Learning to Trade Like an Expert: Cognitive Fine-Tuning for Stable Financial Reasoning in Language Models
For researchers and practitioners building LLM-based trading agents, this work provides a structured method to train and evaluate financial reasoning, though it is an incremental improvement over existing fine-tuning approaches.
The paper proposes a training and evaluation framework for LLMs as autonomous trading agents, using a curated MCQ dataset and a two-stage protocol. Their fine-tuned open models achieve competitive, risk-aware trading performance, outperforming open-source baselines and approaching frontier models at smaller scale.
Recent deployments of large language models (LLMs) as autonomous trading agents raise questions about whether financial decision-making competence generalizes beyond specific market patterns and how it should be trained and evaluated in noisy markets lacking ground truth. We propose a structured framework for training and evaluating such models. Central to our approach is a curated, multiple-choice question (MCQ) dataset derived from classic textbooks and historical markets, verified by an AI committee, enriched with structured reasoning traces, and augmented to reduce shortcut learning. To evaluate whether performance on isolated MCQs generalizes to real-world trading, we introduce a two-stage protocol combining test-set evaluation with an MCQ-based chronological trading simulation. Extensive evaluations across market regimes provide statistically robust evidence that open models trained with our framework exhibit competitive, risk-aware behavior over time, outperform open-source baselines, and approach frontier-model performance at smaller scale. We release the dataset and evaluation framework to support further research.