Seagate Technology Holdings plc (STX) stock intrinsic value fair value estimate & owner earnings analysis

Seagate makes hard disk drives (HDDs) — most of it now the high-capacity nearline drives that cloud and AI data centers use for bulk storage. In fiscal 2026, 80% of revenue came from data-center customers, up from 75% the year before, and one customer alone accounted for 14% of revenue. Sales split roughly 50% Americas, 40% Asia Pacific, 10% Europe/Middle East/Africa. The company manufactures across six countries, including China, and some recording-head materials (rare-earth elements) are sourced predominantly from China — a supply-chain point worth knowing given recent export-control friction, even though the business itself is Irish-domiciled and not China-controlled. The HDD industry has consolidated to three real competitors (Seagate, Western Digital, and Toshiba), which matters because it has made pricing far more disciplined than in prior decades. Seagate is mid-way through rolling out a new recording technology (HAMR, on its Mozaic drive platform) that packs more data onto each disk; management has said capital spending in fiscal 2027 will step up specifically to support this ramp. How we valued it. Seagate's cash flow is extremely cyclical — revenue went from $11.7 billion (fiscal 2022) to $6.6 billion (fiscal 2024) and back to $12.2 billion (fiscal 2026) in the space of four years, and our owner-earnings figure (operating cash flow, adjusted to strip out swings in receivables/inventory/payables and stock compensation, and after real reinvestment spending) swung from about -$3 per share in the depressed fiscal 2023 to nearly $14 per share in the blowout fiscal 2026. Because a single recent year — especially a record one — is a poor guide to what this business earns through a full cycle, we anchored the valuation to the median of the last ten fiscal years (about $4.40/share), which spans a prior upswing, a severe downswing, and the current AI-storage-driven recovery. A useful reality check: Seagate's revenue in fiscal 2026 is barely 9% higher than it was ten years earlier, despite two full boom-bust cycles in between — the business has grown earning power mostly through cost discipline and buybacks, not a rising tide, and we built our base case around that discipline rather than assuming the current boom simply continues. The bear case assumes the pattern that has repeated twice in the last decade repeats again: a sharp post-peak digestion (as happened in fiscal 2019 and fiscal 2023), made worse by the customer concentration noted above and by continued share-count drift (no buyback support) during the downturn. The base case assumes the HAMR ramp Seagate has already committed capital to delivers a few years of above-trend growth before settling back into the roughly flat, cyclical-but-disciplined pattern the business has shown for a decade. The bull case assumes that ramp executes cleanly, debt keeps coming down (already cut from about $5.0 billion to $3.6 billion this past year), and the $4.8 billion in unused share-repurchase authorization gets put to work. In all three cases we treated the three-player industry structure as a durable, already-proven fact, so none of the cases assumes the business's earning power decays to zero over time — but none of them assumes indefinite growth either, since Seagate's own decade-long revenue record doesn't support that. On this basis we get roughly $58/share in the bear case, $91/share in the base case, and $140/share in the bull case. Because the swings here are real and repeat on a multi-year cycle, and because the business still carries meaningful debt and a concentrated customer base, the margin-of-safety price sits at a significant discount to the base-case estimate, and it assumes the HAMR ramp proceeds in the quarters ahead without the qualification delays management has flagged as a risk. A stumble there would look a lot like the last down-cycle. What would make this estimate wrong: if the current AI-storage demand proves durable rather than cyclical, both the base and bull cases understate value; if HAMR qualification slips or a large customer pulls back sharply, the bear case is the better guide. Note also that our owner-earnings figure adds back depreciation and other non-cash charges dollar-for-dollar, as the standard method requires, and we separately capped how much of Seagate's capital spending we treat as growth (rather than upkeep) using the company's own ten-year average capex-to-revenue ratio — so the swings above reflect real cash generation, not an accounting artifact.