INVESTORBENCH: A Benchmark for Financial Decision-Making Tasks with LLM-based Agent
Haohang Li, Yupeng Cao, Yangyang Yu, Shashidhar Reddy Javaji, Zhiyang Deng, Yueru He, Yuechen Jiang, Zining Zhu, Koduvayur Subbalakshmi, Guojun Xiong, Jimin Huang, Lingfei Qian, Xueqing Peng, Qianqian Xie, Jordan W. Suchow
TL;DR
InvestorBench addresses the lack of standardized benchmarks and adaptable frameworks for evaluating LLM-based financial decision-making across diverse tasks. It introduces an open-source benchmark with multi-source market data, three asset-class tasks, a memory-augmented LLM agent framework, and a unified evaluation protocol tested across 13 backbone models. Stock trading results show proprietary backbones generally outperform open-source and domain-tuned models, and memory-reflection mechanisms improve robustness in open-ended decision contexts. The platform provides a scalable, multi-modal testbed for rigorous comparison of financial reasoning by LLM agents, enabling faster development and deployment of robust decision-making tools.
Abstract
Recent advancements have underscored the potential of large language model (LLM)-based agents in financial decision-making. Despite this progress, the field currently encounters two main challenges: (1) the lack of a comprehensive LLM agent framework adaptable to a variety of financial tasks, and (2) the absence of standardized benchmarks and consistent datasets for assessing agent performance. To tackle these issues, we introduce \textsc{InvestorBench}, the first benchmark specifically designed for evaluating LLM-based agents in diverse financial decision-making contexts. InvestorBench enhances the versatility of LLM-enabled agents by providing a comprehensive suite of tasks applicable to different financial products, including single equities like stocks, cryptocurrencies and exchange-traded funds (ETFs). Additionally, we assess the reasoning and decision-making capabilities of our agent framework using thirteen different LLMs as backbone models, across various market environments and tasks. Furthermore, we have curated a diverse collection of open-source, multi-modal datasets and developed a comprehensive suite of environments for financial decision-making. This establishes a highly accessible platform for evaluating financial agents' performance across various scenarios.