A year ago, Seagate Technology Holdings (STX) was trading near $95 a share. Today it touched $802.13 — a new 52-week high — before settling back to around $782. That’s a 700%+ gain in twelve months from a company most investors had filed under “legacy hardware.” The market cap now sits in the $175–$178 billion range. Nobody is calling it legacy anymore.

What happened here isn’t complicated, though the scale of it is hard to fully process. Seagate manufactures the physical drives that store AI training data, model weights, and inference outputs at hyperscale. The world built a lot of AI compute over the past two years. It turns out you need somewhere to put all of that data. And the companies supplying that storage ran out of capacity before demand ran out of momentum.

That supply-demand mismatch is the engine behind everything that follows.

The Demand Side: Why Storage Is Sold Out

Hard disk drive prices have risen roughly 46% since September 2025. That’s not a temporary pricing event — it’s the result of a market where production capacity has been fully committed. Western Digital has reportedly allocated its entire 2026 HDD supply. Seagate’s own nearline capacity is on allocation through calendar 2026, with supply agreements in place with major cloud customers through 2027. Multiple cloud providers are already negotiating demand projections for 2028.

Seagate’s Chief Commercial Officer described the current pricing environment as “the new normal.” That phrasing is worth sitting with. It’s not a company talking up its own quarter. It’s a company with full visibility into forward demand telling you, in plain language, that this isn’t a spike. The hyperscalers — Amazon, Microsoft, Google, Meta — have not pulled back on AI data center investment. If anything, the urgency has accelerated as competitive pressure among cloud platforms intensifies. Every new AI workload requires training data and inference capacity. All of it needs to be stored somewhere.

The economics of flash storage remain prohibitive for the cold storage and nearline workloads at the core of large-scale AI operations. Seagate’s CTO has stated publicly that the cost-per-byte gap between HDDs and SSDs is unlikely to close over the next decade. At current scale, HDDs hold a 5-to-1 or greater cost advantage per byte. That gap is structural, not cyclical, and it defines why Seagate’s product sits at a chokepoint in the AI infrastructure stack that no one has yet found a cheaper way around.

Q3 Fiscal 2026: Eight Consecutive Quarters of Revenue Growth

Seagate reported fiscal Q3 2026 results on April 28 for the quarter ended April 3, 2026. Revenue of $3.11 billion came in ahead of the $2.96 billion analyst consensus, up 44% year-over-year. Non-GAAP diluted EPS of $4.10 beat consensus of roughly $3.50 by 17%. GAAP diluted EPS was $3.27. It was the eighth consecutive quarter of revenue growth.

The margin story is where the quarter gets genuinely interesting. GAAP gross margin hit a record 46.5%, up from 35.2% in the year-ago period. Non-GAAP gross margin reached 47.0%, versus 36.2% a year earlier. Non-GAAP operating profit grew 30% sequentially to $1.2 billion, or 37.5% of revenue — a level the company described as exceeding its own long-term operating margin target ahead of schedule. Free cash flow for the quarter reached $953 million, a decade high, up 57% sequentially. Operating cash flow hit $1.1 billion. During the quarter alone, the company retired $641 million in debt and returned $191 million to shareholders through dividends and buybacks.

• TTM Revenue: $11.01 billion (+28.92% year-over-year)
• TTM Diluted EPS: approximately $10.54–$10.57
• Q3 Non-GAAP EPS: $4.10 (beat by ~17%)
• Q3 Free Cash Flow: $953 million (decade high)
• Q3 Non-GAAP Gross Margin: 47.0% (vs. 36.2% year ago)
• Debt Retired in Q3: $641 million

For Q4 fiscal 2026, management guided revenue of $3.45 billion (plus or minus $100 million) and non-GAAP diluted EPS of $5.00 (plus or minus $0.20). With Q1 through Q3 non-GAAP EPS already totaling roughly $9.00 or more, full fiscal year 2026 non-GAAP EPS is on course to reach approximately $14 per share or higher. Q4 results are expected on July 27, 2026.

Two variables worth naming. The 5.7% shareholder dilution rate is a real number that affects long-term EPS trajectory — it doesn’t break the bull argument, but it belongs in any serious model. And the debt-to-equity ratio remains elevated, though $641 million in single-quarter debt retirement suggests the company is addressing it aggressively. Fitch recently upgraded Seagate’s credit to investment grade, citing margin expansion and balance sheet improvement.

Mozaic 4+: The Technology Doing the Heavy Lifting

The margin expansion and the supply allocation story both trace back to the same source: Seagate’s Mozaic 4+ platform, currently the only Heat-Assisted Magnetic Recording (HAMR) storage product deployed at commercial scale in the industry. As of March 2026, Mozaic 4+ drives at up to 44 terabytes per drive are qualified and shipping in production volumes to two leading hyperscale cloud providers, with additional customer qualifications underway.

HAMR uses a custom laser integrated into the recording head to briefly heat the disk surface during writes, enabling far higher data densities without sacrificing long-term stability. Seagate designs and manufactures its own laser and nanophotonic components internally — a vertical integration decision that reduces supply chain dependency and compresses customer qualification timelines. It also makes the technology harder to replicate quickly. That vertical integration is not theoretical. It is what produced the 47.0% non-GAAP gross margin in Q3.

In a one-exabyte deployment, Mozaic 4+ delivers approximately 47% better infrastructure efficiency versus standard 30TB deployments, reduces the required data center footprint by around 100 square feet, and cuts annual energy consumption by roughly 0.8 million kilowatt-hours. At hyperscale, those numbers translate directly into meaningful operational cost savings for customers — which is exactly why cloud providers are signing multi-year supply agreements rather than procuring quarter to quarter. Seagate’s roadmap targets scaling from current areal densities toward a future 10TB per disk, a trajectory that would support drive capacities approaching 100TB.

Western Digital’s HAMR-based 44TB drive is not expected to reach customers until 2027 at the earliest. Its current flagship nearline product for 2026 is a 40TB drive using energy-assisted PMR technology. That’s a meaningful product gap, and Mozaic 4+ is expected to represent a majority of Seagate’s HAMR exabyte shipments by the end of fiscal 2027. The window of competitive differentiation is real, and it has not closed yet.

Analyst Targets: The Street Is Chasing the Stock

Following the Q3 report, price target increases came in across the board. Mizuho raised to $875 from $700 (Outperform). TD Cowen raised to $850 from $500 (Buy). Morgan Stanley raised to $767 from $582 (Overweight). Citi raised to $740 from $595 (Buy). Barclays raised to $750 from $625 (Overweight). Evercore ISI raised to $750 from $550 (Outperform).

Here’s what’s interesting about that list: STX, trading between $780 and $802 on May 11, has already moved through or beyond Morgan Stanley, Citi, Barclays, and Evercore ISI’s recently raised targets. The stock outpaced the street’s revised thinking within two weeks of the earnings report. That doesn’t make it wrong to own — it makes it important to understand that the consensus price target is no longer a ceiling for this move. It’s a trailing indicator.

On valuation, the forward P/E sits at roughly 31–32x based on current prices and fiscal 2027 estimates. The PEG ratio of approximately 0.54 is the number bulls keep citing — it suggests the stock is undervalued relative to projected earnings growth even at these levels. Revenue is forecast to grow roughly 15% annually over the next three years, against a 7.3% average for the broader U.S. technology sector. Management raised its medium-term growth outlook to at least 20% annually. Those two figures are not yet reconciled in most models, which may partially explain why institutional accumulation has continued at elevated prices.

Price Structure and Key Levels to Watch

STX set a 52-week high of $802.13 intraday on May 11, 2026. The stock is trading well above its 200-day simple moving average, near the top of its 52-week range. The 30-day return has been north of 60% from the post-earnings base in the $490–$510 range. Velocity that steep rarely continues without a period of digestion. That observation is about risk management, not direction.

• Current 52-week high: $802.13 (May 11, 2026). A sustained close above this level on strong volume opens the path toward the $850–$875 analyst target cluster.
• Near-term support: $750–$766 represents the early-May consolidation range. A retest on declining volume would be considered healthy. A breakdown on heavy volume changes the risk picture materially.
• Moving averages: Price is significantly extended above both the 50-day and 200-day moving averages. The historical spread at current levels increases mean-reversion risk in the near term.
• Volume: Watch for expansion on up days as confirmation of continuation. Volume divergence — price advancing while volume declines — is the early warning sign worth monitoring.
• Momentum: Weekly RSI is elevated. Not a standalone reversal signal, but it narrows the acceptable margin for error at these price levels.
• Key date: Q4 fiscal 2026 earnings are expected July 27, 2026.

Three Scenarios From Here

Bull Case – Hyperscaler capex holds at current levels through 2027. Seagate delivers on Q4 guidance of $3.45 billion in revenue and $5.00 non-GAAP EPS. Mozaic 4+ adoption deepens across the hyperscale customer base, driving non-GAAP gross margin above 47% as higher-density drives displace lower-margin legacy capacity. Fiscal 2027 non-GAAP EPS tracks toward $18–$20 per share. The market continues assigning a secular-growth multiple rather than a cyclical-hardware multiple. Mizuho’s $875 and TD Cowen’s $850 become reference points that the stock absorbs rather than reverses from.

Base Case – STX consolidates in the $750–$820 range over the next four to eight weeks. The post-earnings advance gets digested, implied volatility normalizes, and the stock builds a base before the next catalyst. Q4 results on July 27 come in at or near guidance, providing confirmation without a large incremental upside surprise. Full fiscal year 2026 non-GAAP EPS lands at or above $14. Institutional buyers use any softness to build positions, creating a demand layer under the market price. The stock resumes its upward move as Q4 approaches.

Bear Case – Hyperscaler capex guidance begins to moderate in mid-2026 earnings calls. Cloud customers signal reduced storage commitments beyond 2027, triggering downward revisions to Seagate’s forward revenue model. HDD pricing softens as new supply enters the market or QLC NAND continues improving its economics for nearline workloads. The elevated debt-to-equity ratio becomes a concern in a tightening credit environment. Insider selling activity among senior executives — already a watch item — accelerates. Under these conditions a reversion into the $600–$650 range is plausible, representing an 18–23% drawdown from current levels. The July 27 earnings call becomes critical for understanding whether demand visibility through 2028 remains intact.

What Investors Should Watch Now

STX at $780–$802 after a 700%+ advance from its 52-week low requires different thinking than STX at $150 or $200. The fundamental argument hasn’t deteriorated — by most measures it has strengthened across three consecutive quarters. But position sizing discipline and entry timing matter considerably more at these price levels than they did a year ago.

• For active investors: A pullback into the $750–$766 range on declining volume presents a more defined risk entry point relative to today’s extended levels. The $802.13 high becomes the reference point above for defining risk. Entry closer to support levels allows for better-defined position parameters.
• For longer-term holders: The fundamental case is intact. The Q4 guidance of $3.45 billion and multi-year supply agreements through 2027 provide a credible revenue floor. Consider scaling incrementally rather than adding full exposure at peak price levels.
• Options considerations: Implied volatility has likely expanded post-earnings and may remain elevated ahead of the July 27 report. Covered calls or cash-secured puts at lower strikes may offer more favorable risk-adjusted positioning than outright long exposure at current levels.
• Catalysts to track: Q4 fiscal 2026 earnings on July 27; capex commentary from Amazon, Microsoft, Google, and Meta earnings calls; HDD pricing data; Mozaic 4+ customer qualification updates; and any management commentary on 2028 demand projections.
• Risk level to watch: A decisive weekly close below $700 would suggest a more significant structural break and warrant a full reassessment of the position thesis.

Nvidia gets the AI headlines. The hyperscalers get credit for the infrastructure buildout. But somewhere between the GPU cluster and the model output, there are racks of hard drives storing everything those chips are working on. That’s the part of the AI stack that doesn’t get a keynote presentation. It also doesn’t run out of capacity without causing real problems for everyone upstream.

Seagate figured out, earlier than most gave it credit for, that being the company controlling that chokepoint was worth a great deal. The past year has been the market recognizing that fact. Whether the next 12 months look anything like the last 12 depends almost entirely on whether the demand signals coming out of hyperscaler earnings calls in late July and beyond hold up.

July 27. That’s the next real data point. Everything until then is positioning around a thesis that has not yet been challenged at scale.

For informational and educational purposes only. Not investment advice. Trading involves risk, including loss of principal.