Report contents
Industry landscape
Where the company competes: three suppliers sell nearline capacity worldwide
Market position: concentrated. Seagate sells hard drives into a market with three suppliers, and that structure does more to set its pricing and returns than anything management controls internally. The arena is not "storage," which spans NAND flash, tape and cloud object storage. It is nearline hard drives: high-capacity 3.5-inch units that hyperscale data centers fill with data that has to sit cheaply and stay online. Seagate reported $12.2 billion of revenue in fiscal 2026, up 34% year over year, and non-GAAP gross margin of 46.1%, up from 35.8%, driven by higher nearline exabyte shipments, firmer pricing and a richer mix of high-capacity drives.
The value chain runs from precision components to a handful of very large buyers. Seagate designs and assembles the drive and makes the two hardest parts itself: the read/write heads and the magnetic media. It buys motors, suspensions, controllers and substrates from outside suppliers. Downstream, hyperscalers negotiate directly and take the largest share of nearline capacity; server OEMs bundle drives into systems; distributors serve NAS, surveillance and small-business buyers. Buyer concentration is high, but with three sellers the bargaining leverage is roughly balanced, which shows up in the margin line above.
The geographic boundary of competition is global. A drive qualifies once at a hyperscaler's design and then ships to whichever data center region needs capacity, so share is won or lost in a single worldwide pool. Manufacturing is the localized part. All three suppliers build in a small number of Asian sites, and Toshiba's current expansion is in the Philippines, TrendForce reported in October 2026. That matters for the moat: research spending on areal density, the number of bits packed per square inch, is spread across global exabyte volume, so a supplier with a shrinking shipment base cannot fund the next generation.
Consolidation produced this structure. Samsung's drive business went to Seagate and Hitachi's to Western Digital, leaving three sellers of any consequence.
| Supplier | Recording technology | Projected 2028 share of HDD supply |
|---|---|---|
| Seagate | HAMR, developed in-house | ~45% |
| Western Digital | HAMR, partly externally sourced | ~43% |
| Toshiba | MAMR | ~12% |
Shares are TrendForce projections published in October 2026, not current figures. HAMR (heat-assisted magnetic recording) uses a laser to heat the disk surface so bits can be written smaller and packed more densely; MAMR is a competing approach that gains less density per platter.
Barriers to entry are high, and the best evidence is what a determined incumbent gets for its money. Toshiba is spending about JPY 60 billion on capacity, which TrendForce reckons exceeds 70% of Western Digital's annual capex. Because investment takes 12 to 24 months to reach output, Toshiba's 2027 production rises only 23% to 250 exabytes before jumping 52% to 380 exabytes in 2028. That effort lifts it to roughly 12% share. A challenger starting from zero would need the head and media expertise, a multiyear customer qualification cycle and a capacity commitment of similar scale to win a foothold. None has tried in over a decade.
Two things in the market definition remain unsettled. First, sizing. Mordor Intelligence estimates the HDD market at $51.8 billion in 2026, growing about 6% a year to $69.7 billion by 2031. Seagate alone booked $12.2 billion; even allowing for two rivals of comparable scale, the three together look well short of that figure, so the estimate probably uses a broader definition than factory revenue. Exabyte shipments and supplier-reported revenue are the more reliable yardsticks here. Second, substitution. QLC NAND has taken the low-capacity end of the market for years, and where it stops is a live question for the nearline franchise; the evidence so far only shows that the highest-capacity tiers keep moving toward hard drives as areal density advances. What would change this assessment is a fourth supplier entering at scale, or a rival matching Seagate's areal density lead, neither of which the current record shows.
Competitive advantage
Does the company have a moat? Yes, a three-supplier structure protects it
Seagate has a real moat, and it is Fairly Strong rather than strong. The protection comes from an industry that has supported three hard-drive makers for more than a decade without admitting a fourth. What the moat does not do is smooth the cycle: the operating margin ran from negative in fiscal 2023 to 33.5% in fiscal 2026, and anyone who reads the second number as the moat's strength will overpay for it.
Two tests settle this. Concentration shows whether entrants are being kept out. Sustained returns show whether incumbents keep the profits that concentration creates. Both have to hold at once, because a concentrated industry that competes its returns away has no moat, and a profitable firm in a crowded industry is usually just executing better than rivals it will eventually be copied by.
Market position: about 45% of supply, level with Western Digital
Market position: Fairly Strong. TrendForce's October 2026 projection puts Seagate at roughly 45% of hard-drive supply in 2028, Western Digital at about 43% and Toshiba near 12%, measured in exabytes shipped. Seagate sits at or near the top of a three-firm market, which is where a moat usually shows up. The conventional threshold in this framework is more than 40% share plus a clear first place; Seagate clears the share level but not the clear lead, since Western Digital is within a couple of points.
The consolidation that produced this structure is old, and that is the useful part. Samsung's drive business went to Seagate, Hitachi's to Western Digital, and no one has built a new head-and-media operation at scale since. Seagate does not publish its own exabyte share and TrendForce's public figures are forward projections, so a clean five-year share series is not available from either source.
The stability test still gets an answer from a different direction. Toshiba is spending about JPY 60 billion on capacity, an amount TrendForce reckons exceeds 70% of Western Digital's annual capex, and the payoff is a move to roughly 12% share by 2028. A technically capable incumbent spending at that scale and gaining a few points of share is the entry barrier working as intended.
That same projection carries the warning. Seagate and Western Digital converge to near-parity by 2028, so the industry stays closed while Seagate's lead inside it narrows.
Profitability: high through the cycle, violent within it
Profitability: Fairly Strong. Return on equity is unusable here. Buybacks and losses left shareholders' equity negative at the end of fiscal 2024 and fiscal 2025 (negative $1.49 billion and negative $0.45 billion), which turns ROE into noise. Return on invested capital is the better tool, though it carries its own distortion: years of repurchases shrank the capital base, so the ratio flatters the return on gross capital employed.
| USD bn unless stated | FY2022 | FY2023 | FY2024 | FY2025 | FY2026 |
|---|---|---|---|---|---|
| Revenue | 11.66 | 7.38 | 6.55 | 9.10 | 12.20 |
| Gross margin | 29.7% | 19.2% | 23.4% | 35.2% | 45.6% |
| Operating income | 1.96 | (0.34) | 0.45 | 1.89 | 4.09 |
| Net income | 1.65 | (0.53) | 0.34 | 1.47 | 3.18 |
| Free cash flow | 1.28 | 0.63 | 0.66 | 0.82 | 3.10 |
| ROIC, est. | 37% | (4%) | 21% | 48% | 85% |
Fiscal 2026 ended July 3, 2026. ROIC figures are estimates built from reported EBIT, each year's effective tax rate and net debt.
The five-year record reads as two different businesses. From fiscal 2022 to fiscal 2024, revenue fell 44%, operating income went from $1.96 billion to negative $0.34 billion, and free cash flow never exceeded $1.28 billion. From fiscal 2025 to fiscal 2026, revenue rose 34% and operating income reached $4.09 billion. Cumulatively, Seagate earned $8.05 billion of operating income on $46.9 billion of revenue across the five years, a 17.2% operating margin, and generated $6.49 billion of free cash flow.
Estimated ROIC averages about 37% over the five years, which clears any reasonable cost of capital, and that is what the framework asks. The spread is the caveat: negative in fiscal 2023, above 80% in fiscal 2026. Returns swinging that widely are being set by supply and demand timing inside the oligopoly; they are not the product of a stable pricing formula.
Overall: a structural moat wrapped around a cyclical earnings stream
Fairly Strong. Both tests pass, with reservations on each. Concentration has held for more than a decade, and the entry barrier is demonstrated by the cost of Toshiba's attempt. Through-cycle returns sit well above the cost of capital, but no single year's margin should be annualized. The September 2026 quarter, at 57.3% gross margin and $1.76 billion of net income on $4.23 billion of revenue, is a peak observation, not a run rate.
This falls short of Strong because Seagate does not dominate its market; it shares it almost evenly with one rival. A fourth supplier, a Western Digital lead in areal density, or evidence that buyers have learned to play the three vendors off against each other harder than they managed in fiscal 2024 would push the rating down. A durable pricing mechanism would push it up, and the industry has not shown one.
Where the advantage comes from: process depth is the one real source
The barrier to entry is shared by all three suppliers, so it explains the industry's returns better than it explains Seagate's. Inside the oligopoly, Seagate's own edge rests on manufacturing, and it is thinner than the industry structure around it.
Manufacturing: heads and media made in-house, and a density lead (Fairly Strong)
Cost per terabyte governs nearline economics. Areal density, the number of bits packed per square inch, sets how many platters and heads a drive needs to reach a given capacity. A supplier that leads on density builds the same 30-terabyte drive with fewer platters and fewer heads, so its bill of materials is lower at the same price, or it ships a capacity point a rival cannot match and takes the premium tier of a hyperscaler's order.
Seagate designs and makes its own heads and media, the two process-intensive steps, and it shipped heat-assisted magnetic recording (HAMR) ahead of Western Digital. HAMR uses a laser to heat the disk surface so bits can be written smaller and packed more densely. Vertical integration matters because it keeps the yield learning inside the company. Head and media process know-how is where cost differences actually accumulate, and it does not travel well through a supplier relationship.
The mechanism compounds within a generation. Density leadership wins the highest-capacity qualification at the largest buyers first, those orders fund the next density step, and the sequence repeats. Western Digital's HAMR is partly sourced externally, according to TrendForce, which puts some of its process learning outside its own walls.
Durability is where the rating gets tested. HAMR is no longer proprietary in any meaningful sense: Western Digital ships it too, and Toshiba is pursuing the competing MAMR approach, which gains less density per platter. TrendForce's projection that Seagate and Western Digital finish 2028 within two points of each other implies the density lead gets competed down to near-parity. A one-generation lead in a three-firm industry is worth something, roughly a richer mix and first call on the top capacity tier, but it is not a multiyear franchise. The checkable markers are announced areal density per platter and qualification wins at the largest hyperscalers.
Scale: four times Toshiba's exabyte base, near parity with Western Digital (Moderate)
Nearline drives are a fixed-cost business. R&D for the next recording technology and the fabs that make heads and media get paid for once and then spread across exabytes shipped. Seagate's roughly 45% share against Toshiba's 12% means close to four times the base over which to spread the next technology. That gap is the scale advantage that matters.
It is nearly absent against Western Digital. Relative share drives this, and a 45-to-43 split produces no meaningful cost edge.
Capital intensity cuts against the usual scale story. In fiscal 2026 Seagate generated $3.67 billion of operating cash flow and $3.10 billion of free cash flow, which implies capex near $0.57 billion, under 5% of revenue. The fabs are largely built, so incremental exabytes cost little and a demand upturn turns into cash quickly. That is good for owners, but it also means the barrier is not "we can outspend you." Toshiba's JPY 60 billion buys a viable third position, which tells you the capital hurdle is high enough to deter a newcomer and low enough that the incumbents can clear it whenever returns look attractive.
Scale also lacks the reinforcement that would make it decisive. Without customer captivity, a challenger that could fund capacity could also win the sockets. Scale here deters entry more than it protects price. Where the current returns come from is the shortage, and shortages end.
Customer captivity: buyers qualify three suppliers on purpose (Weak)
Nearline drives carry switching costs on paper. A drive gets designed into a hyperscaler's rack, and requalifying a supplier takes quarters of joint testing, which is why the same three vendors keep the sockets.
The buyers manage those costs deliberately. Hyperscalers qualify multiple suppliers for every slot, write specifications that make drives comparable, and negotiate with full visibility into each vendor's capacity. The margin record is the evidence. Seagate's gross margin went from 19.2% in fiscal 2023 to 45.6% in fiscal 2026 on essentially the same product family and the same customers. That swing reflects who holds scarce capacity in a given year and shows that buyers set the terms when supply is long. There is no brand preference in a data center procurement decision and no network effect in a drive.
Durability: substitutes at the low end, a supply response at the top
The substitute is flash. QLC NAND has taken the low-capacity end of the market over several years, and the open question for nearline is where it stops. Hard drives hold the highest-capacity tiers on cost per terabyte, and the write-endurance limits of NAND make it a poor fit for some write-heavy workloads. Those are real protections, but they are boundary conditions that move with both cost curves. If NAND's cost per terabyte falls faster than areal density improves, the nearline pool shrinks from below.
The nearer risk is ordinary. A 57.3% gross margin in the September 2026 quarter, on $4.23 billion of revenue and $2.06 billion of operating income, is an invitation for the other two suppliers to add capacity. Toshiba's exabyte additions land in calendar 2027 and 2028, the same window in which the projection has the top two converging. If that supply meets slower hyperscale spending, margins revert toward the fiscal 2024 pattern. That would not break the moat; it would show what the moat is worth, which is protection from new entrants without protection from the incumbents' own capacity decisions.
Two developments would change the assessment. A fourth supplier entering at scale, which no one has attempted in over a decade, would weaken the structural case. So would a Western Digital density lead that forces Seagate to compete on price for the top capacity tier. Neither appears in the record yet. What the record shows is a three-firm industry with high entry barriers and a wide profit cycle. The structure has proven durable; the margins have not.
Generated by DeepVal AI. Data and models may contain errors or be outdated. For research only; not personalized investment advice.
Continue reading the full report
- Comparison and summaryHow Seagate Technology PLC compares with its closest rivals, and our moat rating
Free to sign up · 20 research points · unlocking this report uses 10
20 points cover 2 full reports. No card needed.
Research reflects its generation date. Preview the first two modules.
More on Seagate Technology PLC
Seagate Technology PLC
Business, moat and key questions
Open company page