We are long shares of Seagate Technology Holdings PLC (NASDAQ:STX), a leading supplier of hard disk drives (HDDs). HDDs hold almost 90% of the data residing in large-scale data centers, and as a result over 80% of Seagate’s revenue is derived from high-capacity, high-performance drives sold into the major hyperscalers (Amazon, Google, Meta, and Microsoft) and enterprise server and storage OEMs. The prospects for massive, sustained data growth are undeniable, and the incremental storage demand driven by the AI wave has barely begun.
Read hedge fund letters here
Already benefitting from these secular tailwinds and its position in a dominant duopoly, Seagate is executing a generational technology transition to heat-assisted magnetic recording (HAMR) that promises to supercharge density gains and company profitability. Due to the foresight of Seagate management, they are years ahead of primary competitor Western Digital and will thus be the disproportionate beneficiary as HAMR ramps. Seagate has qualified its first hyperscaler (Google), and we expect to hear additional announcements in the coming weeks. Robust HDD demand continues to outstrip supply, so the market is primed for HAMR products.
Seagate’s initial 30TB HAMR offering will ramp in earnest over the next several quarters, but it’s the company’s next-generation 40TB product in 2026 that will really transform HDD unit economics. With a 33% capacity gain and a likely reduction in bill of materials, Seagate’s 40TB products will dramatically expand gross margins as they slash the company’s production cost per terabyte relative to its two HDD peers. By contrast, Western Digital doesn’t expect to ship its first generation 36TB HAMR drive until at least 2027, which is still likely overly optimistic.
HAMR ensures HDDs will remain the optimal technology for the bulk of data center storage for at least the next decade. On a cost/TB basis, HDDs are currently 6x cheaper than flash memory-based solid-state drives (SSDs). Given HAMR-fueled density gains, a flattening of the flash cost curve, and the 9x greater capital efficiency of HDD manufacturing, the value proposition of HDDs is secure, which is why forecasters see little change in HDD deployments.
The industry operates in an effective duopoly, with third player Toshiba hopelessly behind on the technological and cost curve. Price wars are a thing of the past, and the players are increasingly judicious about adding supply. Customer supply agreements now provide multiple quarters of visibility and enable improved supply chain planning, and HDD makers demand a proper return for value delivered. Seagate’s capital allocation has been tremendous for a long time – the company has returned over $8 billion to shareholders over just the past five fiscal years, and there are clear signs that future returns will grow further.
Despite being an obvious beneficiary of the AI and data center boom – the most powerful secular trend of our time – and a clear technology leader, Seagate shares trade at an inexplicable 11x 2026 earnings. The market is failing to anticipate the massive gross margin upside and resultant rich future cash generation. Seagate shares are ripe for a major re-rating and are a double or triple from current levels.
I. Investment Highlights
The Age of HAMR is Upon Us.
Technology transitions in the HDD industry are infrequent but highly impactful. Seagate is on the cusp of ramping its new products based on heat-assisted magnetic recording (HAMR) technology. HAMR promises to re-accelerate HDD bit density scaling, shift product cost curves down, and drive a period of sustained profitability improvement. Relative to the ~10TB average capacity of nearline drives in Seagate’s installed base, HAMR drives offer 3x the storage capacity while consuming about 70% less power per terabyte. For the customer, higher storage density results in more data on a given server, enabling fewer racks and reduced floor space for a given capability, which reduces aggregate power, cooling, compute, and floor space costs. Seagate’s first generation 30TB HAMR drive has been qualified at Google (which arguably runs the most demanding process), and at least one other hyperscaler qualification should be announced in the coming weeks. HAMR shipments ramp in earnest in 2H 2025.
The next generation 40TB HAMR drive, expected to ship in mid-2026, should transform Seagate’s unit economics and result in material additional gross margin expansion. The dramatic 33% increase in capacity will be accompanied by a likely reduction in (or at worst flat) bill of materials, allowing Seagate to offer customers a superior cost/TB and generate the richest gross margins in the company’s history. With HAMR technology already qualified and accepted as viable by customers, hyperscalers will rapidly embrace these products. We think HAMR will account for over half of total exabytes shipped by 2027 and that Seagate’s 2026 and 2027 profit and cash flow will shock analysts.
Given the current robust demand environment and a rational HDD oligopoly, Seagate gross margins have already been rising for the past eight quarters (even in the March quarter despite a $200 million revenue shortfall due to an anticipated supply chain issue). HAMR will supercharge that trajectory.
Importantly, HAMR also promises to transform the competitive landscape. Seagate has been working on HAMR for twenty years, meticulously solving some of the industry’s most difficult engineering challenges and developing the requisite manufacturing expertise to ensure successful productization. By contrast, primary competitor Western Digital at one point abandoned HAMR for an alternative technology (since de-emphasized), has already seen roadmap slippage, and is not expected to begin shipping its first generation 36TB HAMR product until sometime in 2027. And given WD’s lack of demonstrated progress to date, we are skeptical of their ability to close the gap anytime soon. Thus, while WD is coming to market with what it admits is its final generation of products using the older PMR technology, Seagate stands alone as the HAMR technology leader, poised to capture market share and benefit from gross margin expansion for years to come.
Seagate is a Data Center and AI Play.
Almost 90% of all data stored in the cloud resides on HDDs, and the world is increasingly generating and deriving greater value from massive datasets and high-density video content. It is mindboggling to think that some of the most data-intensive applications, such as IoT sensors, video content and security, smart cities, and autonomous vehicles, are still in their relative infancy. The coming exponential growth of data will require zettabytes of low-cost enterprise-grade storage. Revenue from the major hyperscalers represents an estimated 40%-50% of Seagate revenue, and well over 80% of exabytes are shipped in the form of high-density, high-performance HDDs targeting these growth applications.
And we haven’t even mentioned the impact of AI. Many in the industry have struggled to try to size the impact of AI on storage demand, but the consensus is that it will be huge. High performance SSDs are central to model training and deployment (inference) to fully utilize expensive compute resources. But HDDs play a key role in storing massive datasets and saving interim model versions (checkpointing), and they will inevitably be the primary storage destination for model outputs and new content generated. Seagate has noted that, in just 1.5 years from 2022 to mid-2023, AI created 15 billion images, and, by 2028, image and video creation with AI models will grow 167 times. HDDs excel at this type of large object storage. AI training data is also retained on HDDs for the benefit of reanalysis by more powerful systems in the future. Ultimately, AI will utilize and produce tremendous amounts of data, and it will be a boon for both SSDs and HDDs.
HDDs Will Remain the Low Cost Per Terabyte Solution For a Very Long Time.
Tech forecaster IDC predicts HDDs will continue to comprise almost 80% of storage used in hyperscale and cloud data centers through 2028. The main reasons are their dramatically lower cost/TB relative to flash drives and the practical reality that the vast bulk of storage is earmarked as “warm” or “cold” because it is accessed less frequently and less sensitive to latency concerns. SSDs are currently about 6x more expensive than HDDs on a cost/TB basis. IDC and others fully expect this price premium to persist, because the massive increase in areal density offered by HAMR technology will enable continued declines in HDD cost/TB, while the flash cost curve is potentially about to flatten. Additionally, the HDD industry is almost 90% more capital efficient than the flash industry, meaning they spend far less on capital equipment to grow their exabyte production capacity. The idea that the flash industry will choose to spend over $200 billion on the fab capacity required to displace HDD exabyte demand is laughable.
The HDD Industry is an Oligopoly.
The HDD industry structure has never been so attractive. Through consolidation, the sector has been whittled down to an effective duopoly going forward. Number three player Toshiba is still used as a price cudgel by the hyperscalers, but their lack of vertical integration and history as a technology follower mean that they will almost certainly be an also ran as it regards HAMR. The new industry practice of seeking long-term supply agreements with its customers has given Seagate newfound demand visibility and enhanced its ability to manage its supply chain. More important has been the change in tone from managements, with Seagate and Western Digital focused on receiving due compensation for value delivered and maintaining stable and attractive profitability.
The AI Supercycle Could Delay the Next Cyclical Peak Until 2030.
We think the sector’s supply discipline, improved visibility, and rational competitive behavior can dampen deleterious business volatility when the industry hits its next downcycle. However, as we remain in the early innings of a data generation and AI supercycle, we believe that the industry’s cyclical peak is many, many years away. Seagate’s last cyclical peak, defined as the last quarter before sequential exabyte growth turned negative, was in the June 2022 quarter. That cycle began in September 2019, translating to a cyclical upturn of three years. With AI poised to revolutionize much of global business, we anticipate the current cyclical upswing, which began a year ago in the June 2024 quarter, will be double the size of the last upturn. With a cyclical peak potentially as far away as 2029-2030, Seagate is poised to witness a lengthy runway of revenue growth.
Importantly, current Street consensus revenue projections for Seagate assume an industry downcycle sometime in 2026 to 2027. This is despite the complete absence of any evidence of demand slowdown. A delay in this cycle’s peak by two years, in combination with HAMR-fueled growth and margin expansion, sets up Seagate for a period of sustained financial outperformance, both on the top line and particularly with gross margins, a focus metric historically. For fiscal 2027 (FYE June), we model Seagate exceeding consensus revenue and gross margins estimates by 25% and 456 basis points, respectively. This should catalyze significant share price appreciation.
HAMR Will Drive Expanded Capital Returns.
Seagate has an impressive history of generating 30-40% returns on invested capital and returning significant amounts of cash to shareholders. In fact, Seagate annually pays out over $600 million in dividends and, in the decade prior to the post-COVID downturn, spent an average of $1.2 billion annually on share repurchases. Moreover, the company is largely expanding effective production capacity through areal density gains as opposed to massive incremental capital expenditures. Seagate management has also demonstrated sound financial management and stewardship of corporate resources, staying lean during upturns, remaining profitable during downcycles, and only pursuing selective, scale-enhancing M&A within its technology wheelhouse. Powered by HAMR-driven revenue growth and margin expansion and benefitting from a working capital normalization following the initial HAMR ramp, Seagate could generate over $8 billion in free cash flow over the next three fiscal years, spurring a re-initiation of massive share repurchases and substantial returns of capital to shareholders.
Seagate is Insulated From Tariff Issues.
We see little risk of direct disruption to Seagate’s business from the ongoing trade disputes and tariffs. Like many semiconductor products, HDDs have been included on exempt product lists, and most every category of technology hardware has been granted a reprieve from steep reciprocal tariffs. At the same time, despite being a key AI enabler, HDD sales to China have not become politicized like advanced processor technology or semiconductor capital equipment. Seagate fabricates drive heads in the US and Ireland and performs most drive assembly operations in Thailand and Malaysia, with separate operations in China to service local demand. Western Digital has a similar footprint, meaning neither company has an inherent advantage. Much of hyperscaler server assembly is performed by Taiwanese ODMs, many of which have significant Mexican operations that are currently insulated from tariffs on such products. In addition, we believe there are tariff escalators in Seagate’s long-term customer supply agreements and that the company would be able to pass along any increased costs to its customers should the situation worsen. Seagate management said it expects minimal financial impact on June quarter financials from tariffs.
It’s Hard to Imagine a More Catalyst-Rich Set-Up.
We see a plethora of potentially impactful stock catalysts over the next six to nine months as the HAMR story unfolds, including:
- The company’s May 22nd investor day, its first since September 2019. We expect to hear about (1) additional hyperscaler qualifications of HAMR, and (2) an upward revision of the company’s long-term target financial model, which should include material upside revision to target gross margins.
- The company’s June 2025 earnings report, which should include (1) more HAMR qualifications, (2) an update on the HAMR volume ramp, and (3) evidence of a recovery in the video and imaging segment that suffered seasonal weakness in the March 2025 quarter.
- Initial FY2026 financial guidance, which will say a lot about the shape of the HAMR adoption curve and potential cyclicality.
- Execution on guided QoQ growth in revenue, gross margins, non-GAAP net income, and free cash flow throughout the year, which shouldn’t be a problem to deliver given Seagate’s capacity is essentially booked through the end of the year.
- Upward revision of the dividend at the October Board meeting.
- The release and commencement of qualification for the highly awaited 40TB HAMR product late in the year.
- Resumption of share buybacks in the first half of 2026, if not sooner.
Seagate Isn’t Getting Sufficient Credit For Its Demand Drivers or Cash Generation; We See a Structural Re-Rating on the Horizon.
Trading at a paltry 11x 2026 P/E, Seagate is notable for receiving little meaningful valuation benefit despite being nearly fully exposed to and a key enabler of the technology investment themes of the century: cloud computing and AI. We believe the stock has suffered from a lack of thoughtfulness regarding comparable company and valuation methodologies. HAMR-driven technology leadership and explosive gross margin expansion underpin a powerful bull thesis. Our proposed universe of defensible peer companies yields median 13x EV/EBITDA and 18x P/E multiples which, applied to our Seagate CY2026 forecasts, implies 96% and 80% stock price upside, respectively. Moreover, giving Seagate due credit for its highly cash-generative model, our discounted cash flow analysis suggests over 130% upside. We fully expect Seagate shares to experience a re-rating, as investors embrace the promise of HAMR and acknowledge the power of the company’s competitive position and underlying financial model.
Read the full report here by Kerrisdale Capital