EvolveTrade: Experience-Driven Policy Refinement for Self-Evolving LLM Trading Agents
Sehee Kim, Yumin Choi, Minki Kang and Sung Ju Hwang (KAIST, DeepAuto.ai) present EvolveTrade, which treats a trading agent's system prompt as a text policy that a Policy Agent rewrites from decision traces and realized portfolio returns while the backbone LLM stays fixed.
Ask this paper
Policy updates. After each interval the Policy Agent revises how the agent gathers evidence, calls tools, verifies signals and manages risk, and the new prompt governs the next batch of trades.
Returns. Across several market regimes and two backbones, EvolveTrade improves Sharpe ratio and cumulative return over fixed-policy LLM baselines in most settings.
Behavior change. Evolved policies use more code-based analysis and trigger computations relevant to the current regime, and case-level attributions link policy changes to return differences.
Abstract
Large language model (LLM) trading agents can combine market data, news, and executable analysis, but their behavior is often controlled by static hand-written tool-use policies that are fixed before deployment. This limits their ability to adapt how they gather evidence, invoke tools, verify signals, and manage risk under changing market regimes. We introduce EvolveTrade, a self-evolving framework that treats the system prompt of a tool-using trading agent as a text-parameterized policy. After each update interval, a Policy Agent revises this policy using accumulated decision traces and realized portfolio feedback, while keeping the backbone LLM fixed. The updated policy is then used for the next batch of trading decisions, enabling the agent to refine its information-acquisition and portfolio-construction procedure over time. Experiments across multiple market regimes and two LLM backbones show that EvolveTrade often improves Sharpe Ratio and Cumulative Return over fixed-policy LLM baselines, achieving the improved SR and CR in most evaluated settings. Behavioral analyses further show that self-evolved policies increase code-mediated analysis and activate regime-relevant computations; case-level policy-to-return attributions trace how policy-induced allocation changes contribute to realized return differences. These results suggest that adapting the reusable procedure governing tool use is a key direction for building more robust LLM trading agents.