Fully Allocated Through 2026, Records on Every Line, and a March Guide 15% Above the Street — Maintaining Outperform as the Debate Shifts From the Business to the Multiple

Results vs. Consensus

FQ2 FY2026 Scorecard

Year-Over-Year Comparisons

Quarter-Over-Quarter Comparisons

Quality of Beat

Revenue: The $2.83B print cleared not just consensus but the top of the guided range, the second consecutive quarter of that pattern, and it happened with unit capacity deliberately flat. The upside mechanism is now well understood: qualification-driven pull-ins of higher-capacity product, this quarter aided by another HAMR customer coming online. Growth was broad (sequential improvement across nearly all end markets), with data center revenue of $2.2B up 28% year-over-year and even edge IoT recovering seasonally to $601M. Nothing in the composition suggests channel stuffing or pull-forward: under build-to-order, the customer takes what was contracted, and the backlog extends regardless.

Margins: An eleventh consecutive record at 42.2% non-GAAP, with the GAAP figure (41.6%) just 60bp behind, and the drivers compounding rather than fading: like-for-like contract escalators, rising revenue per terabyte, HAMR mix (>1.5M units per quarter now), and operating expenses held flat at $290M so that gross margin gains flow through at nearly full weight. Operating margin at 31.9% is the highest in company history, and opex as a share of revenue (10.3%) is a quarter away from the long-term 10% target. The March guide implies gross margin in the 44% neighborhood, which would have been dismissed as fantasy four quarters ago.

EPS: The $3.11 absorbed a full 16% tax rate and 226M diluted shares and still grew 19% sequentially, 53% year-over-year. The convertible-note retirement ($500M of the 2028 exchangeables taken out this quarter) is the quiet quality item: it converts a future dilution overhang into a fixed retirement cost, capping the share-count creep that had been adding roughly 2M shares per quarter. Below the line, interest expense continues to fall with gross debt down to $4.5B.

Segment Performance

Revenue by Segment — FQ2 FY2026

Key Volume & Mix KPIs

Data Center — Cloud Compounding, Enterprise Joining

Data center shipped 165 exabytes (+31% YoY) and $2.2B of revenue with the growth now two-engined: global cloud demand continued its multi-year compounding, and the enterprise OEM market delivered a second consecutive sequential improvement, driven by traditional server units and, more interestingly, storage servers purchased to hold data at the enterprise edge for AI applications. Average cloud nearline capacity reached nearly 26 terabytes and rises mechanically as the HAMR mix builds. Management's tiering argument sharpened this quarter: hard drives anchor the mass-capacity tier that holds training checkpoints and the vector databases behind inference and agentic workloads, and drives in these architectures run 24/7 at streaming-optimized workloads rather than sitting as cold archives.

Assessment: The enterprise edge is quietly becoming the second demand engine we flagged as optionality in October. If agentic AI adoption (management cited a CSP survey showing more than half of customers already using agents) pulls storage-dense infrastructure on-premises, the demand base broadens beyond the eight hyperscalers, which would be the single most de-risking development possible for the multi-year thesis.

Edge IoT — Seasonal Lift, Structural Deprioritization

Edge IoT revenue of $601M rose seasonally on consumer and VIA client demand. The more interesting datapoint arrived mid-January, between the quarter's end and the print: Seagate pushed 32TB capacities (Mozaic-class) into the retail and channel market globally, at prices roughly double what equivalent capacity cost in September, against a channel where average HDD prices rose 46% in four months. The segment remains supply-deprioritized relative to cloud, and correspondingly more profitable per unit than at any point in its history.

Assessment: The channel price environment is a real-time referendum on scarcity, and it is unambiguous. Watch the VIA nearline crossover: management expects the VIA market's largest growth to come from nearline-class products reported inside data center, which will make the edge IoT line look structurally flatter than the underlying demand.

Key Topics & Management Commentary

Overall Management Tone: Assured and, for the first time in this coverage, openly expansive about the duration of the cycle: the prepared remarks reached to calendar 2028 demand discussions and a 10TB-per-disk roadmap into the next decade without hedging. Management engaged pricing questions with unusual directness (volunteering that flat-to-up per-exabyte pricing is "certainly possible"), pushed back on nothing, and reserved caution only for the pace of the Mozaic 4 ramp, which was framed as deliberately prescriptive rather than maximal. The word that recurred was "predictable": management wants the market to understand that the model's upside comes from executing transitions, not from chasing spot demand.

1. From Committed to Fully Allocated: CY2026 Is Sold, CY2027 Is Opening

"Our nearline capacity is fully allocated through calendar year 2026, and we expect to begin accepting orders for the first half of calendar year 2027 in the coming months. Further out, demand visibility is strengthening based on the long-term agreements in place with major cloud customers through calendar '27. Additionally, multiple cloud customers are discussing their demand growth projections for calendar '28." — Dave Mosley, Chair & CEO

The language hardened materially from October's "largely committed." In Q&A, the CFO made the commercial structure explicit: for calendar 2026, purchase orders are in place for every quarter with volume and pricing both locked; any production upside gets sold "in the open market at a good profitability." Calendar 2027 has volume indications and agreements but pricing deliberately not yet fixed, with the longest-lead-time wafer starts for 2027 beginning shortly. Customers are behaving accordingly: the behavior change management most values is buyers who miss allocation responding by planning the next two years rather than bidding for spot.

Assessment: The 2027 pricing negotiation is now the single most valuable open item in the model. Seagate enters it with capacity oversubscribed, a competitor equally allocated, pricing momentum positive for the first time ever, and customers self-selecting into multi-year predictability. The setup argues 2027 like-for-like pricing resets meaningfully higher, which is precisely what the current multiple is betting on; the December quarter's disclosure pattern (open 1H27 orders "in coming months") means evidence will arrive within two quarters.

2. Revenue per Terabyte Turned Positive: The Regime Change Quantified

"This improvement reflects the ongoing execution of our pricing strategy and the growing adoption of our latest generation high capacity products, which collectively drove a modest sequential increase in revenue per terabyte, a trend we expect to continue into the March quarter." — Gianluca Romano, CFO

Asked whether blended per-exabyte pricing, which has improved from double-digit annual declines to a 4% decline exiting the quarter, could go flat or positive year-over-year, the CEO answered directly: "flat to slightly up is certainly possible," dictated by demand, with the renegotiation cycle as the mechanism. The trajectory across our three quarters of coverage: $/TB declining at a "consistent" clip in October, "relatively stable" in the prepared remarks tonight, and now modestly rising sequentially with a repeat expected in March.

Assessment: This is the fulcrum metric of the entire investment case, because it separates a mix story (which eventually exhausts) from a pricing story (which compounds). A structurally positive $/TB in an industry shipping 25%+ more terabytes per year converts exabyte growth into revenue growth at better than 1:1, which is what makes the March guide's +34% YoY credible and the FY27 consensus too low. We now track this line above every other KPI.

3. HAMR: All Major U.S. CSPs Qualified, 1.5M Units a Quarter, Mozaic 4 Ramping

"We qualified the last big cloud service provider in the U.S. and we have qualified 6 out of 8 of the top cloud service providers. So the transition from PMR technology to HAMR technology is progressing very well." — Gianluca Romano, CFO

The qualification ledger now reads: all major U.S. CSPs on Mozaic 3, six of the eight top global providers, the remaining two on track for the first half of calendar 2026, and quarterly HAMR shipments exiting the year above 1.5 million units (roughly 45–54 exabytes at 30–36TB per drive, or about a quarter of total volume). Mozaic 4 begins its ramp later this quarter with two major customers in qualification and multiple CSP qualifications expected "in the coming months." The lab demonstrated 7TB per disk, extending the roadmap visibly beyond the 5TB generation, with 10TB targeted early next decade. Management deliberately constrained Mozaic 3 volume because factories are full and the 4TB transition is the priority.

Assessment: HAMR unit volume grew roughly 50% sequentially while being deliberately constrained, which says the binding limit is factory transition cadence, not demand or yield. The Mozaic 4 ramp beginning this quarter is the fiscal year's central execution event: its yield curve dictates the cost-down trajectory (management expects the 40TB drive to drive "a fairly important reduction in cost per terabyte") and the calendar-2027 supply plan. The self-graded "on the plan or slightly ahead" report card, alongside the 7TB lab demo, leaves the execution record unblemished.

4. The March Guide: +34% YoY in the Former Seasonal Trough

March was historically Seagate's weakest quarter; the guide calls for $2.90B ± $100M (+2.7% sequentially, +34% YoY at the midpoint), operating margin approaching the mid-30s, and $3.40 ± $0.20 of EPS against the $2.96 the Street had modeled. Management stated flatly that data center demand "will more than offset typical March quarter seasonality" and, more importantly, extended the commitment: sequential top- and bottom-line improvement "throughout calendar 2026." The implied gross margin at the guide midpoint sits in the ~44% neighborhood, a twelfth consecutive record, with incremental margins again near or above 100% of the prior quarter's rate.

Assessment: A 15% EPS guide beat is the kind of number that forces the entire Street to rebuild models, which is mechanically what a +19% session is. The calendar-2026 sequential commitment is the forward anchor: taken literally, it implies a December 2026 quarter above $3.1B and quarterly EPS approaching $4, which is the earnings power the current price is capitalizing. Management has met or beaten every such commitment in our coverage window; the commitment itself is now the thesis.

5. Balance Sheet Endgame: Converts Retired, Leverage at 1.1x

"We retired approximately $500 million of exchangeable senior notes due 2028, which serves to limit further dilutive impact from business and optimized cash deployed for future share repurchases." — Gianluca Romano, CFO

The quarter's capital allocation went to the convertible overhang rather than open-market repurchases: $500M of the 2028 exchangeable notes retired, gross debt down to $4.5B, net leverage at 1.1x, and the highest quarterly free cash flow in eight years ($607M) funding it all alongside the $154M dividend. Management expects leverage to keep trending lower and is evaluating further debt reduction.

Assessment: Retiring converts at the source is economically superior to buying back shares to offset their dilution, even if it makes the "buyback cadence" we asked for in October look absent on the surface: the dilution simply stops existing rather than being offset quarterly at ever-higher prices. With leverage at 1.1x and FCF compounding toward a $2.5B+ annual run-rate, the capital-return story is no longer about capacity; it is about willingness, and the >75%-of-FCF commitment now has real numbers behind it.

6. The Demand Architecture: Video, Agentic AI, and the Warm Tier Defense

"Among the most promising of these is agentic AI, which relies on persistent access to large volumes of historic data to enable effective planning, reasoning and independent decision-making… one recent survey conducted by a leading cloud service provider report[ed] more than half of participating customers were actively using AI agents." — Dave Mosley, Chair & CEO

The demand argument added two layers this quarter. First, scale datapoints: YouTube now absorbs 20 million video uploads daily, up from 2 million three years ago, before counting the AI-generation wave. Second, the agentic argument: agents require persistent access to large historical datasets for planning and reasoning, which makes stored data an input to compute rather than an output of it. On the perennial warm-tier question (whether NAND claims the middle tier in AI architectures), the CEO's answer was his most architectural yet: random small-block workloads belong to memory, big-data streaming belongs to drives running 24/7, and "the architectural tier that stores the data will probably remain constant for the next decade."

Assessment: The agentic framing matters because it makes storage demand usage-driven and cumulative: every agent deployment adds a persistent data dependency that outlives the session. The "next decade" architecture claim is aggressive but currently supported by the capital math (NAND replacing HDD exabytes remains prohibitively expensive, and NAND itself is now shortage-priced). We note that management's demand narrative has been directionally right, and conservative, for four consecutive quarters.

7. Pricing Philosophy Under Scarcity: Deliberate, Not Opportunistic

Multiple questions probed why Seagate is not repricing more aggressively when NAND vendors are pushing 40–100% quarterly contract increases. Management's answer was a philosophy statement: honor existing commitments, reprice at renewal and product transition ("as one long-term agreement rolls into the next… we have constrained supply of those new products, then we look at what the demand is, and we dictate where our pricing is"), and preserve the predictability that makes customers willing to commit years ahead. The one early-HAMR customer discount continues rolling off. Notably absent: any commitment to match the NAND industry's spot-price aggression.

Assessment: Leaving near-term price on the table is the premium Seagate pays for a multi-year order book, and it is the right trade: the NAND vendors' 50% hikes are the behavior of an industry that cannot contract forward, and it invites demand destruction and substitution the moment supply normalizes. Seagate's escalator model monetizes scarcity over years instead of quarters. The cost is that current-quarter margins understate achievable pricing; the benefit is the 2027 book gets written against a demonstrated, sustained scarcity rather than a spike.

8. Costs: Mid-Teens Declines Now, Mozaic 4 the Next Leg

Cost per terabyte has been falling at a mid-teens annual rate through the HAMR transition, and management expects the 4TB-per-platter generation to sustain and potentially steepen that curve: unit costs stay similar while capacity per unit jumps a third, with the full effect landing "well into the next calendar year" as two major customers complete qualification and the ramp scales through calendar 2026. Depreciation stays modest (10-year useful life, capex at the bottom of the 4–6% band), so the cost curve is driven by physics and yield, not by capital intensity.

Assessment: The mid-teens cost decline against now-rising revenue per terabyte is a widening scissors that mechanically expands gross margin every quarter it persists. The dependency is the Mozaic 4 yield ramp; management's willingness to guide March gross margin to the 44% neighborhood while that ramp begins suggests the early yield data is good.

9. OpEx at the Target: The Reinvestment Signal

Operating expenses held at $290M (10.3% of revenue), reaching the long-term 10% target a quarter ahead of plan, and management flagged what comes next: reinvestment. The CEO was explicit that the lean years under-invested in the roadmap and that, with the HAMR transition cleared, incremental dollars go back into areal-density development "even staying within our same model."

Assessment: A company choosing to reinvest R&D at the moment its moat is widest is playing the long game correctly; the 7TB lab demonstration is what that spending buys. We model opex dollars growing modestly from here with the ratio held near 10%, which preserves the operating-leverage math while funding the roadmap that sustains it.

10. Calendar 2025 in Review: The Transformation, Audited

The year's ledger, as management presented it: revenue up more than 25%, gross margin up nearly 740 basis points, operating margin up more, HAMR from one ramping customer to six of eight global CSPs qualified with 1.5M+ units a quarter, the first CSP shipping 3TB-per-disk product, and records in every financial metric that matters. Set against the commitments made at the May 2025 Investor Day (low-to-mid-teens growth, 40% gross margin milestone, 50% incrementals, 10% opex, 4–6% capex), every target was met early and exceeded.

Assessment: The FY2028 model is now functionally obsolete as a ceiling; it survives only as a floor. The next strategic checkpoint (most plausibly an analyst event or the fiscal-year close in July) should bring a formally raised long-term framework, and the December quarter's exit economics suggest the new margin frame starts with a 4. That re-guide is a genuine catalyst, and the Morgan Stanley-reported management expectation of 50%+ HDD gross margins within 12 months (from December's investor meetings) previews its shape.

Guidance & Outlook

The March guide obliterates the seasonality question that lingered from October (when management would only offer "we expect a good quarter"): +34% year-over-year growth in what was historically the softest quarter, with data center demand explicitly guided to more than offset edge IoT seasonality. The sequencing commitment extends beyond the quarter: revenue and profitability are guided to improve sequentially through all of calendar 2026, which implies the December 2026 quarter exits above $3.1B with EPS approaching $4.

Implied margin trajectory: operating margin "approaching the mid-30s" with $290M of opex at $2.9B implies non-GAAP gross margin around 43.5–44.5%, a twelfth consecutive record. The margin drivers stack rather than trade off: revenue per terabyte guided up again, HAMR mix rising with the Mozaic 4 ramp beginning, the early-customer discount continuing to roll off, and cost per terabyte falling at a mid-teens rate.

Street at: consensus entering the print sat at $2.77B and $2.96 for March; the guide midpoints exceed both by 4.7% and 14.9% respectively. For the fiscal year, the first-half actuals ($5.45B, $5.72 EPS) plus the March guide put nine-month EPS at $9.12, implying full-year consensus in the mid-$12s versus roughly $11.20 modeled after October's print. The revision cycle is still steepening: each of the past three prints has forced a double-digit percentage rebuild of forward numbers.

Guidance style: the established pattern held again this quarter (revenue and EPS both landed above the top of October's guided ranges), so the same logic applies forward: $2.90B and $3.40 are supply-committed floors with upside from qualification pull-ins, and the guide's own record implies the high ends ($3.0B, $3.60) are live outcomes.

Analyst Q&A Highlights

Can Per-Exabyte Pricing Actually Turn Positive?

The opening question crystallized the quarter's central economic development, noting the improvement from double-digit annual price declines to a 4% decline exiting the quarter, and asked whether flat or positive year-over-year pricing is achievable, along with a framework for incremental gross margins.

Q: "As you think about overall average pricing per exabyte, we've gone from kind of down double digits to high single digits. And I think we just exited the quarter down 4% year-on-year. Do you see a world where pricing could [be] flat or even move positive year-over-year?"
— C.J. Muse, Cantor Fitzgerald

A: "The pricing will be dictated by the demand. Right now, the demand is really strong. So I think as we roll through into '27 and '28, we look at how much capacity we're having… We'll bring more exabytes to bear and then people go out there and renegotiate for those. I think flat to slightly up is certainly possible. And that's the way we're really managing it as we talk to our customers."
— Dave Mosley, Chair & CEO

Assessment: A CEO volunteering "flat to slightly up is certainly possible" for per-exabyte pricing, in an industry that has never in its history sustained positive pricing, is the most economically significant sentence on the call. Even held flat, pricing at these levels converts the entire mid-20s exabyte CAGR into revenue growth; positive pricing makes the March guide's +34% the template rather than the peak.

Why LTA Pricing Beyond 2026 Cannot Simply Be Marked Higher

The follow-on pressed the obvious aggressive case: with supply this tight, why should embedded LTA pricing for 2026 and beyond not be dramatically higher? The answer revealed how management weighs demand persistence against spot maximization.

Q: "The opportunity for pricing… you just said sort of flat to up as possible. But as we think about the pricing that might be getting embedded within these LTAs and sort of beyond '26. Why can't that be a lot higher just given the tightness in the supply-demand environment?"
— Wamsi Mohan, Bank of America

A: "I think this gets into how persistent is the demand going to be… The one behavior change that I really like in the last year is that people are starting to say, if I can't get it now, I'll plan next year better and the following year better. So we're having great dialogues on that front… my perspective on this is I think demand will stay strong for quite some time. So in that kind of world, we're having great discussions with customers further out in time."
— Dave Mosley, Chair & CEO

Assessment: The answer is that Seagate is selling predictability, not scarcity, and predictability is what converts a price spike into a repriced curve. The CFO's addendum ("we expect revenue and profitability to continue to improve sequentially every quarter… it's actually now getting better somehow") tells you the escalator embedded in that predictability is doing its work without headline shock therapy.

The Shape of Exabyte Supply Growth Through the HAMR Crossover

With demand visibly outpacing supply, the question turned to how much exabyte supply can actually grow this calendar year and whether growth accelerates through the crossover. The answer was a candid description of a full factory in mid-transition.

Q: "At your Analyst Day last year, you kind of pointed to a mid-20% exabyte growth CAGR. And I'm just wondering where you think that supply growth can land this calendar year. And as you get closer to that HAMR crossover point later this year, like does that pace of exabyte growth accelerate?"
— Erik Woodring, Morgan Stanley

A: "We are planning to transition to 4 terabytes [per] platter… fairly aggressively, but I think what people have to keep in mind is that we were fairly tight all throughout manufacturing. So we have products that are in the pipeline already that are committed to customers… It won't be as fast as maybe we've done some ramps in the past, but it will be very profitable, and that's the way we look at it."
— Dave Mosley, Chair & CEO

Assessment: "Prescriptive, not maximal" is the ramp philosophy, and "it will be very profitable" is the tell: Seagate is optimizing the transition for margin rather than share, which only a vendor with a sold-out book and no competitive pressure at the leading capacity points can afford. Supply growth stays in the mid-20s band near-term with acceleration deferred to the Mozaic 4 volume phase.

What Locks 2026, and What Is Actually Agreed for 2027

The most valuable clarification in the Q&A pinned down the commercial mechanics of the order book: what exactly is committed for calendar 2026 versus 2027.

Q: "Is it right to assume that pricing is also locked in for all of '26? And what sort of agreement are you referring to for 2027 if volume and pricing is not locked in next year?"
— Timothy Arcuri, UBS

A: "For this calendar year, we said basically, we have PO in place for all the quarters, so volume and pricing is well defined. As Dave said before, if in a quarter, we can produce a little bit more, of course, we will sell those exabytes in the open market at a good profitability… Calendar '27, we will start working on that fairly soon. Of course, we have very good indication and agreement on volumes, but we have not fixed the price yet."
— Gianluca Romano, CFO

Assessment: This is the cleanest disclosure yet of the model's structure: 2026 is a closed book (POs, volume, price), upside production clears at spot ("good profitability" in a 46%-channel-inflation market), and 2027 volumes are agreed with pricing to be set from a position of maximum leverage. The 2027 price-setting round, beginning within months, is the next structural catalyst and the one we will grade hardest.

What Is Driving the March Quarter's Extraordinary Incremental Margins

The guide's implied incremental gross margins (in the vicinity of 100%) prompted a direct question about whether something unusual is flattering the March quarter.

Q: "It seems to be a really sizable uptick in gross margins. I think it's up like 250 basis points or 100% plus incrementals. Could you just — is there any[thing] you would call out in March quarter that's unique that's helping drive that kind of margin expansion? And is this really all coming from the core HDD business?"
— Amit Daryanani, Evercore ISI

A: "I would say, we expect it to be a very good quarter. I don't think it's different than what we have done before. It's always based on the pricing strategy and the mix, as you know. We qualified another customer on HAMR, so we will ramp a little bit more volume on HAMR. This is helping us to get better margin. But fundamentally, [it] is not really different in how we think we are going to execute the quarter."
— Gianluca Romano, CFO

Assessment: "Nothing unique" is the bullish answer: the March margin step is the same four-driver stack (escalators, mix, HAMR scale, cost-down) that produced the last eleven records, not a one-time item that reverses. The CEO's addendum ruling out the systems business as a driver removes the last benign-mix alternative explanation.

Could HDD Pricing Follow NAND's 40–100% Hikes?

With NAND contract prices reportedly rising 40–100% quarter-over-quarter, the question was whether rolling LTA renewals give Seagate the opportunity for similarly dramatic repricing, plus an update on HAMR mix trajectory.

Q: "Do you think there's an opportunity here for more significant price increases [like] in NAND flash. We're hearing things like 40% to 100% up Q-on-Q for some contracts… as the LTAs roll off, is there an opportunity for some of those to be repriced at a more significantly higher price to change the trajectory[?]"
— Mark Newman, Bernstein

A: "As we — as one long-term agreement rolls into the next year or the next year, we've satisfied our existing supply commitments, people are looking at the new products, we have constrained supply of those new products, then we look at what the demand is, and we dictate where our pricing is… And again, what we're seeing is people who can't get what they need today, they're saying, okay, I need to be able to plan my data center procurement out in the future, let's get more predictable in the future."
— Dave Mosley, Chair & CEO

Assessment: "We dictate where our pricing is" is as much pricing power as a CEO can claim on a public call without inviting customer retaliation. The refusal to chase NAND-style hikes is discipline, not weakness: contract escalators compound across a multi-year book, while spot spikes reprice one quarter and invite substitution. The HAMR mix answer (deliberately constrained Mozaic 3, aggressive pivot to Mozaic 4, mix-up to the highest capacity points) confirms margin, not volume, is the optimization target.

Decomposing the Margin: Pricing vs. Mix vs. Cost

An attempt to bucket the gross margin expansion into pricing, mix, and cost-reduction components drew a deliberate non-answer on the split but a useful confirmation of the HAMR mix arithmetic.

Q: "If there's a way to put it in 3 buckets, like how much of the gross margin upside is driven by pricing? How much is driven by product mix? How much is driven by cost reduction…? How much of HAMR as a percentage of the mix?"
— Krish Sankar, TD Cowen

A: "We don't give specific details on the impact of pricing, mix and cost. But they are somehow interrelated. I would say the change in mix is helping with the cost reduction and the supply-demand situation is, of course, supporting our pricing strategy. So they are all very good contributor[s] to the increase in gross margin. And as we said before, this is going to continue through the calendar year… Dave gave an indication of the unit[s] that we shipped in the last quarter. So I think you can fairly easily calculate that."
— Gianluca Romano, CFO

Assessment: The pointed refusal to decompose is rational (the split is competitive intelligence for both customers and the rival), but the "you can fairly easily calculate" nudge on HAMR mix (1.5M units × 30–36TB ≈ 45–54 EB, roughly a quarter of shipments) is the closest thing to a formal mix disclosure yet. At this trajectory HAMR crosses half of nearline exabytes on schedule in the second half of calendar 2026.

What They're NOT Saying

1. No share repurchases at all this quarter: The October report card asked for buyback cadence; the December answer was $0 of open-market repurchases, with capital directed to the $500M convert retirement instead. The economics favor the choice, but the $5B authorization is now two quarters old with $29M used, and the stock has tripled since it was set. At some price, management's revealed preference (debt first, dilution second, buybacks third) is itself a valuation signal worth hearing.
2. The 50% gross margin conversation happened off the call: Management told investors in December (as reported by a major sell-side desk) that 50%+ HDD gross margins are expected within 12 months, yet nothing resembling that number was said on this call, where the framing stayed at "continue to improve sequentially." The gap between the private trajectory and the public guide is where the next several beats live.
3. CY2027 pricing is genuinely unset: The candor about 2027 (volumes agreed, price not fixed) cuts both ways: it is the upside case, and it is also an acknowledgment that the biggest variable in the FY27–28 model has no contractual floor yet. If AI infrastructure spending pauses before the 2027 book prices, the negotiation leverage inverts quickly.
4. Mozaic 4 yield state remains unquantified: Third consecutive call describing 4TB-platter yields as early/improving without numbers, now with the ramp starting. The March quarter's ~44% implied gross margin is an implicit yield claim; a miss there would be the first hard evidence of trouble.
5. The exabyte CAGR framework went undefended: Asked directly for calendar-2026 exabyte growth, management declined to guide and restated the mid-20s framework it is visibly exceeding. The reticence keeps the framework beatable, but it also means the eventual formal raise (which the FY28-model refresh will force) is pre-announced to anyone listening.
6. Nothing on the competitor's next-day print: The rival reported the following evening, and no question addressed competitive dynamics at all: extraordinary for a two-vendor industry, and a measure of how completely the scarcity regime has suspended competitive tension. The dormant risk: both vendors' HAMR-generation products meet at the same CSP qualification tables in 2027.

Market Reaction

• Pre-print setup: STX closed January 27 at $371.76, itself the 52-week closing high, up 35.0% year-to-date in under a month (S&P 500: +1.9%), up 259% over the trailing twelve months, and up 29.9% over the trailing 30 days, a stretch that included an 11% single-day surge on January 6 when a major AI-industry keynote called AI storage a "completely unserved market" that would become "the largest storage market in the world."
• After-hours move (January 27): Initially fractionally lower (roughly -1%) on the headline beat, which was largely pre-priced after the month's run; the re-rate built through the 5:00 PM call as the fully-allocated disclosure, the calendar-2026 sequential commitment, and the March guide landed.
• Reaction session (January 28): Opened at $395.19 (up 6.3%) and climbed all day to an intraday high of $449.50 (up 20.9%), closing at $442.93, up 19.1%, on 14.7M shares (3.7x the 30-day average), on a flat tape. It was the second consecutive +19.1% earnings-day move.
• Peer tape: The pure-play HDD comparable rose 10.7% ahead of its own print the next evening; the NAND pure-play gained 9.6%; the DRAM major added 6.1%. Coverage framed the session as the storage supply crunch extending an already explosive complex-wide rally.

A call-driven re-rate: the tape's sequencing is informative. The print alone (a 3% revenue beat, 10% EPS beat) was worth roughly nothing after hours because three consecutive beats had taught the market to expect it. What was worth 19% was the forward architecture disclosed on the call: every 2026 exabyte sold at locked pricing, 2027 volumes agreed with pricing leverage intact, sequential growth committed for four more quarters, and a March guide that made the Street's March number look like a typo. The market repriced the duration and slope of the cycle, not the quarter.

The chase dynamic: the session's second notable feature was the sell-side response: a wave of same-day target raises landed with a median new objective almost exactly at the closing price, and the one cautious desk raising its numbers 13–22% while warning the stock "seems to be discounting the current uptrend sustaining for a long time." When targets converge onto the price rather than leading it, the marginal buyer is no longer buying against a Street framework; the stock is discounting management's own commitments directly. That is not a reason to sell a compounding story, but it removes the "Street is behind" cushion that supported the last two entries, and it is why our rating now leans on the revision cycle rather than the gap to fair value.

Street Perspective

Debate: Does "Fully Allocated Through 2026" Make FY27 Estimates a Floor or a Ceiling?

Bull view: With volume and pricing locked for calendar 2026 and 2027 volumes agreed, the only variables left are the 2027 price reset (negotiated from record scarcity) and Mozaic 4 economics (a cost-down machine). Consensus FY27 numbers are mechanical underestimates: sequential growth through calendar 2026 alone walks the exit rate to ~$12.5B annualized revenue and $15–16 of EPS before any 2027 price uplift.

Bear view: A locked 2026 tells you nothing about 2027 demand; it tells you hyperscalers panic-secured supply at the peak of an AI capex cycle. The 2028 "discussions" are wish lists, and management admitted as much last quarter. If AI infrastructure digests in late 2026, the 2027 book prices into a falling market, and locked-2026 revenue becomes the top from which comparisons fail.

Our take: The bull case has the contractual evidence and the bear case has the base rates. The tiebreaker is the demand composition shift now visible (enterprise edge, agentic workloads, sovereign deployments broadening the buyer base beyond eight hyperscalers) plus the industry's structural refusal to add unit capacity, which caps the downside amplitude of any digestion. We model FY27 above consensus but below the full sequential extrapolation, and we treat the 1H27 order-book pricing (disclosed within two quarters) as the decisive datapoint.

Debate: Is Positive Revenue-per-Terabyte Sustainable, or a Scarcity Artifact?

Bull view: Pricing turned positive with contracts, not spot: escalators embedded in renewals, mix-up to premium capacities, and a discount roll-off, all mechanisms that persist as long as supply discipline holds. Both vendors have forsworn unit capacity additions; scarcity is structural policy, and positive $/TB is the new equilibrium, exactly as it became in DRAM after that industry's consolidation.

Bear view: Every commodity in history has printed "structurally positive pricing" at the top of a shortage, and channel prices up 46% in four months are the signature of a bubble, not an equilibrium. QLC NAND is being redirected into the gap at +20% quarterly contract prices; two years of that and the substitution boundary moves for real. Positive $/TB will last precisely as long as the shortage does.

Our take: The DRAM-consolidation analogy is the strongest version of the bull case and mostly survives scrutiny: a two-vendor industry with year-long lead times, contract-locked demand, and density-led (not unit-led) supply growth has genuinely different clearing dynamics than past HDD cycles. But we are careful about what we underwrite: our model assumes $/TB flat-to-slightly-positive through FY27, not the compounding-inflation scenario, and the channel's 46% spike is exactly the kind of demand-destruction test the thesis needs to survive next.

Debate: What Is the Right Multiple for the Second Year of a Re-Rating?

Bull view: At ~$443 the stock trades near 35x FY26E, but under 28x the FY27 path, for a business compounding EPS at 50%+ with two years of contracted revenue, a 1.1x-levered balance sheet, and a coming capital-return inflection. Scarce, structurally repriced franchises (equipment monopolies, DRAM leaders in their supercycles) sustained 30x+ while their revision cycles ran, and this revision cycle is demonstrably still running: three consecutive prints have forced double-digit estimate rebuilds.

Bear view: The stock has tripled in twelve months, sits 19% above where the smartest desks' targets stood yesterday morning, and now trades at a premium to the S&P on earnings that are, by management's own architecture, cyclical with a two-year visibility window. When targets chase price and the marginal buyer is momentum, the multiple is no longer a valuation; it is a positioning statement. The January 6 keynote-driven 11% day is the tell of a crowded trade.

Our take: Both descriptions are true simultaneously, which is what a late-stage re-rating looks like. Our discipline: we hold the rating while estimates rise faster than the price (still true this quarter: forward EPS moved ~13–15% on the guide against a 19% price move; roughly a wash, with the 2027 pricing option un-priced) and we pre-commit to acting when that reverses. The fair-value range of $460–510 assumes 28–30x a de-risked FY27 path; it does not assume the 50%-gross-margin scenario, which is the un-modeled upside that keeps us from trimming into strength.

Model Update Needed

Valuation impact: Our fair-value range moves from $280–315 to $460–510, driven by the estimate rebase (FY26E to ~$12.80, FY27E to ~$16.00) with the multiple held at 28–30x FY27E rather than expanded: at this stage of the re-rating we want estimate growth, not multiple growth, doing the work. Against the January 28 close of $442.93, the midpoint (~$485) implies roughly +9%. The un-modeled options are asymmetric to the upside (the CY27 price reset, the 50%-gross-margin trajectory management has discussed with investors, a formally raised long-term framework), but the entry margin of safety is the thinnest of our coverage, and we flag it honestly rather than stretching the range to flatter the rating.

Thesis Scorecard Post-Earnings

Overall: Thesis strengthened again on the fundamental pillars, with the pricing pillar (revenue per terabyte turning positive) delivering the single most important confirmation of our coverage to date. The offsetting move is Bear #2 escalating to materializing: the stock now trades inside the Street's freshly-raised target range, and the margin of safety has compressed to single digits against our framework.

Action: Maintain Outperform; fair-value range $460–510. Signposts for the March quarter: gross margin at or above 43.5%, the 1H CY2027 order acceptance opening with pricing commentary, the final two global CSP qualifications on Mozaic 3, Mozaic 4 ramp evidence (a shipping announcement or customer count), and the June guide extending the sequential commitment. Downgrade triggers, now explicitly two-sided: fundamentally, a Mozaic 4 yield stumble or any retreat from the sequential-growth commitment; on price, a further material re-rating without matching estimate revisions would take the stock through our range's ceiling and force the valuation question regardless of execution.