Generating actionable financial advice from business records demands that models integrate numerical reasoning, domain knowledge, and sound judgment, while avoiding recommendations that could harm the business. Direct supervision is difficult: historical decisions are not necessarily optimal, and high-quality free-form labels are expensive to obtain.
arXiv:2608. 11787v1 Announce Type: cross Abstract: Generating actionable financial advice from business records demands that models integrate numerical reasoning, domain knowledge, and sound judgment, while avoiding recommendations that could harm the business.
By Ofir Ben Shoham, Shrutendra Harsola, Vignesh Subrahmaniam, Shravan Mohan, Yakov Gazman, Oded Vainas
PaGNet is a hybrid panel‑aware model that combines a LightGBM branch with panel‑temporal summaries and a Panel‑MLP branch using attention‑pooled aggregation and shared‑trunk multi‑task learning. It produces predictions for corporate tax avoidance proxies while providing a per‑target diagnostic that indicates which branch contributed most to each forecast. On a panel of 1,754 Korean firms, PaGNet improves explained variance over six baselines by 0.08–0.11 on key accrual targets and demonstrates stable routing and diagnostic consistency across different feature regimes.
By Wonho Song, Hyungjoon Kim
arXiv:2607. 24889v1 Announce Type: cross Abstract: Financial models combine public disclosures with analyst assumptions to produce forecasts and valuations.
By Jiacheng Lu, Sinuo Wang, Wentao Zhao, Rui Sun, Cheng Hua, Tao Song, Hui Cai, Beidi Luan, Zhengze Wu, Lingjing Teng, Yijia He, Jing Li, Daxin Jiang, Zuo Bai, Haibing Guan
FinRiskAtlas is a Chinese-language benchmark designed to evaluate large language models (LLMs) for financial risk review by focusing on decision‑aligned tasks rather than generic financial knowledge. It contains 9,742 instances across 53 task families, including 42 domain‑knowledge families and 11 downstream review operations defined by explicit evaluation contracts. The extended FinRisk‑Ask framework replays 680 pre‑action states from 104 professional trajectories, withholding future evidence during inference to assess evidence‑state control and request targeting. Results across 33 model configurations show that operation‑level evaluation yields distinct rankings and that knowledge‑based shortlisting can incur significant regret, while frequent use of the Ask branch does not necessarily improve evidence acquisition, highlighting gaps in broad financial capability scores.
By Suyang Zhong, Jingzhe Zhu, Qi Xu, Liyao Sun, Yin Wang, Qingqing Sun, Shuai Chen, Tianyi Zhang
arXiv:2607. 18867v1 Announce Type: new Abstract: Large language models leak parametric knowledge of realized outcomes into historical financial decision tasks.
By Haozhe Jia