Toward Expert Investment Teams:A Multi-Agent LLM System with Fine-Grained Trading Tasks

Researchers developed a multi-agent LLM trading framework that decomposes investment analysis into fine-grained tasks rather than coarse-grained instructions. Testing on Japanese stock data showed the approach significantly improved risk-adjusted returns and achieved superior performance through portfolio optimization.

• →Fine-grained task decomposition in LLM trading systems outperforms conventional coarse-grained instruction approaches.
• →Alignment between analytical outputs and downstream decision preferences is critical for trading system performance.
• →The framework was successfully tested on Japanese stock data including prices, financial statements, news, and macro information.
• →Portfolio optimization exploiting low correlation with stock indices achieved superior performance results.
• →Multi-agent LLM systems show promise for autonomous financial trading when properly structured with detailed task breakdown.