2 trading firms invest billions in AI data center race

Algorithmic trading firm XTX Markets is constructing a $1 billion-plus data center complex in Finland, while rival Jane Street just committed $6 billion to rent computing power from AI cloud specialist CoreWeave, signaling a capital-intensive pivot from speed to raw intelligence in financial markets. The moves represent a new arms race in quantitative trading, where an edge is no longer gained by shaving milliseconds off trade times but by building massive artificial intelligence models to predict market moves.

“All these HFT firms…are trying to leapfrog each other by becoming faster and faster,” XTX founder Alex Gerko said in a 2019 discussion. “We decided we’re just not a part of this game.” Gerko, a mathematician who has been a vocal critic of the high-frequency trading speed race, is doubling down on the belief that superior predictive models, not just faster execution, will generate the highest returns.

The scale of the new AI infrastructure push is immense. XTX’s bet involves building its own complex of five data centers, starting operations this year and housing its 25,000 Nvidia AI chips. Jane Street’s deal, meanwhile, includes a $6 billion contract for CoreWeave’s cloud services and a direct $1 billion equity investment in the fast-growing AI infrastructure provider, according to a recent Cantor Fitzgerald note.

This strategic shift effectively raises the cost of competition, creating a significant moat around firms with the capital to build or rent massive-scale AI compute. The billions being poured into data centers suggest the future of trading belongs to firms with the most powerful AI, a trend that is highly bullish for the specialized companies building the underlying infrastructure, including CoreWeave and chipmaker Nvidia.

The New Competitive Moat

The quantitative trading industry is undergoing a fundamental change, moving from a physical arms race of microwave towers and fiber optic cables to a computational one. XTX Markets, which generated $2.3 billion in profit on $5.3 billion in revenue from its U.K. business alone last year, has proven the viability of this AI-centric model. With only 250 employees, the firm’s efficiency highlights the power of its deep learning models, which trade an average of $250 billion daily across asset classes. By building its own data centers in a northern climate to ensure cheap and efficient cooling, XTX is locking in a long-term structural advantage in training its AI.

CoreWeave at the Center

For firms that choose not to build, CoreWeave is emerging as a kingmaker. The so-called “neocloud” company, which provides dedicated access to Nvidia’s latest GPUs, now has a revenue backlog approaching $100 billion, according to market analysis. The $6 billion Jane Street contract adds to a roster of clients queuing for its AI computing capacity. This booming demand has sent CoreWeave’s stock soaring 49% in April, though it still trades at 10 times sales—a multiple that may seem reasonable given its explosive growth trajectory. Cantor Fitzgerald recently raised its price target on the stock to $156, citing the firm's key strategic deals.

The massive investments from trading firms and other sectors are fueling a broader rally in technology stocks. The tech-heavy Nasdaq-100 has erased its first-quarter losses, rallying 11% in April as a ceasefire in the Middle East boosted investor confidence. Analysts expect Nasdaq-100 companies to report net income growth of 19% for the first quarter, far outpacing the 11% expected from the S&P 500, driven by AI-related spending. This spending is not limited to a few firms; the five largest U.S. hyperscalers are projected to spend $720 billion on capital expenditures in 2026.

This AI arms race in finance provides a powerful, long-term tailwind for the companies supplying the picks and shovels. While specialized players like CoreWeave offer a direct investment in AI infrastructure, tech giants like Microsoft also stand to benefit. Microsoft is seeing remarkable demand for its Copilot AI assistant and trades at an attractive 22 times forward earnings, a slight discount to the Nasdaq-100. For investors, the multi-billion-dollar data center commitments from the world's most sophisticated trading firms serve as a strong confirmation of the AI infrastructure investment thesis.

This article is for informational purposes only and does not constitute investment advice.