I rate Seagate Technology Holdings plc (STX) a “Sell.” I admit the operating results are excellent, as FY Q4 revenue increased 48% YoY to $3.63 billion, and non-GAAP gross margin expanded to ~53%, with free cash flow at $1.12 billion. For FY 2026, revenue increased 34% to $12.2 billion and free cash flow reached $3.1 billion. Moreover, management guides for a strong September quarter with revenue at $4.1 billion and an operating margin of ~50%.
My disagreement is with the durability implied by those numbers. Seagate remains a manufacturer of storage capacity in a market that is (long-term) in my view dominated by price competition, customer concentration, technological substitution, and large swings in supply and demand. HAMR materially improves Seagate's cost position, but it does not remove those characteristics. More importantly, Seagate's own disclosures indicate that pricing and mix, not HAMR alone, drove the recent margin step-up. Western Digital simultaneously reaching even higher gross margins without comparable HAMR penetration adds to my concern.
My base case sees shares at fair value closer to ~378.
The Margin Expansion Is Primarily a Pricing and Mix Story
Arguably, the most important insight in Seagate's 2026 10-K is the company explicitly stating that the 11 percentage-point increase was primarily driven by pricing actions and a product-mix shift toward higher-capacity products. Revenue increased 34%, primarily due to higher nearline shipments and favorable pricing.
Nearline exabytes increased from 497 EB to 695 EB, or 40%. The data center increased from 75% to 80% of revenue. Non-nearline shipments actually declined from 98 EB to 94 EB. Virtually all incremental capacity came from nearline data-center HDDs.
On the Q4 call, Morgan Stanley noted that Seagate's price per exabyte increased approximately 10% YoY in June and that September guidance appeared to imply something close to 20%. CFO Gianluca Romano did not reject the premise. He instead explained that the gap between supply and demand had widened and that incrementally available output was being sold at a “very good price.”
This is favorable for near-term earnings. It is also important analytically, as Seagate is not merely generating more bits at lower cost. It is currently benefiting from more volume, better mix, and higher price simultaneously. That may be unusually cyclical.
Western Digital Shows This Is an Industry Phenomenon
If Seagate's 50%+ gross margin were primarily a proprietary return on HAMR, I would assign significantly greater durability to current profitability. Western Digital makes that interpretation difficult. WD reported Q4 fiscal 2026 revenue growth of 44% and a 54% non-GAAP gross margin, versus Seagate's 53%. WD's operating margin reached 44.2%, and quarterly free cash flow was $1.28 billion.
Yet Seagate is substantially further ahead in HAMR deployment. HAMR represented approximately 40% of Seagate's nearline exabyte run rate exiting fiscal 2026. WD still expects its first commercial 40 TB-44 TB HAMR products to move into volume production around 2027.
Two HDD vendors with different recording technologies have simultaneously reached gross margins above 50%. The common factors are strong hyperscaler demand, constrained supply, high-capacity mix, and disciplined pricing.
But Seagate is not as clean a bet for the AI-CapEx thesis as Nvidia Corporation (NVDA) or Micron Technology, Inc. (MU).
I agree with the secular data-growth thesis. AI increases data creation, retention, and reuse. HDD remains economically attractive for mass-capacity storage. That supports long-run exabyte growth. What I would not do is translate GPU deployment directly into HDD demand. Management does not do that either.
Asked specifically about HDD attachment to agentic AI and KV-cache workloads, Mosley said requirements depend on the application and that the outcome “could be a lot; it could be a little.” He also characterized several of these new applications as difficult to forecast. In other words, the AI boom does not make Seagate's demand function equivalent to Nvidia's.
HDD Cyclicality Did Not Disappear
The strongest empirical argument against treating today's margins as permanent is Seagate's own recent history. Fiscal 2022 gross margin was approximately 30%. In fiscal 2023, mass-capacity demand collapsed. HDD shipments fell from 631 EB to 441 EB, revenue declined 37% from $11.7 billion to $7.4 billion, price per TB fell from $17 to $15, and gross margin collapsed from 30% to 18%. Gross margin then recovered to roughly 23% in fiscal 2024, 35% in fiscal 2025, 46% in fiscal 2026 and 52.7% in Q4 2026.
To highlight it more cleanly, isolated:
30% -> 18% -> 23% -> 35% -> 46% -> 53%.
There is clearly little stability in the profitability picture, implying a cyclical hardware industry that currently experiences an exceptionally favorable point in its cycle.
2028 Allocation Is Valuable, but It Is Not the Same as Revenue Backlog
Management says the vast majority of nearline exabytes are allocated into calendar 2028 and discussions are extending into 2029. Contracts covering calendar 2027 define both product configuration and pricing. This materially improves near-term visibility.
But note that there is still a strong distinction between allocation and unconditional backlog. Seagate's customer relationships historically operate through master purchase agreements, forecasts, and purchase orders. The company has strengthened longer-term commitments and introduced cancellation charges, but its filings still warn that major customers may reduce, defer, or cancel purchases.
Customer concentration is also increasing. One customer represented approximately 14% of fiscal 2026 revenue. That gives hyperscalers significant economic leverage when supply normalizes and potentially makes the cyclical swings in profitability more aggressive.
I assign high confidence to fiscal 2027 demand, but I assign less confidence to extrapolating today's environment through 2028-2029.
Valuation Compounds the Risk
I use a residual earnings model to value STX. For FY27: I would not fight the cycle yet. Q1 guidance is already $4.1B revenue / $7.30 EPS and implies roughly 50% operating margin; management also says FY27 revenue growth should exceed FY26's 34% and margins should improve sequentially. ~$35 EPS is reasonable. For FY28, I see ~$31 EPS, as I stop assuming further pricing expansion, with operating margin falling toward 40% on ~21 billion revenue. Post-2028, ~$20-25 EPS is my normalized earnings power. That is still well above historical Seagate economics; I am explicitly giving credit for HAMR, higher nearline mix, and lower unit costs. In addition, I anchor on a ~10% cost of equity and a ~4% terminal growth rate. That returns a fair implied target value of $378 per share.
Investor Takeaway
I rate Seagate “Sell.” FY2027 earnings will likely remain exceptional, but I believe investors are capitalizing on an unusually favorable combination of volume growth, mix, and pricing as if it were structural. Seagate's own history argues otherwise, while Western Digital's simultaneous margin expansion suggests today's profitability is industry-wide rather than uniquely HAMR-driven. I give Seagate substantial credit for better technology and structurally higher earnings power, but at a fair value of roughly $378 per share, I see insufficient compensation for the risk that pricing and margins normalize from current peak levels.
