EvolveTrade Paper: Evolving Agent System Prompts Beats Fixed-Prompt Trading Baselines

dair_ai · x · 2026-09-23

A recommended paper for agent harness tuning, EvolveTrade treats a trading agent's system prompt as its policy: after each trading interval, a separate Policy Agent reads decision traces and realized returns and rewrites the prompt while the backbone model stays frozen. Across several market regimes and two backbones, the evolved agent beats fixed-prompt baselines on Sharpe ratio and cumulative return in most settings, with rewritten prompts pushing the agent toward more code-based analysis and regime-fitting signals.

Original post →

More from coding & agent

coding & agent channel →