AlphaPilot uses AI agents to improve stock ranking and returns
AlphaPilot: Multi-Turn Tool-Using LLM Agents for Adaptive Alpha Mining with Reinforcement Learning
Computation and LanguageMachine Learning
Summary
Investors want formulas that predict which stocks will go up, but making these formulas is hard because it requires understanding market news, past data, and testing ideas. The authors created AlphaPilot, an AI system that talks to itself in multiple steps to come up with better stock-picking formulas, checking each idea against real market data. This lets the system adapt day by day as the market changes, leading to better investment returns and risk measures than other methods.
What this means in practice
- •For quantitative investment teams: Create adaptive stock ranking formulas that incorporate market news and data for improved portfolio performance.
- •For financial software developers: Build AI-powered tools that generate and validate investment factors using multi-step reasoning and data feedback.
Authors
Peiqi Jiang, Fan Zhang, Xin Xu, Tianhao Chen, Can Yang
Abstract
Formulaic alpha mining seeks interpretable expressions that rank assets by future returns. The task is challenging because it demands simultaneous processing of unstructured market descriptions, historical data, and quantitative evaluation of candidate factors, yet most existing methods exploit only a subset of these signals. We introduce AlphaPilot, a reinforcement-learning framework that trains multi-turn tool-using LLM agents for adaptive alpha mining. The agent conditions on market descriptions, proposes factor expressions, queries real-data evaluator feedback, and updates a retained factor pool across turns. This design jointly leverages market text, LLM financial knowledge, real data, and evaluation feedback. At inference time, AlphaPilot performs day-by-day search and pool updates to adapt to shifting market regimes. Across three independent runs in out-of-sample rolling backtests on CSI300 and CSI500, AlphaPilot achieves mean total returns of 42.93% and 54.19%, with mean Sharpe ratios of 2.46 and 2.86, respectively. It attains the highest mean total return, Sharpe, and RankIC among the evaluated methods on both universes.