When Valid Signals Fail: Regime Boundaries Between LLM Features and RL Trading Policies

TL;DR

This paper investigates the use of large language models as feature extractors for reinforcement learning trading agents, revealing that valid features do not always enhance policy robustness under distribution shifts.

Contribution

It introduces an automated prompt-optimization method for extracting predictive features from LLMs and analyzes their impact on RL trading performance during macroeconomic shocks.

Findings

Optimized prompts produce features with high predictive correlation (IC > 0.15).

Features improve performance in stable conditions but add noise during shocks.

Macroeconomic variables remain the most robust drivers of policy improvement.

Abstract

Can large language models (LLMs) generate continuous numerical features that improve reinforcement learning (RL) trading agents? We build a modular pipeline where a frozen LLM serves as a stateless feature extractor, transforming unstructured daily news and filings into a fixed-dimensional vector consumed by a downstream PPO agent. We introduce an automated prompt-optimization loop that treats the extraction prompt as a discrete hyperparameter and tunes it directly against the Information Coefficient - the Spearman rank correlation between predicted and realized returns - rather than NLP losses. The optimized prompt discovers genuinely predictive features (IC above 0.15 on held-out data). However, these valid intermediate representations do not automatically translate into downstream task performance: during a distribution shift caused by a macroeconomic shock, LLM-derived features add noise, and the augmented agent under-performs a price-only baseline. In a calmer test regime the agent recovers, yet macroeconomic state variables remain the most robust driver of policy improvement. Our findings highlight a gap between feature-level validity and policy-level robustness that parallels known challenges in transfer learning under distribution shift.