{"paper_id":"2510.15691","title":"Exploring the Synergy of Quantitative Factors and Newsflow Representations from Large Language Models for Stock Return Prediction","authors":["Tian Guo","Emmanuel Hauptmann"],"paper_date":"2025-10-17","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2510.15691v3","page_url":"https://thequant.space/flowcharts/exploring-the-synergy-of-quantitative-factors-and-newsflow-r/","scores":{"empirical_rigor":6,"math_complexity":5.5,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes a multimodal fusion learning framework and a mixture model with a decoupled training scheme to combine quantitative factors and LLM-generated newsflow for stock return prediction.","quote":"First, we introduce a fusion learning framework to learn a unified representation from factors and newsflow representations generated by an LLM.","summary":"In quantitative investing, return prediction supports various tasks, including stock selection, portfolio optimization, and risk management. Quantitative factors, such as valuation, quality, and growth, capture various characteristics of stocks. Unstructured data, like news and transcripts, has…","strategy":"Mixture Decoupled"},"assumptions":{"status":"available","items":[{"assumption":"News information is quickly priced in and absorbed into quantitative factors.","type":"market","stated":"implied","quote":"like news, is often quickly priced in and absorbed into factors"},{"assumption":"Fine‑tuning the LLM may over‑emphasize already‑priced or noisy news information, potentially weakening performance.","type":"model","stated":"implied","quote":"fine‑tuning the LLM in multimodal models might cause it to overemphasize the already‑priced or noisy news information, thereby marginally affecting or even weakening performance"}]},"datasets":{"status":"available","items":[{"name":"North American investment universe","provider":"commercial data vendor","kind":"text-news","access":"commercial","period":null,"universe":"up to ~1,000 stocks","quote":"We have three datasets corresponding to the North American (NA), Emerging Markets (EM), and European (EU) investment universes, each containing up to ∼1, 000 stocks. We use company‑level financial news data provided by a commercial data vendor."},{"name":"Emerging Markets investment universe","provider":"commercial data vendor","kind":"text-news","access":"commercial","period":null,"universe":"up to ~1,000 stocks","quote":"We have three datasets corresponding to the North American (NA), Emerging Markets (EM), and European (EU) investment universes, each containing up to ∼1, 000 stocks. We use company‑level financial news data provided by a commercial data vendor."},{"name":"European investment universe","provider":"commercial data vendor","kind":"text-news","access":"commercial","period":null,"universe":"up to ~1,000 stocks","quote":"We have three datasets corresponding to the North American (NA), Emerging Markets (EM), and European (EU) investment universes, each containing up to ∼1, 000 stocks. We use company‑level financial news data provided by a commercial data vendor."}],"known":[],"sample":{"start":"2023","end":"2024","market":"North American Universe stocks","asset_class":"equities","frequency":"monthly"},"data_resolution":"unclear"},"transaction_costs":{"status":"available","modeled":"no","headline_net_of_costs":false,"detail":null,"quote":null,"source":"front_matter","rule":"R1","candidates":{"front_matter":"no","evidence_card":"unclear","evidence_card_quoted":false},"meaning":"a trading or execution result is reported without costs"},"results":{"status":"available","evaluation_type":"trading-pnl","findings":[{"finding":"Within fusion learning, the representation combination method generally outperforms other fusion methods despite its low architectural complexity.","kind":"performance","quote":"Within fusion learning, the representation combination method, despite its relatively low architectural complexity, generally outperforms other fusion methods."},{"finding":"Fine‑tuning the LLM generally has a detrimental effect on most multimodal methods, while the News Alone method benefits from fine‑tuning.","kind":"negative","quote":"for most of the multimodal methods (Fusion and Mixture groups), fine‑tuning appears to have a detrimental effect. Conversely, the single‑modal News Alone method, which relies solely on the LLM without factors, shows a notable improvement."},{"finding":"Fusion Combination remains the top performer in the long‑only portfolio even after fine‑tuning, though its performance declines relative to the no‑fine‑tuning baseline.","kind":"performance","quote":"Fusion Combination remains the top performer in the long‑only portfolio, although its performance declines."}],"reported_performance":{"strategy":"Mixture Decoupled","sharpe":1.78,"sharpe_benchmark":null,"annual_return_pct":33.77,"total_return_pct":null,"max_drawdown_pct":null},"out_of_sample":{"method":"holdout","headline_out_of_sample":true}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":18,"citations":{"cited_by_in_corpus":0,"by_independent_authors":0,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":0,"unclassified":0},"independent_papers":[],"scope":"Citations between papers in this corpus only; citations from outside it are not counted."}},"implementation":{"status":"available","public":false,"code_url":null,"note":"No repository presented by the authors as their own was found in the paper. Absence here is not proof that no code exists."},"replication":{"status":"none_found","card":null,"note":"No replication of this paper has been run by thequant.space yet."},"extraction":{"text_source":"fulltext","model":"openai/gpt-oss-120b","version":1,"extracted_at":"2026-11-14T17:24:22Z"},"provenance":{"claims":"Everything here is what the authors state in the paper. Nothing was re-run or reproduced by the platform.","quotes":"Assumptions, datasets, cost detail and findings were extracted by a language model. Every quote is verbatim paper text. Where an item has `support`, \"full\" means every figure in it appears in the quoted sentences and \"partial\" means a figure is stated elsewhere in the paper; items with a figure found nowhere in the paper were discarded.","datasets_known":"datasets.known lists well-known vendors and public datasets by a stable id, found by a dictionary over the full text where the sentence says the data is used; it is independent of the model and covers only names in the dictionary.","flags":"transaction_costs.modeled, results.evaluation_type and results.out_of_sample.method are platform classifications (structured-decision models); see https://thequant.space/score-guide/#penalty-flags","comparisons":"supports / contradicts / qualifies are a language model's reading of two papers' quoted findings, restricted to closely related papers that share no author. They are pointers to read, not verified replications, and coverage is partial.","implementation":"code_url is a repository the paper itself presents as its own (rule-scored, ambiguous cases judged by a model), checked for liveness when it was recorded."},"docs":"https://thequant.space/agents/"}