arXiv AI October 31, 2026 By Yanbin Yin, Chuyu Xu, Liqin Ye, Jing Peng, Chao Zhang

EPSBench: Benchmarking LLM Agents for Earnings Forecasting

EPSBench is a benchmark that evaluates large language model agents on quarterly earnings forecasting by having them connect financial evidence to income‑statement line items for 149 firm‑quarters from 86 US companies. Agents read filings, earnings‑call transcripts, and news, estimate company‑specific drivers, and a shared accounting calculator converts these into GAAP diluted earnings per share (EPS). Six models were assessed, with three showing no significant difference from analyst consensus in point forecast error, while the others performed worse; nominal 80% prediction intervals contained reported EPS only 37–70% of the time, largely due to low operating‑income estimates.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

arXiv Machine Learning

Aug 4

Latent-Regime Bias Auditing for Volatility Forecasting

arXiv:2608. 01599v1 Announce Type: new Abstract: Volatility forecasts are commonly evaluated with aggregate accuracy metrics such as RMSE and MAE, but these metrics can hide conditional failures that matter for risk management.

By Arthur Chagas, Pedro Bento, Yan Aquino, Arthur Buzelin, Wagner Meira Jr., Cristiano Arbex Valle

safety

arXiv AI

Oct 16

Momentum, Commentary Tone, and Earnings Surprise in LLM Financial Forecasts

The study investigates which inputs influence a language model’s financial forecasts by creating synthetic earnings announcements that independently reverse 20‑day momentum, commentary tone, and EPS surprise. Six instruction/chat checkpoint conversions from four model families (0.49‑7.61 B parameters) were tested with 2,016 prompts each, covering 384 primary scenarios. Results show that the contrast in primary EPS ranges from –0.19 to 86.80 percentage points, with signed tone and momentum contrasts often exceeding EPS effects; numerical diagnostic accuracy varied from 51.0 % to 95.8 %. The experiment measures observable input reliance and numerical recognition but does not establish internal reasoning mechanisms or return predictability.

By Alina Khaybullina

llms benchmarks

arXiv Machine Learning

Jun 15

Detecting Lookahead Bias in LLM Forecasts

arXiv:2512. 23847v2 Announce Type: replace-cross Abstract: We develop a statistical procedure to detect lookahead bias in economic forecasts generated by large language models (LLMs).

By Zhenyu Gao, Wenxi Jiang, Yutong Yan

llms safety

arXiv Machine Learning

Sep 10

From Narrative to Auditable Forecasts: A Structured Scaffold for Agentic Forecasting

arXiv:2609.05905v1 Announce Type: cross Abstract: LLM agents are increasingly used for live forecasting, where they retrieve up-to-date information and produce estimates for unresolved future events....

By Yuanpu Cao, Yongkang Du, Yurui Chang, Lu Lin, Jinghui Chen

llms agents benchmarks

arXiv Machine Learning

Jul 21

FinBench: Time-Gated Calibration and Uncertainty Benchmarking for Agentic Financial Forecasting

arXiv:2607. 16229v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly used as components of agentic systems that observe, plan, and act.

By Rishab Ghosh, Vinay Devarakonda

llms agents benchmarks safety

arXiv Machine Learning

Jun 16

Multi-Agent Framework for Audit Risk Assessment with Explicit Uncertainty and Evidence Conflict Modeling

arXiv:2606. 15640v1 Announce Type: new Abstract: Audit risk assessment increasingly benefits from combining heterogeneous evidence sources, yet existing approaches typically produce point predictions without quantifying how well different evidence streams agree.

By Yuhan Wang, Manqing Wang, Yixuan Lu, Zhaoyue Peng, Shengda Lin

llms agents

One email a morning, machine-written

Related stories

Latent-Regime Bias Auditing for Volatility Forecasting

Momentum, Commentary Tone, and Earnings Surprise in LLM Financial Forecasts

Detecting Lookahead Bias in LLM Forecasts

From Narrative to Auditable Forecasts: A Structured Scaffold for Agentic Forecasting

FinBench: Time-Gated Calibration and Uncertainty Benchmarking for Agentic Financial Forecasting

Multi-Agent Framework for Audit Risk Assessment with Explicit Uncertainty and Evidence Conflict Modeling