MadEvolve: Evolutionary Optimization of Trading Systems with Large Language Models

2026-05-21

AI Summary

This work addresses the algorithmic challenges in generating Alpha signals and optimizing trading strategies in quantitative finance by proposing MadEvolve, a general-purpose evolutionary optimization framework inspired by Alpha-Evolve and powered by large language models (LLMs). MadEvolve uniquely integrates LLM-based agents with evolutionary algorithms and backtesting simulation systems to enable end-to-end joint evolution of feature engineering and trade execution strategies. Evaluated in Bitcoin trading scenarios, the framework significantly outperforms existing agent-based search methods such as Claude Code, achieving superior strategy performance while effectively mitigating overfitting and p-hacking risks. This approach establishes a novel paradigm for automated algorithm discovery in quantitative finance.

Abstract

We explore the application of LLM-driven algorithm optimization to several common tasks in quantitative finance. MadEvolve, a general-purpose algorithm optimization framework inspired by DeepMind's Alpha-Evolve, was recently developed to optimize algorithms in computational cosmology. Here we demonstrate the utility of MadEvolve to optimize algorithmic trading strategies and alpha generation at the example of Bitcoin trading. On our simulation and backtesting setup, we achieve significant improvements on all tasks we considered, such as evolving feature sets for signal generation, optimizing separate components of the trading strategy, and jointly evolving the feature pipeline together with the execution strategy. Additionally, we compare our method to other agentic search approaches, specifically Claude Code, and carefully evaluate p-hacking probabilities on our simulation setup. Our findings strongly support the utility of AI-driven agentic and evolutionary algorithms for algorithmic trading and quantitative finance.

Problem

• algorithmic trading
• alpha generation
• quantitative finance
• evolutionary optimization
• trading strategy

Innovation

• LLM-driven optimization
• evolutionary algorithms
• algorithmic trading
• alpha generation
• agentic search