{"paper_id":"2505.06864","title":"NewsNet-SDF: Stochastic Discount Factor Estimation with Pretrained Language Model News Embeddings via Adversarial Networks","authors":["Shunyao Wang","Ming Cheng","Christina Dan Wang"],"paper_date":"2025-05-11","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2505.06864v1","page_url":"https://thequant.space/flowcharts/newsnet-sdf-stochastic-discount-factor-estimation-with-pret/","scores":{"empirical_rigor":6,"math_complexity":8,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes NewsNet‑SDF, a novel deep learning framework that integrates pretrained language model news embeddings with stochastic discount factor estimation via adversarial networks.","quote":"We introduce NewsNet‑SDF, a novel deep learning framework that seamlessly integrates pretrained language model embeddings with financial time series through adversarial networks.","summary":"Stochastic Discount Factor (SDF) models provide a unified framework for asset pricing and risk assessment, yet traditional formulations struggle to incorporate unstructured textual information. We introduce NewsNet-SDF, a novel deep learning framework that seamlessly integrates pretrained language…","strategy":"NewsNet-SDF"},"assumptions":{"status":"available","items":[{"assumption":"The fundamental asset pricing moment condition Et[Mt+1Re t+1] = 0 must hold.","type":"model","stated":"explicit","quote":"These diverse information sources are fused and fed into parallel SDF and conditional networks trained adversarially to enforce the fundamental asset pricing equation Et[Mt+1Re t+1] = 0."},{"assumption":"Stocks have all required firm characteristics available each month for inclusion.","type":"data","stated":"explicit","quote":"Following [4], we restrict our analysis to stocks with all required characteristics available in a given month, yielding approximately 10,000 unique securities."}]},"datasets":{"status":"available","items":[{"name":"News articles","provider":"The New York Times","kind":"text-news","access":"unknown","period":"1980-2022","universe":"U.S. equities","quote":"Our textual data consists of approximately 2.5 million articles from The New York Times (1980-2022), processed through our attention-based neural framework with careful temporal alignment to financial data."},{"name":"U.S. equity data","provider":null,"kind":"market-prices","access":"unknown","period":"1980-2022","universe":"approximately 10,000 unique securities","quote":"Empirical evaluations on U.S. equity data (1980-2022) demonstrate NewsNet‑SDF substantially outperforms alternatives with a Sharpe ratio of 2.803."},{"name":"Firm characteristics","provider":null,"kind":"fundamentals","access":"unknown","period":null,"universe":"56 firm‑specific characteristics","quote":"We construct 56 firm‑specific characteristics following [37] and [38], spanning valuation, profitability, financial soundness, capitalization, operational efficiency, and momentum categories."},{"name":"Macroeconomic time series","provider":"FRED‑MD database","kind":"macro","access":"unknown","period":null,"universe":"157 series","quote":"For macroeconomic data, we incorporate 157 time series from the FRED‑MD database and other sources, covering output, employment, housing, and financial market indicators.","known_id":"fred"}],"known":[{"id":"fred","name":"FRED","provider":"Federal Reserve Bank of St. Louis","kind":"macro","access":"public","quote":"For macroeconomic data, we incorporate 157 time series from the FRED-MD database and other sources, covering output, employment, housing, and financial market indicators.","mentions":1,"used_in":1}],"sample":{"start":"1980","end":"2022","market":"U.S. equity market","asset_class":"equities","frequency":null},"data_resolution":"weekly-monthly"},"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":"no","evidence_card_quoted":false},"meaning":"a trading or execution result is reported without costs"},"results":{"status":"available","evaluation_type":"trading-pnl","findings":[{"finding":"NewsNet‑SDF achieves a Sharpe ratio of 2.80, far exceeding traditional factor models.","kind":"performance","quote":"Empirical evaluations on U.S. equity data (1980-2022) demonstrate NewsNet‑SDF substantially outperforms alternatives with a Sharpe ratio of 2.803."},{"finding":"The model improves Sharpe ratio by 471% over CAPM and over 200% versus conventional SDF implementations.","kind":"performance","quote":"The model shows a 471% improvement over CAPM, over 200% improvement versus traditional SDF implementations."},{"finding":"Pricing errors are reduced by 74% compared to the Fama‑French five‑factor model.","kind":"pricing-calibration","quote":"We enhance SDF estimation through a novel loss function that dynamically reweights pricing errors based on factor exposure, reducing pricing errors by 74% compared to the Fama‑French five‑factor model."},{"finding":"Ablation of news embeddings cuts Sharpe ratio by 41%, while removing macroeconomic features cuts it by 31%, indicating news provides unique predictive signals.","kind":"predictive-accuracy","quote":"Ablation studies reveal that eliminating news text causes a 41% decrease in Sharpe ratio versus a 31% reduction when removing macroeconomic features."},{"finding":"News‑derived principal components rank among the most influential determinants of SDF dynamics.","kind":"statistical","quote":"Ablation studies confirm that text embeddings contribute significantly more to model performance than macroeconomic features, with news‑derived principal components ranking among the most influential determinants of SDF dynamics."}],"reported_performance":{"strategy":"NewsNet-SDF","sharpe":2.8,"sharpe_benchmark":null,"annual_return_pct":null,"total_return_pct":null,"max_drawdown_pct":null},"out_of_sample":{"method":"holdout","headline_out_of_sample":true}},"independent_evidence":{"status":"available","supports":[],"contradicts":[],"qualifies":[],"compared_with":62,"citations":{"cited_by_in_corpus":1,"by_independent_authors":1,"how_cited":{"used_as_baseline":0,"uses_method_or_data":1,"extends":0,"discusses_findings":0,"mentions":0,"unclassified":0},"independent_papers":[{"paper":{"paper_id":"2602.12490","title":"Transformer-based CoVaR: Systemic Risk in Textual Information","paper_date":"2026-02-13","authors":["Junyu Chen","Tom Boot","Lingwei Kong","Weining Wang"],"empirical_rigor":8.5,"page_url":"https://thequant.space/flowcharts/transformer-based-covar-systemic-risk-in-textual-information/"},"how":"uses_method_or_data","quote":null}],"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":"2026-11-14T17:38:52Z"},"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/"}