Abstract
Financial trading is a crucial component of the markets, informed by amultimodal information landscape encompassing news, prices, and Kline charts,and encompasses diverse tasks such as quantitative trading and high-frequencytrading with various assets. While advanced AI techniques like deep learningand reinforcement learning are extensively utilized in finance, theirapplication in financial trading tasks often faces challenges due to inadequatehandling of multimodal data and limited generalizability across various tasks.To address these challenges, we present FinAgent, a multimodal foundationalagent with tool augmentation for financial trading. FinAgent's marketintelligence module processes a diverse range of data-numerical, textual, andvisual-to accurately analyze the financial market. Its unique dual-levelreflection module not only enables rapid adaptation to market dynamics but alsoincorporates a diversified memory retrieval system, enhancing the agent'sability to learn from historical data and improve decision-making processes.The agent's emphasis on reasoning for actions fosters trust in its financialdecisions. Moreover, FinAgent integrates established trading strategies andexpert insights, ensuring that its trading approaches are both data-driven androoted in sound financial principles. With comprehensive experiments on 6financial datasets, including stocks and Crypto, FinAgent significantlyoutperforms 9 state-of-the-art baselines in terms of 6 financial metrics withover 36% average improvement on profit. Specifically, a 92.27% return (a 84.39%relative improvement) is achieved on one dataset. Notably, FinAgent is thefirst advanced multimodal foundation agent designed for financial tradingtasks.