FCMR: Robust Evaluation of Financial Cross-Modal Multi-Hop Reasoning

TL;DR

This paper introduces FCMR, a new benchmark for evaluating multimodal large language models' ability to perform complex, multi-hop reasoning across financial data modalities, revealing current models' limitations.

Contribution

The paper presents FCMR, a challenging, multi-level benchmark for assessing cross-modal reasoning in financial contexts, addressing limitations of previous benchmarks.

Findings

State-of-the-art models achieve only 30.4% accuracy on hard tasks.

Models struggle with three-hop cross-modal reasoning.

Analysis uncovers a bottleneck in information retrieval.

Abstract

Real-world decision-making often requires integrating and reasoning over information from multiple modalities. While recent multimodal large language models (MLLMs) have shown promise in such tasks, their ability to perform multi-hop reasoning across diverse sources remains insufficiently evaluated. Existing benchmarks, such as MMQA, face challenges due to (1) data contamination and (2) a lack of complex queries that necessitate operations across more than two modalities, hindering accurate performance assessment. To address this, we present Financial Cross-Modal Multi-Hop Reasoning (FCMR), a benchmark created to analyze the reasoning capabilities of MLLMs by urging them to combine information from textual reports, tables, and charts within the financial domain. FCMR is categorized into three difficulty levels-Easy, Medium, and Hard-facilitating a step-by-step evaluation. In particular, problems at the Hard level require precise cross-modal three-hop reasoning and are designed to prevent the disregard of any modality. Experiments on this new benchmark reveal that even state-of-the-art MLLMs struggle, with the best-performing model (Claude 3.5 Sonnet) achieving only 30.4% accuracy on the most challenging tier. We also conduct analysis to provide insights into the inner workings of the models, including the discovery of a critical bottleneck in the information retrieval phase.