Quant hedge funds expand AI research investment amid intensifying talent war

Quantitative hedge funds and investment firms are beefing up their support for academic research and AI lab initiatives as the battle for machine-learning talent heats up against tech companies, according to a Bloomberg report. Firms like G-Research, Qube Research & Technologies, Capital Fund Management (CFM), and Jump Trading are financing PhD programs, backing university researchers, and setting up specialized labs to lure scientists who can build more advanced trading models.

The war for talent has pushed first-year salaries for machine-learning PhDs to between $450,000 and $650,000, with select cases hitting $800,000, per recruitment industry data. However, quant firms are finding themselves up against AI developers who can offer researchers higher public recognition and potential equity gains. Qube Research & Technologies has set up QRT Labs, sponsoring over 70 PhD and postdoc researchers, while G-Research has backed 36 researchers via the European Laboratory for Learning and Intelligent Systems and supports 43 PhD students across eight UK universities.

Paris-based CFM, managing about $35bn, has established a machine-learning lab focused on longer-term academic research, separate from the need to generate immediate trading signals. Other companies, including Hudson River Trading, XTX Markets, JPMorgan, and Optiver, have also launched dedicated AI research programs, and Jump Trading has expanded its academic collaborations to bolster hiring.

The fierce competition is spilling over, with hedge fund employees jumping to AI-focused tech firms like OpenAI, and researchers leaving established institutions such as Two Sigma for startups like Anthropic. Meanwhile, former DeepMind scientists have transferred to Jump Trading and Citadel Securities. This talent hunt highlights the mounting significance of machine learning in systematic investing, pushing firms not just on compensation but also on giving researchers access to cutting-edge computing resources, research freedom, and opportunities to create sophisticated predictive models.