{"paper_id":"2402.04775","title":"Cyber risk and the cross-section of stock returns","authors":["Daniel Celeny","Loïc Maréchal"],"paper_date":"2024-02-07","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2402.04775v2","page_url":"https://thequant.space/flowcharts/cyber-risk-and-the-cross-section-of-stock-returns/","scores":{"empirical_rigor":6,"math_complexity":4.5,"quadrant":"Street Traders"},"proposes":{"status":"available","statement":"The paper proposes a Paragraph Vector machine‑learning method to quantify firms' cyber risk from their 10‑K disclosures and uses this score in asset‑pricing tests.","quote":"the “Paragraph Vector” algorithm to estimate firms’ cyber risk based on their 10‑K statements.","summary":"We extract firms’ cyber risk with a machine learning algorithm measuring the proximity between their disclosures and a dedicated cyber corpus. Our approach outperforms dictionary methods, uses full disclosure and not devoted-only sections, and generates a cyber risk measure uncorrelated with other…","strategy":"high cyber risk quantile portfolio"},"assumptions":{"status":"available","items":[{"assumption":"The cyber risk score can be derived without calibrating to past cyber‑attack events.","type":"model","stated":"explicit","quote":"Our method is free of cyber‑attack calibration, focuses on the cybersecurity context, and consistently generates a non‑zero score."},{"assumption":"All firms in the sample have at least one 10‑K filing available.","type":"data","stated":"explicit","quote":"We write a Python script that queries all available information from WRDS’ API and filters the firms based on the existence of a 10‑K filing with the SEC so that all of the retained firms have at least one 10‑K statement available."},{"assumption":"Monthly stock returns and financial ratios are accurately captured from CRSP and Compustat for the sample period.","type":"data","stated":"explicit","quote":"We extract monthly stock returns and financial ratios for 7,059 firms between January 2007 and December 2022."},{"assumption":"The factor data (HML, SMB, MOM, CMA, RMW) from Kenneth French are appropriate benchmarks.","type":"model","stated":"explicit","quote":"We also download the one‑month Treasury bill rate and returns on the market, book‑to‑market (HML), size (SMB), momentum (MOM), investment (CMA), and operating profitability (RMW) factors from the Kenneth French data repository."},{"assumption":"The Paragraph Vector algorithm can capture latent cyber‑risk exposure rather than post‑attack recovery effects.","type":"statistical","stated":"implied","quote":"Using the Paragraph Vector machine learning algorithm of Le and Mikolov (2014), our method"}]},"datasets":{"status":"available","items":[{"name":"CRSP monthly stock returns","provider":"CRSP","kind":"market-prices","access":"public","period":"January 2007–December 2022","universe":"7,059 US‑listed firms","quote":"... for 7,059 firms between January 2007 and December 2022.","known_id":"crsp"},{"name":"Compustat fundamentals","provider":"S&P Global Market Intelligence","kind":"fundamentals","access":"commercial","period":"January 2007–December 2022","universe":"7,059 US‑listed firms","quote":"... S&P Global Market Intelligence’s Compustat database.","known_id":"compustat"},{"name":"SEC 10‑K filings","provider":"SEC Edgar","kind":"text-news","access":"public","period":"2007–2022","universe":"All firms with a 10‑K filing in the sample","quote":"We use 10‑K statements to build a cyber risk score. To download these statements, we use the index files from the SEC’s Edgar archives."},{"name":"MITRE ATT&CK knowledgebase","provider":"MITRE","kind":"alternative","access":"public","period":null,"universe":null,"quote":"We use a machine learning algorithm, the “Paragraph Vector”, in combination with the MITRE ATT&CK cybersecurity knowledgebase..."},{"name":"Kenneth French factor data","provider":"Kenneth French data repository","kind":"benchmark","access":"public","period":"2007–2022","universe":null,"quote":"We also download the one‑month Treasury bill rate and returns on the market, book‑to‑market (HML), size (SMB), momentum (MOM), investment (CMA), and operating profitability (RMW) factors from the Kenneth French data repository.","known_id":"french-library"}],"known":[{"id":"wrds","name":"WRDS","provider":"Wharton Research Data Services","kind":"other","access":"commercial","quote":"Market data We download public equity data from Wharton Research Data Services (WRDS), using the data from the Center for Research in Security Prices (CRSP) and S&P Global Market Intelligence’s Compustat database.","mentions":3,"used_in":1},{"id":"crsp","name":"CRSP","provider":"Center for Research in Security Prices","kind":"market-prices","access":"commercial","quote":"Market data We download public equity data from Wharton Research Data Services (WRDS), using the data from the Center for Research in Security Prices (CRSP) and S&P Global Market Intelligence’s Compustat database.","mentions":2,"used_in":1},{"id":"compustat","name":"Compustat","provider":"S&P Global","kind":"fundamentals","access":"commercial","quote":"Market data We download public equity data from Wharton Research Data Services (WRDS), using the data from the Center for Research in Security Prices (CRSP) and S&P Global Market Intelligence’s Compustat database.","mentions":1,"used_in":1}],"sample":{"start":null,"end":null,"market":"US-listed stocks","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":"A portfolio of high‑cyber‑risk stocks yields an excess return of 18.72% per annum.","kind":"performance","quote":"We find that a portfolio of US‑listed stocks in the high cyber risk quantile generates an excess return of 18.72% p.a."},{"finding":"A long‑short cyber‑risk portfolio earns a risk premium of 6.93% per annum, robust to factor benchmarks.","kind":"performance","quote":"Moreover, a long‑short cyber risk portfolio has a significant and positive risk premium of 6.93% p.a., robust to all factors' benchmarks."},{"finding":"The long‑short strategy on cyber‑risk sorted portfolios produces an average monthly excess return of 0.56% with a statistically significant alpha.","kind":"performance","quote":"Indeed, a long‑short strategy on cyber risk sorted portfolios has a positive and statistically significant alpha over traditional factor models and an average monthly excess return of 0.56%."},{"finding":"Including the cyber‑risk factor improves the pricing of stocks in Bayesian factor‑model selection, showing it is an essential factor.","kind":"pricing-calibration","quote":"Bayesian asset pricing method, we show that our cyber risk factor is the essential feature that allows any multi‑factor model to price the cross‑section of stock returns."},{"finding":"The cyber‑risk score is largely uncorrelated with other firm characteristics and outperforms a dictionary‑based score.","kind":"statistical","quote":"... generates a cyber risk measure uncorrelated with other firms' characteristics."}],"reported_performance":{"strategy":"high cyber risk quantile portfolio","sharpe":null,"sharpe_benchmark":null,"annual_return_pct":18.72,"total_return_pct":null,"max_drawdown_pct":null},"out_of_sample":{"method":"unclear","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":20,"citations":{"cited_by_in_corpus":2,"by_independent_authors":0,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":0,"unclassified":2},"independent_papers":[],"scope":"Citations between papers in this corpus only; citations from outside it are not counted."}},"implementation":{"status":"available","public":true,"code_url":"https://github.com/technometrics-lab/17-Cyber-risk_and_the_cross-section_of_stock_returns","note":null},"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":"2026-11-14T18:15:24Z"},"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/"}