{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,9]],"date-time":"2026-10-09T06:49:41Z","timestamp":1791528581612,"version":"4.3.4"},"posted":{"date-parts":[[2026]]},"group-title":"SSRN","reference-count":0,"publisher":"Elsevier BV","license":[{"start":{"date-parts":[[2026,1,1]],"date-time":"2026-01-01T00:00:00Z","timestamp":1767225600000},"content-version":"unspecified","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"abstract":"<jats:p>Conference-call-based volatility prediction has mainly followed an isolated information modeling paradigm, treating each transcript as an isolated information source and each forecast horizon as an independent task. This limits comparison with economically comparable peers and the use of risk information shared across horizons. We propose an LLM-Guided Hypergraph Learning framework to address these limitations. We first use LLMs to construct a shared pool of standardized business-condition factors and then build factor-transcript hypergraphs that connect disclosures across firms through shared factors, enabling the model to retrieve comparable historical peer information. The model then explicitly compares focal and peer information and jointly predicts volatility over 5-, 10-, 20-, and 60-day horizons through multi-task learning with horizon-specific modulation. Experiments on 14,896 U.S. earnings conference calls show that the framework consistently outperforms strong transcript-based, graph-based, and financial baselines. Ablation studies confirm the importance of factor-guided peer retrieval, explicit comparison, and cross-horizon learning.<\/jats:p>","DOI":"10.2139\/ssrn.7587198","type":"posted-content","created":{"date-parts":[[2026,10,9]],"date-time":"2026-10-09T06:00:21Z","timestamp":1791525621000},"source":"Crossref","is-referenced-by-count":0,"title":["Beyond Isolation: LLM-guided Hypergraph Learning across Documents and Horizons for Volatility Prediction from Conference Calls"],"prefix":"10.2139","author":[{"given":"Peizhe","family":"Li","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Liang","family":"Zhang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"78","container-title":[],"original-title":[],"deposited":{"date-parts":[[2026,10,9]],"date-time":"2026-10-09T06:00:21Z","timestamp":1791525621000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.ssrn.com\/abstract=7587198"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026]]},"references-count":0,"URL":"https:\/\/doi.org\/10.2139\/ssrn.7587198","relation":{},"subject":[],"published":{"date-parts":[[2026]]},"subtype":"preprint"}}