2025 will go down in market lore as a year when artificial intelligence didn’t just reshape technology — it reshaped portfolios. While everyone else was piling into the usual AI suspects like chipmakers and cloud giants, one unusual investor quietly demolished expectations by backing the infrastructure that actually makes AI possible.
That investor wasn’t a traditional Wall Street guru with decades of tech investing under his belt. No, he was Xavier Majic, a former professional ice-hockey player who transitioned into finance and now runs Maple Rock Capital Partners, a $3 billion hedge fund that gained more than 60% in 2025 through November thanks to a savvy bet on data storage and memory stocks.
From Ice Rinks to Financial Models
Majic’s story sounds like a script Hollywood passed over. Born and raised in British Columbia, he spent a decade as a center on professional hockey teams around the world, including the Jacksonville Lizard Kings and the Manitoba Moose. He even served as captain of Canada’s men’s national team during the late 1990s, long before AI was on most investors’ radar.
After hanging up his skates, Majic pursued an MBA at Harvard Business School, then cut his teeth at hedge funds in San Francisco before launching Maple Rock in 2015. Over the next decade, he developed a reputation for finding value in cyclical and overlooked sectors — not for chasing tech fads.
That background helped him approach the AI trade with a contrarian’s eye.
Betting the Infrastructure, Not the Hype
In 2025, most headlines around AI stocks were dominated by companies like Nvidia — whose GPUs power many AI models — but Majic saw something different. His thesis wasn’t “AI will take over the world” — it was “AI will consume more data than we currently have room to store.”
To power AI training and inference, data centers need massive amounts of storage hardware — everything from traditional hard disks to flash memory. While companies like Nvidia and Alphabet were busy optimizing performance, no one was giving the same attention to the storage layer underpinning global data needs.
Majic’s fund positioned itself heavily in stocks like Western Digital, Seagate Technology Holdings, and Kioxia Holdings — companies that make the drives and memory modules that store the tens of petabytes AI demands. As companies worldwide raced to deploy large language models and generative AI applications, demand for storage capacity exploded, validating Majic’s unglamorous but prescient bet.
Shares of Western Digital and Seagate surged — with Western Digital nearly quadrupling and Seagate more than tripling through 2025 — far outpacing traditional AI leaders and broader indexes. In comparison, Nvidia rose about 32% over the same period, while the S&P 500 returned around 18%.
The Logic Behind the Win
What made Majic’s strategy work wasn’t a crystal ball. It was careful analysis of how technology ecosystems evolve and a willingness to think beyond the obvious. Most investors were focused on the “sexy” end of AI: chips, software, platforms. Majic looked where the rubber meets the road: data persistence. Every AI model needs storage — and not just any storage, but enterprise-grade capacity that can keep up with rapid reads, writes, and backups across distributed cloud systems.
Financial analysts at Morgan Stanley captured this dynamic when they wrote that data is the oil that keeps AI running, and that hard-disk drive makers benefit from the vast data storage needs inherent in AI’s future.
Majic’s team didn’t just buy and hold; they timed their positions, building stakes gradually as their thesis played out from 2016 into the present cycle and then scaling up amid fresh demand signals by 2022 — well before the 2025 AI surge.
Beyond AI: Diversification and Market Savvy
While the AI trade was the standout, it wasn’t Majic’s only successful play. Maple Rock also scored with positions in other cyclical sectors, such as mining equities, underscoring the fund’s broader philosophy: find areas of mispricing before the market does.
But 2025’s AI trade is what turned heads in both financial and tech circles. Hedge funds with explicit AI mandates launched with fresh capital, and even tech giants talked up structural AI demand — yet few achieved the kind of returns Majic’s portfolio delivered.
The success also highlights how investors are now thinking about AI beyond just algorithms: hardware supply chains, data infrastructure, advanced networking, and storage are all integral to the technology’s real-world deployment and expansion.
Lessons for Investors and Tech Watchers
Majic’s journey from ice hockey to hedge fund success reveals several broader trends:
1. AI isn’t a single investment
Investing in AI infrastructure — such as data storage, memory, and networking — can yield significant gains alongside traditional hardware bets like GPUs and cloud services.
2. Value meets structural growth
Majic’s strategy combined value investing principles (looking for mispriced assets) with secular growth trends (AI adoption), creating a hybrid playbook that many retail investors overlook.
3. Timing and patience matter
The thesis was developed over years — not months — showing how long-term thinking can yield outsized results when paired with deep sector knowledge.
All of which serves as a reminder: sometimes the biggest market opportunities come not from the loudest sectors, but the ones quietly powering them.
The Human Angle: Not Your Typical Fund Manager
Unlike many high-profile fund managers who build personal brands via TV appearances or social feeds, Majic keeps a lower profile — coaching youth hockey and occasionally reconnecting with old teammates on the road, according to friends and associates.
It’s a grounded contrast to the glitzy world of tech investing and AI hype cycles, and it underscores a deeper truth: sound investing isn’t about riding the trendiest names — it’s about understanding the engines behind them.
Looking Ahead: What Comes After the AI Trade?
Even as the AI boom continues, the market is already shifting toward the next set of bottlenecks: energy efficiency for data centers, semiconductor fabrication capacity, AI-specific memory technologies (like HBM and CXL-enabled modules), and innovations that reduce storage latency and increase throughput.
Some of the top-performing AI-linked stocks in recent months — such as Seagate and Micron — highlight how data handling and storage may remain central to investors’ focus. Analysts continue to track these segments for future growth opportunities as demand evolves.
But for now, Majic’s 2025 AI trade stands as a case study in seeing beyond headlines, finding value where others aren’t looking, and backing that insight with conviction.
