Seagate FY2026 Earnings: $12.2B Revenue and a 52.3% Q4 Gross Margin as AI Data Centers Pay for Storage

A disk-drive maker just printed a software quarter. Seagate's fiscal fourth quarter delivered $3.6 billion of revenue at a 52.3% GAAP gross margin and a 43.0% operating margin, with $1.3 billion of net income falling to the bottom line. For the full fiscal year, revenue grew 34% to $12.2 billion while cost of revenue grew only 12.5% — the operating leverage of a factory running full tilt for AI data centers. And management guided the September quarter to $4.1 billion, which would be the biggest quarter in company history.

The Headline Numbers​

Seagate's fiscal year ends on the Friday closest to June 30, so FY2026 is the twelve months ended July 3, 2026 — a 53-week year, meaning roughly two points of the growth below is one extra week of sales. Every figure comes from the Form 10-K and the earnings release.

Two things stand out before we go deeper. First, the profit growth is more than triple the revenue growth — every incremental dollar of revenue carried about 76 cents of gross profit (an incremental margin of $2,358M on $3,098M of added sales). That is what a fixed-cost factory looks like at high utilization. Second, free cash flow of $3.1 billion is nearly equal to net income, so this is cash profit, not accrual profit: operating cash flow was $3,674M against just $569M of capital expenditure.

The fourth quarter alone tells the acceleration story: revenue of $3,629M (+48.5% year over year), gross margin of 52.3% (up from 37.4%), net income of $1,294M (+165%), and diluted EPS of $5.58. Seagate also guided fiscal Q1 2027 to revenue of $4.1 billion (plus or minus $100M) and non-GAAP diluted EPS of $7.30 (plus or minus $0.20) — at the midpoint, 13% sequential revenue growth with implied non-GAAP net margin near 40%.

Revenue Deep Dive​

Seagate reports a single operating segment, so the mix story lives in the 10-K's MD&A splits: 80% of FY2026 revenue came from Data Center customers, up from 75%, with Edge IoT holding the remaining 20%. In dollars, that is roughly $9.8 billion of Data Center revenue (+43%) against $2.4 billion of Edge IoT (+7%) — the AI build-out did essentially all of the growing.

The channel and geography rows barely moved, which is itself the point: this was not a story about finding new customers in new places. It was about selling far more exabytes to the same cloud customers. Nearline HDD shipments hit 695 exabytes, up 40% from 497, while non-nearline volume slipped 4% to 94 exabytes. Total shipments of 789 exabytes grew 32.6% — a touch slower than revenue's 34.1%, implying blended pricing was up about 1%. Seagate grew on volume, not price.

Management's own wording supports that reading. "Our performance is being driven by robust cloud data center demand and disciplined execution, and we see the momentum continuing in 2027," chief executive Dave Mosley said in the release, adding: "As AI accelerates data generation and its value, we see durable long-term demand for mass capacity storage. Seagate is well positioned to address strengthening exabyte demand through our Mozaic platform and differentiated HAMR technology roadmap." Note what is absent: no "demand exceeds supply," no "pricing momentum," no "supply constrained." For a quarter this strong, the restraint is conspicuous — the ledger confirms growth came from utilization and mix, and management is not claiming price increases it cannot show. The receivables line backs the demand story without overstating it: receivables grew 60% to $1,534M on 34% revenue growth, consistent with a back-half-loaded year rather than channel stuffing, since the fourth quarter ran at a $3.6B quarterly pace against a $3.0B average.

The Margin Story​

Four years of margins show a company climbing out of a hole and then sprinting past its old highs:

The mechanics are straightforward. Cost of revenue rose $740M (+12.5%) while revenue rose $3,098M (+34.1%) — factories, once paid for, mint margin on incremental wafers and sliders. Operating expenses barely moved: product development up 4.3% to $755M, marketing and administrative up 2.9% to $577M. Seagate absorbed a $105M litigation-settlement charge and a $151M loss on early debt redemptions and still expanded operating margin by nearly 13 points.

The $105M charge deserves one sentence: it is Seagate's net share of a $175M settlement of a securities class action covering 2020–2023, with insurers paying about $70M; the court granted preliminary approval in July 2026 ahead of a November final hearing. It sits above operating income, so the 33.6% operating margin is after the charge, not before it. Below the line, the $151M debt-transaction loss is the toll paid to retire $1.4 billion of debt early — expensive in the year, accretive to every year after, since interest expense already fell from $321M to $284M.

One caution for the margin celebration: the tax line is normalizing. The provision jumped from $44M (a 2.9% effective rate) to $506M (13.7%), and cash taxes paid were just $40M. The Singapore structure still shelters plenty, but the gap between provision and cash is a future headwind, not a permanent gift.

The One Big Question: Is the $4.1B Guide a Run-Rate or a Peak?​

Seagate's Q1 FY2027 guide — $4.1B of revenue and $7.30 of non-GAAP EPS — implies something remarkable: roughly $1.66B of non-GAAP net income on about 227M shares, or a 40% net margin, up from 36% in Q4. Management is guiding continued expansion, not a plateau. The question is whether HAMR-based Mozaic drives (up to 44TB per unit, 3.5PB systems) constitute a technology cycle that sustains this, or whether we are looking at the top of a memory-adjacent supercycle that always ends the same way.

A peer glance frames the stakes. Memory is earning monopoly-like rents right now while storage equipment earns mortal ones:

Seagate sits between the equipment makers and the memory oligopoly — better than picks-and-shovels, nowhere near supercycle memory. That is the bull case in one table (share gains and HAMR mix can push margins toward memory-like levels on the Data Center book) and the bear case too (disk drives have never held pricing power through a downturn; FY2023 gross margin was 18%). The falsifiable test arrives with Q1: nearline exabytes must keep growing at 30%+ and Data Center mix must hold near 80%, or the guide is mix-and-timing rather than a cycle. Watch also the one customer that bought 14% of FY2026 revenue — concentration this high means a single qualification decision moves the quarter.

Tracking a $12B Company in Plain Text​

Double-entry bookkeeping forces every dollar to reconcile, which is exactly what an earnings season needs: no non-GAAP fog, no adjusted anything — revenue minus expenses must equal net income to the dollar. Read how we model every company for the conventions, then look at Seagate's FY2026 as the ledger records it (income is a credit, so it carries a negative sign):

; FY2026 Income Statement — fiscal year ended July 3, 2026 (53 weeks)
; Check: −12195 + 6637 + 755 + 577 + 105 + 27 + 404 + 506 + 3184 = 0 ✓
 
2026-07-03 * "Seagate Technology Holdings plc" "FY2026 Income Statement"
  Income:Revenue                         -12195 MUSD  ; Data Center 80% / Edge IoT 20% of revenue; nearline HDD exabytes 695 of 789 shipped (10-K MD&A)
  Expenses:CostOfRevenue                   6637 MUSD
  Expenses:ResearchAndDevelopment           755 MUSD  ; product development
  Expenses:SellingGeneralAdministrative     577 MUSD  ; marketing and administrative
  Expenses:OtherNet                         105 MUSD  ; legal settlement, Note 12 ($175M securities settlement, ~$70M paid by insurers)
  Expenses:OtherNet                          27 MUSD  ; restructuring and other, net
  Expenses:OtherNet                         404 MUSD  ; other expense, net: interest expense 284 less interest income 30, debt-transaction loss 151, less other gain 1
  Expenses:IncomeTax                         506 MUSD
  Equity:Adjustments                       3184 MUSD  ; net income offset (RE set by balance assertion)

The balance-sheet number that tells the narrative is the equity repair. Seagate carried negative total equity for three straight years — a ($1,491M) deficit as recently as FY2024 — because decades of buybacks and dividends outran cumulative earnings. In FY2026, retained earnings absorbed $3,184M of net income, and total equity swung to +$2,167M while total debt fell from $4,995M to $3,565M. Deleveraging and profit retention repaired both sides of the balance sheet in a single year. Cash ended at $1,704M, up $813M, after returning $810M to shareholders — and the buyback, dormant for years, restarted with $176M against a $4.8B remaining authorization.

The Multi-Year Arc​

Five years of Seagate is a full cycle in miniature: boom, bust, trough, recovery, boom.

Revenue fell 44% from FY2022 to FY2024, then nearly doubled in two years. The FY2023 loss combines the demand trough with the $300M BIS settlement penalty over Huawei shipments — a one-off that still echoes, since Seagate pays $15M a quarter through 2028 ($75M paid in FY2026, $120M still accrued). Note what management did through the hole: it kept paying the dividend every year (about $600M annually), cut capital expenditure to $254M in FY2024, and waited. Capex is now $569M and property, plant and equipment is growing again ($1,657M to $2,034M) — the factory is being fed because the orders are real. The March 2025 Intevac acquisition ($88M, thin-film deposition equipment) slots into the same logic: owning more of the HAMR supply chain while the cycle pays for it.

The Verdict: Bull vs. Bear​

Bull Case

• The Q1 FY2027 guide ($4.1B revenue, $7.30 non-GAAP EPS) implies ~40% net margin — management is guiding acceleration, and Q4's 52.3% gross margin says the mix supports it.
• Nearline exabytes grew 40% at flat pricing, so any future pricing power drops straight to gross margin; Data Center is 80% of revenue and still rising.
• HAMR/Mozaic plus the Intevac supply-chain tuck-in give Seagate a technology lead in mass-capacity drives that cloud customers must qualify over long cycles — sticky share.
• $3.1B of free cash flow covers the $634M dividend nearly five times over, $1.4B of debt was retired, and $4.8B of buyback authorization remains — capital returns have room to grow.
• The balance sheet crossed from a ($453M) deficit to +$2,167M equity in one year; interest expense is already falling ($321M to $284M) with more to come.

Bear Case

• One customer is 14% of revenue, and the numbers do not yet support any claim of pricing power — management's own release avoids the words, and FY2023's 18% gross margin shows how fast this reverses.
• The last cycle cut revenue 44% peak to trough; non-nearline exabytes already shrank 4%, and HDDs face structural pressure everywhere outside mass-capacity cloud.
• $225M of legal settlements sit accrued with a November 2026 final-approval hearing pending, and the remanded Lambeth patent case heads back for a new trial — the one-offs may not be one-offs.
• The 13.7% effective tax rate (vs. 2.9% last year) is still well below cash reality converging: $506M provisioned vs. $40M paid is a timing benefit with an expiry date.
• The 53-week year flatters FY2026 growth by ~2 points, and Q4's 52.3% margin annualizes to a level no disk-drive maker has ever held — mean reversion is the base rate.

Our Take

Seagate earned this re-rating the hard way — by surviving FY2023–FY2024 without cutting the dividend and keeping the factory ready for the wave that arrived. The FY2026 ledger shows genuine operating leverage, real cash conversion, and a repaired balance sheet, and the Q1 guide dares the market to call the top. We side with the cycle lasting through FY2027: nearline demand tied to AI data-center builds has multi-quarter visibility, HAMR mix is still early, and the deleveraging flywheel (less interest, more buybacks) compounds from here. But this is a cyclical peak until proven otherwise — the bear case is simply the last cycle, and it is only three years old. Size it as a cycle, not a compounder, and let the exabyte count be the stop-loss.