| title | E2A-Bench: Benchmarking Evidence-to-Action Reliability in Financial Chart Reasoning |
|---|---|
| authors | Xiaoya Wang, Yutong Xu, Junjie Wang |
| date | 2026-09-13 |
| venue | arXiv (preprint, not peer reviewed) |
| id | arXiv:2609.14302 |
| upvotes | 3 (Hugging Face Daily Papers, 2026-11-01) |
| links | abs pdf hf |
| abstract | Can financial vision-language models (VLMs) turn chart evidence into reliable action recommendations? Existing hallucination evaluations are mostly claim-centric; they assess whether generated statements are supported, but not whether evidence remains traceable through rationale, confidence, and final action. We introduce E2A-Bench, a 969-query benchmark for financial chart reasoning, constructed from 323 HS300 constituents under three input modalities with deterministic OHLCV-derived evidence anchors. E2A-Bench evaluates grounding, reasoning-action consistency, evidence-confidence calibration, and directional coverage through UCR, RCI, ECI, and NDR, where NDR measures coverage-aware evidence-to-action reliability rather than realized trading performance. Evaluating 20 VLMs reveals three failures hidden by scalar hallucination scores: the lowest-UCR model ranks near the bottom by NDR due to only 6.4% directional coverage; oracle-aided verification reduces unsupported claims but can collapse coverage; and financial fine-tuning amplifies the BUY:SELL ratio by factors of 4.21 to 4.68 across strict base-fine-tuned pairs. These results show that financial VLM evaluation should trace the full evidence-to-action chain rather than rely on a single hallucination score. Code and data: https://github.com/wanng-ide/E2A-Bench |
cite
@misc{wang2026e,
title={E2A-Bench: Benchmarking Evidence-to-Action Reliability in Financial Chart Reasoning},
author={Xiaoya Wang and Yutong Xu and Junjie Wang},
year={2026},
eprint={2609.14302},
archivePrefix={arXiv},
url={https://arxiv.org/abs/2609.14302}
}