JPMorgan Chase & Co. has tested AI agents capable of autonomously allocating capital, outperforming the traditional 60/40 stock-and-bond portfolio in simulated trading environments. The experiments represent a significant step toward automated investment management.
The move comes as investors increasingly embrace artificial intelligence for stock picking, risk management, and now, portfolio construction. By handing allocation decisions to models, the bank signals growing confidence in machine-driven finance.
Bloomberg reported that the AI agents beat the 60/40 benchmark in backtests, though specific performance figures were not disclosed. The tests evaluated the agents' ability to dynamically shift assets without human intervention.
If deployed, autonomous allocation could disrupt traditional wealth management, potentially reducing fees and altering the role of human advisors. Regulators may scrutinize the reliability and transparency of such systems.
The innovation highlights a broader industry pivot toward AI-driven finance, but critics warn that black-box models could amplify risks during market stress.