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

2026-02-26

AI Summary

This work addresses the limitations of existing multi-agent large language model–based financial trading systems, which often rely on coarse-grained instructions that fail to capture the complexity of real-world investment workflows, resulting in constrained reasoning performance and opaque decision-making. To overcome this, we propose a fine-grained task decomposition mechanism that explicitly breaks down investment analysis into subtasks aligned with practical investment workflows. We further refine agent architecture and task configuration to better align analytical outputs with downstream portfolio preferences. Evaluated in a leakage-resistant backtesting environment incorporating Japanese stock prices, financial statements, news, and macroeconomic data—and coupled with standard portfolio optimization techniques—our framework significantly outperforms coarse-grained baselines, achieving higher risk-adjusted returns while maintaining controlled portfolio volatility and low correlation with benchmark indices.

Technology Category

• Multiagent Systems: Mechanism Design
• Cognitive Modeling & Cognitive Systems: Agent Architectures
• Machine Learning: Mixture of Experts (MoE)

Application Category

• Economics, Online Markets and Human Computation: Architectures and workflows that use LLMs for crowd work
• Search and Retrieval-Augmented AI: Agentic search
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

The advancement of large language models (LLMs) has accelerated the development of autonomous financial trading systems. While mainstream approaches deploy multi-agent systems mimicking analyst and manager roles, they often rely on abstract instructions that overlook the intricacies of real-world workflows, which can lead to degraded inference performance and less transparent decision-making. Therefore, we propose a multi-agent LLM trading framework that explicitly decomposes investment analysis into fine-grained tasks, rather than providing coarse-grained instructions. We evaluate the proposed framework using Japanese stock data, including prices, financial statements, news, and macro information, under a leakage-controlled backtesting setting. Experimental results show that fine-grained task decomposition significantly improves risk-adjusted returns compared to conventional coarse-grained designs. Crucially, further analysis of intermediate agent outputs suggests that alignment between analytical outputs and downstream decision preferences is a critical driver of system performance. Moreover, we conduct standard portfolio optimization, exploiting low correlation with the stock index and the variance of each system's output. This approach achieves superior performance. These findings contribute to the design of agent structure and task configuration when applying LLM agents to trading systems in practical settings.

Problem

• multi-agent LLM
• financial trading systems
• task decomposition
• investment analysis
• decision transparency

Innovation

• fine-grained task decomposition
• multi-agent LLM system
• risk-adjusted returns
• decision alignment
• leakage-controlled backtesting