Metraipolis › archive › dc_0d145a4fcd90

Enhancing Financial Question Answering: A Novel Benchmark Dataset of Banks' financial statements — arXiv:2609.03654

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.03654
License (as archived)
CC0-1.0
Archived
2026-10-16T15:05:28.761Z
SHA-256 of the archived text
fc1cfdd28f4f409ec93ebd4ebb9763218f12d021845c9d62e220b185d83fa4ee
Stable URL of this page
https://metraipolis.com/documento/dc_0d145a4fcd90

How to cite

Enhancing Financial Question Answering: A Novel Benchmark Dataset of Banks' financial statements — arXiv:2609.03654 — arXiv (en), https://arxiv.org/abs/2609.03654 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_0d145a4fcd90 (Metraipolis archive dc_0d145a4fcd90, sha256 fc1cfdd28f4f409ec93ebd4ebb9763218f12d021845c9d62e220b185d83fa4ee)
BibTeX
@misc{dc0d145a4fcd90,
  title = {Enhancing Financial Question Answering: A Novel Benchmark Dataset of Banks' financial statements — arXiv:2609.03654},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.03654},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_0d145a4fcd90; sha256 fc1cfdd28f4f409ec93ebd4ebb9763218f12d021845c9d62e220b185d83fa4ee},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.03654. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

The comparative analysis of banks' financial statements poses significant challenges for automated question answering systems due to their complexity, substantial length, technical language, and inhomogeneity of both textual and numerical content across different jurisdictions and institutions. We introduce FinRAG-QA, a novel benchmark dataset for financial question answering, which comprises 999 practitioner-curated questions on 10 standardised indicators, grounded in 209 annual and Pillar 3 reports from 24 major European and U.S. banks spanning 2019-2023. Unlike prior financial QA benchmarks, which centre on U.S. filings and single-institution analysis, FinRAG-QA targets cross-institutional retrieval over documents averaging 198k words, longer than any existing financial QA resource. On this benchmark we evaluate a multi-stage RAG pipeline and isolate the contribution of each component. Contextual chunk enrichment combined with a retrieval-optimised embedding model raises NDCG@10 from 0.322 to 0.710; conditional on the ground truth being retrieved, a reasoning-optimised generator raises answer accuracy from 44.6% to 79.0% (+34.4 percentage points), at roughly 20x the generation latency. We further show that cross-encoder reranking degrades retrieval when the first-stage ranking is already strong, and that a single top-ranked chunk outperforms larger contexts at generation time. Experiments were run in late 2024-early 2025 with the models available at that time.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-09 in cs.IR).
Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI); Computational Engineering, Finance, and Science (cs.CE); Information Retrieval (cs.IR); General Finance (q-fin.GN).
Pagina del lavoro: https://arxiv.org/abs/2609.03654 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-16T14:49:33.971Z dalla pagina https://arxiv.org/catchup/cs.IR/2026-09-09?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).