the shortfall showed up in cost per exabyte. It still gave no number. A disclosure that has become more relevant each quarter remains an inference from the margin line.
2. The size of the external component headwind: asked directly to quantify its gross margin impact, the CFO said only that it is "included in the December guidance." The split between component pass-through and value pricing is the single most important input to the margin's durability, and it is unknown.
3. Mix versus like-for-like price: asked how much of the price increase per terabyte came from HAMR mix rather than like-for-like increases, management answered qualitatively. Without that split, the market cannot separate a technology premium from a shortage premium.
4. Why Mozaic 4 breadth did not move: the Mozaic 4 ramp has been at two cloud customers for two quarters, and the next qualifications moved from "underway" in July to completing "over the next couple of months." Management framed it as planned staging, but did not name how many customers or which quarter each qualification closes.
5. Any view on competitor capacity: asked about peers moving toward 8% capital intensity and a third producer adding capacity, management declined to engage. Given that this was the trigger for a 29% decline in the stock, silence is defensible commercially and costly for the multiple.
6. The terms of the 2030 agreements: "some cloud customer agreements extending through calendar 2030" came with no exabyte, revenue or customer count, and the Q&A made clear that price is set only 12 months out. The headline horizon is long and the binding portion is short.
7. The 14-week comparison: neither the release nor the call mentioned that the year-ago quarter had an extra week. Management said in July 2025 that the extra week did not change build-to-order supply, but with post-print target cuts citing the exabyte deceleration, a one-line statement of the comparison basis would have helped.
8. Why the share count rises despite the buyback: diluted shares are guided up 1M to 233M for December after $400M of repurchases. Share-based dilution at a higher stock price is the likely reason, and it means the first $400M a quarter only holds the count steady.

Market Reaction

• Pre-print setup: STX closed October 27 at $667.85, up 142.5% year-to-date (S&P 500: +12.8%) and up 190.0% over the trailing twelve months, but down 27.2% over the trailing 30 days. The close sat 39.0% below the 52-week closing high of $1,094.04 (June 22) and 17.5% below the 50-day average of $809.96. From $945.57 on October 1, the stock fell 10.21% on October 2 on 13.6M shares as the Toshiba capacity reports landed, reached a closing low of $630.49 on October 23 (33.3% below October 1), and rose 3.67% on print day itself to $667.85 on 9.0M shares.
• After-hours move (October 27): the stock fell roughly 5.5% in extended trading to about $630; post-print coverage described the move as profit-taking after the pre-print rally.
• Reaction session (October 28): opened at $646.84 (−3.1%), traded as low as $621.50 (−6.9%), then reversed to close at $697.97 (+4.5%), 12.3% above the intraday low, on 10.8M shares, 2.2 times the 30-day average. The S&P 500 rose 0.3%. Storage and memory peers rallied too: Western Digital +4.1%, Micron +5.9% and SanDisk +7.7%.
• Since: STX closed October 29 at $700.47 (+0.36%). On publication day the whole memory and storage group was trading sharply lower in a sector-wide selloff after SanDisk's report.

The reaction session is a two-part story. We read the morning as trading the quarter's composition: flat exabytes, cost per exabyte up, and a guide premium well below July's on EPS (7.0% against roughly 25%). By the close, the market had re-read the December guide, the pricing commentary and the capital-return framework, and bought the stock back 12.3% from its low. That close was in line with storage and memory peers on the same day (ahead of Western Digital, behind Micron and SanDisk), so much of the recovery was the sector rallying rather than a verdict that Seagate's print was better than feared.

The more important fact is the starting point. The stock entered the print 29.4% below its October 1 close. Nothing in Seagate's own disclosures changed between October 1 and October 27. What changed was the industry's supply outlook (the reported Toshiba plan, and a peer's capex commentary), the sector's sentiment (memory stocks selling off despite strong results), and the multiple that investors will pay for peak-looking margins. A stock that fell a third in three weeks on a reported supply plan, and then reported a record quarter with its book allocated through 2028, is trading a supply scenario two years out, not the business in front of it.

Sell-side reaction was unusually negative for a quarter this strong: one downgrade, and nearly all of the published target changes we could find were cuts. The reasons were consistent: fewer exabytes than modeled, the risk of more industry supply, and lower peer multiples. Even after the cuts, the targets we could find sit between $750 and $1,350, all above the stock. The Street's estimates went up while its targets went down.

Street Perspective

Debate: Is the Supply Response the Start of the Cycle Turn?

Bull view: The bull camp argues that the supply response is too small and too slow to matter. The reported Toshiba plan would at most double a roughly 11% share over two years, with non-HAMR technology; Seagate's book is allocated through calendar 2028; and management says the supply-demand gap is probably growing. Adding real drive capacity takes years of heads, media and qualification work, and the 2028 allocation is already in place.

Bear view: The bear camp argues that every storage cycle has ended the same way: the shortage ends because suppliers add capacity, and pricing reverses before volume does. A rival reportedly targeting 30% share, a peer raising capex and a peer matching Seagate's gross margin all point to an industry that sees today's prices as an invitation. Calendar 2028 prices will be set while that capacity is coming on.

Our take: The bears are right about the mechanism and early on the timing. The arithmetic matters: a doubling of Toshiba's entire capacity adds about one-fifth of two years' demand growth at a mid-20s rate. That narrows the gap rather than closing it, which slows pricing rather than reversing it. Our model already assumes the price curve slows sharply in fiscal 2028. The point at which we would concede the bear case is evidence that calendar 2028 terms land below calendar 2027 escalators, and that evidence arrives over the next two or three quarters.

Debate: Is a Quarter Built Almost Entirely on Price a Lower-Quality Quarter?

Bull view: The bull view holds that price is the point. A company that can raise revenue per exabyte 16% in a quarter, on contracts rather than spot, with demand still exceeding supply, has pricing power that the multiple does not yet reflect. Volume returns in December as Mozaic 4 qualifications close, and the company then has two engines rather than one.

Bear view: The bear view is that a quarter with flat volume, rising unit cost and all growth from price is the classic late-cycle pattern. Part of the price is a pass-through of component inflation, not value capture, and it will reverse when those costs fall. Cost per exabyte rising sequentially shows the cost curve is no longer helping.

Our take: The bear is right that this quarter's mix is lower quality than June's, and the bull is right that the price is contracted and will hold for the contract term. We do not know how much of the price step is pass-through, and management chose not to say. Our response is to model the price curve slowing from here and to hold management to the December exabyte step-up. If December delivers volume and cost per exabyte falls again, this quarter was a transition dip. If not, the bear's pattern fits better.

Debate: What Multiple Does a Contracted Peak Deserve?

Bull view: The bull argument is that 18x the current fiscal year's earnings, for a business with contracted pricing through 2027, allocated volume through 2028, net cash by year-end and a 75% return framework, is a cyclical multiple on a business with unusual visibility. At current prices, the market assumes peak earnings within two years.

Bear view: The bear argument is that peak margins always look contracted at the peak. A 57.8% gross margin in a business that earned 20% to 25% for most of its history is not a base to capitalize at a growth multiple, and the right comparison is normalized earnings, which would put the stock well below today's price even after a 36% drawdown.

Our take: We cut our multiple to 22x to 25x the fiscal 2027 estimate, from 26x to 30x, because the supply risk the bears describe is now visible rather than theoretical. We do not cut further, because the fiscal 2027 earnings we are capitalizing are mostly contracted, and because a normalized-earnings frame ignores that the industry's cost structure and technology lead have changed. The stock at 18x pays for none of fiscal 2028, which is the year the debate is really about.

Model Update Needed

Valuation impact: Our fair-value range moves from $830–960 to $850–970. The fiscal 2027 estimate rises about 21%, from roughly $32 to roughly $38.80, and we cut the multiple from 26x–30x to 22x–25x to reflect the visible industry supply response and the de-rating of the peer group. Against the $700.47 close on October 29, the midpoint of $910 implies about +30%, and the low end implies about +21%. That is the widest spread of this coverage, and it is wide because the estimate rose about 21% while the stock fell about 8% from its close on the day of the July report. Post-print sell-side targets span $750 to $1,350; our range sits in the lower half because we capitalize only the contracted fiscal 2027 path and nothing from calendar 2028 pricing.

Thesis Scorecard Post-Earnings

Overall: Thesis unchanged in direction and slightly weaker in composition. The demand pillar and the balance sheet are stronger than in July, the capital-return commitment finally has a number, and the pricing mechanism is more contracted than we feared. Against that, the quarter's growth rested almost entirely on price, cost per exabyte rose, Mozaic 4 breadth did not move, and the industry supply response is now visible. As in July, the fundamentals improved while the stock fell into the print, and the valuation gap is now the widest of this coverage.

Action: Maintain Outperform; fair-value range $850–970 against $700.47. Signposts for the December quarter: a sequential exabyte step-up of at least about 7% (roughly 235 EB), new Mozaic 4 qualifications closing at named customer types, cost per exabyte falling again, an implied ~60% gross margin print, and any first calendar 2028 pricing commentary. The pre-committed exit conditions from July did not trip: no verified hyperscaler capex cut, no retreat from the sequential commitment, no reversal of the pricing layer, and a $400M buyback. They are revised to: calendar 2028 terms landing below calendar 2027 escalators, a December quarter without an exabyte step-up, Mozaic 4 qualifications slipping past the March quarter, March-quarter repurchases well below $1B, or a verified hyperscaler capex cut. Any one of those moves us to Hold. A recovery toward the upper end of the fair-value range does the same, on discipline rather than doubt.