F$^2$Agent: Financial Fusion of Agentic Intelligence for Multimodal Trading

2026-08-06

AI Summary

Existing large language model–driven trading agents struggle to accurately model the intricate relationships among multimodal financial data and are highly sensitive to market noise. To address these limitations, this work proposes a hierarchical multi-agent architecture that dynamically captures cross-modal dependencies through a modality-aware adaptive fusion mechanism and enhances decision stability via noise-robust consistency regularization. Evaluated across six asset classes—including equities and cryptocurrencies—the proposed method significantly outperforms 16 baseline models, achieving an average annualized return improvement of over 20%. Notably, it attains annualized returns of 120.48% for GOOG and 148.41% for TSLA, demonstrating its effectiveness in multimodal fusion and noise-resilient trading decision-making.

Abstract

With increasingly diverse and heterogeneous information sources, effectively leveraging multimodal data is becoming pivotal for high-quality financial trading. Although recent advancements in Large Language Model (LLM)-based agents have enabled the ingestion of multimodal inputs, existing methods fail to capture nuanced cross-modal dependencies and remain vulnerable to market noise, due to limited multimodal modeling, ineffective fusion mechanisms, and inadequate robustness. To address these challenges, we propose F$^2$Agent, a novel multimodal agentic paradigm driven by the Financial Fusion of Agentic Intelligence. F$^2$Agent first deploys a hierarchy of specialized agents to comprehensively extract modality-specific signals. It further introduces a modality-aware adaptive fusion mechanism coupled with noise-robust consistency regularization to dynamically capture fine-grained inter-modality dependencies and generate noise-resilient trading signals. Extensive experiments on six stocks and cryptocurrency assets demonstrate that F$^2$Agent consistently outperforms 16 competitive baselines across multiple trading metrics, with over 20% relative improvement in annualized return on average. Notably, F$^2$Agent delivers returns of 120.48% on GOOG and 148.41% on TSLA, demonstrating its efficacy and robustness in varying market dynamics.

Problem

• multimodal trading
• cross-modal dependencies
• market noise
• robustness
• financial intelligence

Innovation

• Multimodal Fusion
• Agentic Intelligence
• Noise-Robust Trading
• Adaptive Fusion Mechanism
• Financial AI