FinPos: A Position-Aware Trading Agent System for Real Financial Markets

TL;DR

This paper introduces FinPos, a novel position-aware trading agent system that explicitly manages continuous positions in financial markets, leveraging large language models and a dual-agent decision structure for improved decision-making.

Contribution

The paper presents FinPos, the first trading agent system explicitly designed for continuous position management using LLMs, with mechanisms for interpretation, decision separation, and multi-timescale rewards.

Findings

FinPos outperforms existing trading agents in position-aware tasks.

LLM-centered systems show significant potential for long-term market decisions.

Position awareness improves trading performance and realism.

Abstract

The exceptional potential of large language models (LLMs) in handling text information has garnered significant attention in the field of financial trading. However, most existing trading agents operate under intraday, independent unit-based trading tasks, where decisions are made as isolated directional actions, and thus lack awareness of continuous position management. Therefore, we propose a position-aware trading task designed to simulate a more realistic market. To address this task, we propose FinPos, a position-aware trading agent system designed to explicitly model and manage continuous positions. FinPos enhances position awareness through three key mechanisms: (1) professional-level interpretation of heterogeneous market information; (2) a dual-agent decision structure that separates directional reasoning from risk-aware position adjustment; and (3) multi-timescale reward signals, allowing the agent to internalize position awareness through experiential feedback rather than static instructions alone. Extensive experiments demonstrate that FinPos surpasses state-of-the-art trading agents in the position-aware trading task, which closely mirrors real market conditions. More importantly, our findings reveal that LLM-centered agent systems exhibit a vast, largely unexplored potential in long-term market decision-making.