Llm agents evaluated for complex option trading decisions under constraints
LiveOption: Evaluating LLM Agents in Structured Option Trading with Nonlinear Payoffs
Artificial Intelligence
Summary
Trading options is more complicated than buying or selling stocks because the outcomes depend on complicated rules and strategies. The authors created a new testing setup called LiveOption to see how well large language models (LLMs) perform in making these complex trading decisions. Their setup mimics real trading conditions like limited money and timing challenges, and it checks not just profits but also how good and safe the trading choices are. Tests showed that current LLM-based agents mostly struggle to earn good returns in these complex tasks. LiveOption provides a new way to test financial AI that looks beyond just whether they make money.
What this means in practice
- •For quantitative trading teams: Evaluate and improve AI agents for structured option trading strategies under realistic constraints.
- •For financial software developers: Develop standardized platforms to test AI decision-making in complex multi-leg option trading scenarios.
Authors
Haochen Luo, Yifan Li, Binh Minh An, Xiaolong Luo, Zhengzhao Lai, Yuan Zhang, Chen Liu
Abstract
Large language models (LLMs) and multi-agent systems (MAS) have shown promise in financial decision-making, yet existing evaluations focus on equity trading and primarily assess directional prediction, overlooking the structural complexity of derivative markets. Option trading introduces fundamentally different challenges, including nonlinear payoffs and multi-leg strategy construction, requiring structured decisions rather than simple directional bets. We introduce LiveOption, an evaluation framework for LLM-based agents in option trading. LiveOption formulates the problem as structured sequential decision-making under realistic execution and capital constraints, and provides a reproducible environment with standardized interaction protocols. The framework includes three task suites covering portfolio overlays, event-driven earnings trading, and 0DTE intraday trading. We further propose a hierarchical metric suite that evaluates action validity, decision quality, risk characteristics, and outcome-level performance. Experiments show that current agents often fail to achieve competitive returns in most scenarios. LiveOption offers a principled testbed for evaluating structured decision-making beyond outcome-based metrics.