aiqbench analysis
The benchmark tests whether LLM agents can link filings, earnings calls, and news to the income-statement drivers behind EPS, with cited evidence. Results are mixed: three of six models match analyst consensus, and their 80% prediction intervals cover reported EPS only 37-70% of the time, mostly because forecasts run low. For AI readers, the useful finding is that the error gap traces largely to operating-income estimates, which points to where agent evaluation and improvement should focus.
EPSBench: Benchmarking LLM Agents for Earnings Forecasting
From arXiv
Full textarXiv:2610.38293v1 Announce Type: cross Abstract: Quarterly earnings forecasting requires an agent to connect financial evidence to the income-statement line items that produce a company's reported earnings. We introduce EPSBench, a benchmark for studying this ability across 149 firm-quarters from 86 US companies. Agents read a fixed collection of filings, earnings-call transcripts, and news, then estimate company-specific income-statement drivers with cited evidence. A shared accounting calculator converts these estimates into GAAP diluted earnings per share (EPS), providing a common basis for comparing forecast assumptions.
We evaluate six models under a common harness, assessing point forecast error relative to date-matched analyst consensus and checking whether their prediction intervals achieve the stated coverage. Three models show no detectable paired difference in point forecast error from consensus, and three have higher error. Across all six models, nominal 80% intervals contain reported EPS in only 37--70% of runs that return an interval, and most misses fall above the upper bound because the forecasts run low. The driver records trace most of the error gap between models to operating-income estimates.
Abstract:Quarterly earnings forecasting requires an agent to connect financial evidence to the income-statement line items that produce a company's reported earnings. We introduce EPSBench, a benchmark for studying this ability across 149 firm-quarters from 86 US companies. Agents read a fixed collection of filings, earnings-call transcripts, and news, then estimate company-specific income-statement drivers with cited evidence. A shared accounting calculator converts these estimates into GAAP diluted earnings per share (EPS), providing a common basis for comparing forecast assumptions.
We evaluate six models under a common harness, assessing point forecast error relative to date-matched analyst consensus and checking whether their prediction intervals achieve the stated coverage. Three models show no detectable paired difference in point forecast error from consensus, and three have higher error. Across all six models, nominal 80% intervals contain reported EPS in only 37--70% of runs that return an interval, and most misses fall above the upper bound because the forecasts run low. The driver records trace most of the error gap between models to operating-income estimates.
Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
arXiv:2610.38293 [cs.CL]
(or arXiv:2610.38293v1 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2610.38293 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 265 words
Related coverage
FinVerse: Financial Time-Series Benchmark
arXiv · Aug 5, 2026
Are We Really Benchmarking Forecasting Models? The Impact of Preprocessing on Time Series Performance
arXiv · Oct 8, 2026
FETS Benchmark: Foundation Models Enable Scalable and Generalizable Energy Time Series Forecasting
arXiv · Jul 21, 2026
WPBench: A Comprehensive Benchmark for Wind Power Forecasting
arXiv · Sep 22, 2026