{"paper_id":"2609.34169","title":"From Word Counts to Context: Topic Models for Asset Pricing","authors":["Kevin Foley","Jonathan Hartadi","Shivesh Prakash","Swapnil Vatsal"],"paper_date":"2026-09-28","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2609.34169","page_url":"https://thequant.space/flowcharts/from-word-counts-to-context-topic-models-for-asset-pricing/","scores":{"empirical_rigor":6,"math_complexity":6.5,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes replacing the LDA topic model with a frozen sentence transformer in a narrative asset‑pricing pipeline.","quote":"the LDA topic model with a frozen sentence transformer","summary":"News may reveal systematic risk, but whether its context enhances the construction of systematic risk factors is still unclear. We seek to test whether utilizing a sentence transformer represents an improvement over techniques such as Latent Dirichlet Allocation (LDA) in the coherence of topic term…","strategy":"Spherical sentence‑transformer with multi‑horizon IPCA"},"assumptions":{"status":"available","items":[{"assumption":"The downstream pipeline (shock construction, covariance, Sparse IPCA, portfolio evaluation) is held fixed while only the text layer changes.","type":"model","stated":"explicit","quote":"the downstream construction of shocks, stock-level covariance characteristics, Sparse IPCA factors, and out-of-sample portfolios is held fixed while the text layer changes"},{"assumption":"LDA observes the full cleaned article whereas the sentence transformer uses up to three lead passages.","type":"data","stated":"explicit","quote":"LDA observes the full cleaned article, whereas the sentence transformer uses up to three lead passages"},{"assumption":"Both models are trained on the same FNSPID financial‑news corpus covering 2015‑2023.","type":"data","stated":"explicit","quote":"Both the LDA and transformer models use the full news corpus from January 2015 to December 2023 to construct topics"},{"assumption":"The evaluation window is limited to 41 months, which may affect statistical significance.","type":"statistical","stated":"explicit","quote":"The evidence remains limited by a short, 41-month evaluation window"},{"assumption":"Topic count is fixed at 60 for both models, and the vocabulary is shared.","type":"model","stated":"explicit","quote":"Both the LDA and transformer models use the same collection of FNSPID articles, and both produce 60 topics in total"}]},"datasets":{"status":"available","items":[{"name":"FNSPID financial‑news corpus","provider":"Dong et al., 2024","kind":"text-news","access":"public","period":"2015-2023","universe":null,"quote":"FNSPID newswire text from 2015 to 2023 (Dong et al., 2024)","known_id":"fnspid"},{"name":"CRSP US Stock Databases","provider":"Center for Research in Security Prices (CRSP)","kind":"market-prices","access":"commercial","period":null,"universe":null,"quote":"Daily stock returns and market‑capitalization data come from the CRSP US Stock Databases, accessed through the Wharton Research Data Services platform","known_id":"crsp"},{"name":"Fama–French daily factor file","provider":"Kenneth R. French Data Library","kind":"macro","access":"public","period":null,"universe":null,"quote":"The Fama–French daily factor file used to obtain the risk‑free rate is publicly available from the Kenneth R. French Data Library","known_id":"french-library"}],"known":[{"id":"fnspid","name":"FNSPID","provider":"Dong et al.","kind":"text-news","access":"public","quote":"Their Wall Street Journal sample runs from 1984 to 2017 and supports 180 topics; ours is FNSPID newswire text from 2015 to 2023 (Dong et al., 2024) and supports 60.","mentions":9,"used_in":3},{"id":"crsp","name":"CRSP","provider":"Center for Research in Security Prices","kind":"market-prices","access":"commercial","quote":"Daily stock returns and market-capitalization data come from the CRSP US Stock Databases, accessed through the Wharton Research Data Services platform (Center for Research in Security Prices, n.d.).","mentions":4,"used_in":2},{"id":"wrds","name":"WRDS","provider":"Wharton Research Data Services","kind":"other","access":"commercial","quote":"Daily stock returns and market-capitalization data come from the CRSP US Stock Databases, accessed through the Wharton Research Data Services platform (Center for Research in Security Prices, n.d.).","mentions":2,"used_in":1}],"sample":{"start":"2020-09","end":"2024-01","market":"US stocks (CRSP panel)","asset_class":"equities","frequency":"monthly"},"data_resolution":"daily"},"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":"The sentence transformer branch yields higher topic coherence measured by NPMI compared to LDA.","kind":"statistical","quote":"We find that the sentence transformer branch had higher observed scores both in terms of coherence (measured by NPMI) as well as financial performance (measured by Sharpe)"},{"finding":"The transformer model achieves higher out‑of‑sample Sharpe ratios than LDA on the same financial window.","kind":"performance","quote":"With the topic count and shared vocabulary held fixed across models, the transformer model yields higher topic coherence and OOS Sharpe ratios"},{"finding":"The spherical ST with grouped multi‑horizon IPCA specification attains the highest observed excess‑return Sharpe of approximately 1.03.","kind":"performance","quote":"combined specification attained a Sharpe ratio of approximately 1.03, which was the highest among the model specifications considered"},{"finding":"The reported out‑of‑sample tests do not establish a statistically significant improvement in Sharpe ratios.","kind":"negative","quote":"the OOS tests performed do not establish a statistically significant improvement in Sharpe ratios"}],"reported_performance":{"strategy":"Spherical sentence‑transformer with multi‑horizon IPCA","sharpe":1.0334,"sharpe_benchmark":null,"annual_return_pct":null,"total_return_pct":null,"max_drawdown_pct":62.23},"out_of_sample":{"method":"unclear","headline_out_of_sample":null}},"independent_evidence":{"status":"available","supports":[],"contradicts":[],"qualifies":[],"compared_with":26,"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":"2611.14325","title":"Point-in-Time Encoders for Earnings Calls","paper_date":"2026-11-10","authors":["Alessio Brini"],"empirical_rigor":8,"page_url":"https://thequant.space/flowcharts/point-in-time-encoders-for-earnings-calls/"},"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":"2026-11-14T17:04:20Z"},"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/"}