QuantaAlpha: An Evolutionary Framework for LLM-Driven Alpha Mining

2026-02-06

AI Summary

This work addresses the challenges posed by market noise and non-stationarity in financial time series, which render traditional alpha factors susceptible to backtest overfitting and ineffective under regime shifts. To overcome the limitations of existing agent-based frameworks—particularly their lack of controllable multi-round search and experience reuse mechanisms—we propose QuantaAlpha, a novel approach that formulates end-to-end alpha factor discovery as a trajectory optimization problem. By employing trajectory-level mutation and crossover operations, QuantaAlpha identifies suboptimal steps and recombines high-return segments to enable effective pattern reuse. Semantic consistency constraints and complexity control are further integrated to ensure factor interpretability and generalization. Combining large language models with evolutionary algorithms, our method achieves an information coefficient (IC) of 0.1501, annualized return of 27.75%, and maximum drawdown of 7.98% on the CSI 300 index, and demonstrates strong transferability with four-year cumulative excess returns of 160% and 137% on CSI 500 and S&P 500, respectively.

Technology Category

• Search and Optimization: Algorithm Configuration
• Multiagent Systems: Mechanism Design
• Machine Learning: Quantum Machine Learning

Application Category

• Search and Retrieval-Augmented AI: Agentic search
• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Financial markets are noisy and non-stationary, making alpha mining highly sensitive to noise in backtesting results and sudden market regime shifts. While recent agentic frameworks improve alpha mining automation, they often lack controllable multi-round search and reliable reuse of validated experience. To address these challenges, we propose QuantaAlpha, an evolutionary alpha mining framework that treats each end-to-end mining run as a trajectory and improves factors through trajectory-level mutation and crossover operations. QuantaAlpha localizes suboptimal steps in each trajectory for targeted revision and recombines complementary high-reward segments to reuse effective patterns, enabling structured exploration and refinement across mining iterations. During factor generation, QuantaAlpha enforces semantic consistency across the hypothesis, factor expression, and executable code, while constraining the complexity and redundancy of the generated factor to mitigate crowding. Extensive experiments on the China Securities Index 300 (CSI 300) demonstrate consistent gains over strong baseline models and prior agentic systems. When utilizing GPT-5.2, QuantaAlpha achieves an Information Coefficient (IC) of 0.1501, with an Annualized Rate of Return (ARR) of 27.75% and a Maximum Drawdown (MDD) of 7.98%. Moreover, factors mined on CSI 300 transfer effectively to the China Securities Index 500 (CSI 500) and the Standard&Poor's 500 Index (S&P 500), delivering 160% and 137% cumulative excess return over four years, respectively, which indicates strong robustness of QuantaAlpha under market distribution shifts.

Problem

• alpha mining
• financial markets
• non-stationarity
• noise sensitivity
• market regime shifts

Innovation

• evolutionary framework
• trajectory-level mutation
• semantic consistency
• alpha mining
• cross-market transfer