[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-120385-en":3,"doc-seo-120385-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},120385,2336464648746,"Skyler","https://ap-avatar.wpscdn.com/davatar_276721f389ce27ea32af1340a28f341c",8,"Research & Report","Out-of-sample predictability of firm-specific stock price crashes - A machine learning approach","Machine learning methods are used to predict firm-specific stock price crashes and to assess out-of-sample predictive accuracy relative to traditional regression models. Financial variables and textual signals from 10-K Management Discussion and Analysis sections are combined to forecast crash risk. Results show that logistic regression with financial inputs performs reasonably and can beat some newer classifiers. A stochastic gradient boosting model delivers consistent improvements, while well-chosen combinations of financial and textual inputs further raise performance, supporting the usefulness of ML for crash prediction.","DOI: 10.1111/jbfa.12831  \nARTICLE  \nOut-of-sample predictability of firm-specific stock price crashes: A machine learning approach  \nDevrimi Kaya1  Doron Reichmann2   Milan Reichmann3  \n1 Friedrich-Alexander-University of Erlangen-Nürnberg, Chair of Business Analytics and Sustainability, Nürnberg, Germany  \n2Accounting Department, Goethe University Frankfurt, Frankfurt, Germany  \n3 Chair of Banking and Finance, Leipzig University, Leipzig, Germany  \nCorrespondence  \nDoron Reichmann, Theodor-W.-Adorno-Platz 1, 60629 Frankfurt am Main, Germany. [Email: d.reichmann@econ.uni-frankfurt.de](Email: d.reichmann@econ.uni-frankfurt.de)  \nAbstract  \nWe use machine learning methods to predict firm-specific stock price crashes and evaluate the out-of-sample prediction performance of various methods, compared to traditional regression approaches. Using financial and textual data from10-K filings, our results show that a logistic regression with financial data inputs performs reasonably welland sometimes outperforms newer classifiers such as random forests and neural networks. However, we find that a stochastic gradient boosting model systematically outperforms the logistic regression, and forecasts using suitable combinations of financial and textual data inputs yield significantly higher prediction performance. Overall, the evidence suggests that machine learning methods can help predict stock price crashes.  \nKEYWORDS  \nmachine learning, natural language processing, stock price crash risk, textual disclosures  \n1  INTRODUCTION  \nStock price crashes are prevalent in international financial markets (An et al., 2018; Jin & Myers, 2006). Studies extensively show that economic factors, such as financial opacity, agency costs and managerial incentives, can help explain stock price crashes (Hong et al., 2017; Hutton et al., 2009; Kim et al., 2019; Kim et al.,  \nThis is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.  \n© 2024 The Author(s). Journal of Business Finance & Accounting published by John Wiley & Sons Ltd.  \nJ Bus Fin Acc. 2025;52:1095–1115. wi[leyonlinelibrary.com/journal/jbfa](leyonlinelibrary.com/journal/jbfa)  1095  \n1096  \nKAYA ET AL.  \n2011a).1 Most studies relate economic determinants to stock price crash risks using within-sample analyses. However, inferences about the predictability of stock price crashes are still limited. If investors fail to detect potential threats, stock prices can deviate from their fundamental values, increasing the risk of future price crashes. Hence, methods that help to identify crash-prone firms would offer significant value to investors.  \nWe use machine learning methods to predict firm-specific stock price crashes.2 Specifically, we evaluate the out-ofsample prediction performance of machine learning, compared to traditional regression approaches. While research on machine learning in accounting and finance typically focuses on numerical data as predictors (Bali et al., 2023; Chenet al., 2022; Gu et al., 2020), we also consider textual disclosures as an integral part of the financial reporting package. For instance, Lewis and Young (2019) document a substantial increase in textual firm disclosures over time. Machine learning methods can uncover complex patterns in both financial and textual data that help predict firm outcomes (e.g., Bertomeu et al., 2021; Bochkay et al., 2023; El-Haj et al., 2020). Hence, our study aims to leverage these empirical methods and test their performance for out-of-sample stock price crash predictions.  \nTo conduct our analyses, we collect a sample of 39,583 US firm-year observations from the period 1996 to 2018 and calculate a wide set of 37 financial data inputs. We further retrieve the Management Discussion and Analysis (MD&A) sections of 10-K filings from the SEC’s online EDGAR system to build textual inputs because resear","cbCaifFtG4kGZrYS","https://ap.wps.com/l/cbCaifFtG4kGZrYS","pdf",901659,1,21,"English","en",105,"# Introduction\n## Data and Research Design\n## Modeling Approaches\n## Text Processing and Feature Construction\n## Training and Evaluation","[{\"question\":\"What is the main goal of the study?\",\"answer\":\"The study aims to predict firm-specific stock price crashes and evaluate how well different methods perform in out-of-sample settings versus traditional regression models.\"},{\"question\":\"Which data sources are used for prediction?\",\"answer\":\"It uses financial data inputs and textual information extracted from the Management Discussion and Analysis (MD\\u0026A) sections of SEC 10-K filings from EDGAR.\"},{\"question\":\"Which models show the strongest out-of-sample performance?\",\"answer\":\"A stochastic gradient boosting model systematically outperforms logistic regression, and combining suitable financial and textual inputs yields significantly higher prediction performance.\"}]","Out-of-sample predictability of firm-specific stock price crashes - A machine learning approach | PDF",1785729770,53,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"out-of-sample-predictability-of-firm-specific-stock-price-crashes-a-machine-learning-approach","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/out-of-sample-predictability-of-firm-specific-stock-price-crashes-a-machine-learning-approach/120385/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-03",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What is the main goal of the study?","Question",{"text":75,"@type":76},"The study aims to predict firm-specific stock price crashes and evaluate how well different methods perform in out-of-sample settings versus traditional regression models.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"Which data sources are used for prediction?",{"text":80,"@type":76},"It uses financial data inputs and textual information extracted from the Management Discussion and Analysis (MD&A) sections of SEC 10-K filings from EDGAR.",{"name":82,"@type":73,"acceptedAnswer":83},"Which models show the strongest out-of-sample performance?",{"text":84,"@type":76},"A stochastic gradient boosting model systematically outperforms logistic regression, and combining suitable financial and textual inputs yields significantly higher prediction performance.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]