@inproceedings{zhang-etal-2025-xfinbench,
    title = "{XF}in{B}ench: Benchmarking {LLM}s in Complex Financial Problem Solving and Reasoning",
    author = "Zhang, Zhihan  and
      Cao, Yixin  and
      Liao, Lizi",
    editor = "Che, Wanxiang  and
      Nabende, Joyce  and
      Shutova, Ekaterina  and
      Pilehvar, Mohammad Taher",
    booktitle = "Findings of the Association for Computational Linguistics: ACL 2025",
    month = jul,
    year = "2025",
    address = "Vienna, Austria",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/landing_page/2025.findings-acl.457/",
    pages = "8715--8758",
    ISBN = "979-8-89176-256-5",
    abstract = "Solving financial problems demands complex reasoning, multimodal data processing, and a broad technical understanding, presenting unique challenges for current large language models (LLMs). We introduce **XFinBench**, a novel benchmark with 4,235 examples designed to evaluate LLM{'}s ability in solving comple**X**, knowledge-intensive **Fin**ancial problems across diverse graduate-level finance topics with multi-modal context. We identify five core capabilities of LLMs using XFinBench, i.e., {\_}terminology understanding{\_}, {\_}temporal reasoning{\_}, {\_}future forecasting{\_}, {\_}scenario planning{\_}, and {\_}numerical modelling{\_}. Upon XFinBench, we conduct extensive experiments on 18 leading models. The result shows that o1 is the best-performing text-only model with an overall accuracy of 67.3{\%}, but still lags significantly behind human experts with 12.5{\%}, especially in temporal reasoning and scenario planning capabilities. We further construct a knowledge bank with 3,032 finance terms for knowledge augmentation analysis, and find that relevant knowledge to the question only brings consistent accuracy improvements to small open-source model. Additionally, our error analysis reveals that rounding errors during calculation and blindness to position and intersection of curves in the image are two primary issues leading to model{'}s poor performance in calculating and visual-context questions, respectively."
}