aiqbench analysis
The XForecast benchmark tests whether models can combine stock price histories with paired news text to predict four-week price movements, with a weighted hit rate that emphasizes cases where news should matter most. The main value for ML readers is the multimodal setup and the evaluation design, since time-series and text fusion benchmarks are still sparse and the reported results will show whether text adds signal beyond price data. Treat the framing as a benchmark proposal for now: the claimed benefit of news inputs depends on the quality of the FinTexTS-generated text and on how the Kaggle results hold up out of sample.
XForecast: The KDD 2026 Finance Day Challenge on News-Grounded Financial Forecasting
arXiv:2610.16131v1 Announce Type: new Abstract: Financial forecasting has traditionally relied on numerical time-series (TS) signals, while textual information such as news remains largely underutilized. We introduce XForecast, a benchmark and competition for news-grounded financial forecasting, where the task is to predict stock price movements four weeks ahead using both historical price data and paired news signals. XForecast combines stock price data with curated financial text generated using FinTexTS, a multimodal dataset generation method. It covers 100 large-cap stocks from 2019--2023, with training on 2019--2022 and evaluation on 2023. To better evaluate models that leverage textual signals, we adopt a weighted hit rate metric that places greater emphasis on instances where news information is expected to have stronger influence. The XForecast Challenge, hosted on Kaggle as part of KDD 2026 Finance Day, aims to benchmark methods for integrating financial TS and textual information.
Abstract:Financial forecasting has traditionally relied on numerical time-series (TS) signals, while textual information such as news remains largely underutilized. We introduce XForecast, a benchmark and competition for news-grounded financial forecasting, where the task is to predict stock price movements four weeks ahead using both historical price data and paired news signals. XForecast combines stock price data with curated financial text generated using FinTexTS, a multimodal dataset generation method. It covers 100 large-cap stocks from 2019--2023, with training on 2019--2022 and evaluation on 2023. To better evaluate models that leverage textual signals, we adopt a weighted hit rate metric that places greater emphasis on instances where news information is expected to have stronger influence. The XForecast Challenge, hosted on Kaggle as part of KDD 2026 Finance Day, aims to benchmark methods for integrating financial TS and textual information.
Machine Learning (cs.LG)
arXiv:2610.16131 [cs.LG]
(or arXiv:2610.16131v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2610.16131 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
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 · 231 words
