{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,23]],"date-time":"2026-10-23T19:49:30Z","timestamp":1792784970347,"version":"4.5.0"},"posted":{"date-parts":[[2024]]},"group-title":"SSRN","reference-count":70,"publisher":"Elsevier BV","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"abstract":"<jats:p>&lt;div&gt;\n &lt;div&gt;\n  We fit more than 1,000 machine&amp;nbsp;&lt;span&gt;learning models to predict stock returns, systematically&amp;nbsp;&lt;\/span&gt;&lt;span&gt;varying key design choices across algor&lt;\/span&gt;&lt;span&gt;ithms, target variables, feature sets, and training&amp;nbsp;&lt;\/span&gt;&lt;span&gt;methodologies. The resulting nonstandard error of portfolio returns exceeds the standard&amp;nbsp;&lt;\/span&gt;&lt;span&gt;error by 59%, with monthly mean top-minus-bottom portfolio returns ranging from 0.13% to&amp;nbsp;&lt;\/span&gt;&lt;span&gt;1.98%. These findings demonstrate that design choices are a first-order source of uncertainty&amp;nbsp;&lt;\/span&gt;&lt;span&gt;in machine learning return pred&lt;\/span&gt;&lt;span&gt;ictions, and we provide recommendations for model design.&amp;nbsp;&lt;\/span&gt;&lt;span&gt;We further identify the conditions under which non-linear models outperform linear models.&amp;nbsp;&lt;\/span&gt;&lt;span&gt;Finally, we document that superior design choices persist over time and are identifiable in&amp;nbsp;&lt;\/span&gt;&lt;span&gt;real time.&lt;\/span&gt;\n &lt;\/div&gt;\n&lt;\/div&gt;<\/jats:p>","DOI":"10.2139\/ssrn.5031755","type":"posted-content","created":{"date-parts":[[2024,12,2]],"date-time":"2024-12-02T13:44:37Z","timestamp":1733147077000},"source":"Crossref","is-referenced-by-count":3,"title":["Design choices, machine learning, and the cross-section of stock returns"],"prefix":"10.2139","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-1223-2373","authenticated-orcid":true,"given":"Minghui","family":"Chen","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0445-8361","authenticated-orcid":true,"given":"Matthias Xaver","family":"Hanauer","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9674-6039","authenticated-orcid":true,"given":"Tobias","family":"Kalsbach","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"78","reference":[{"key":"ref1","doi-asserted-by":"crossref","first-page":"193","DOI":"10.1214\/aoms\/1177729437","article-title":"Asymptotic theory of certain \"goodness of fit\" criteria based on stochastic processes","volume":"23","author":"T W Anderson","year":"1952","journal-title":"The Annals of Mathematical Statistics"},{"key":"ref2","doi-asserted-by":"crossref","first-page":"3034","DOI":"10.1093\/rfs\/hhad006","article-title":"Factor momentum","volume":"36","author":"R D Arnott","year":"2023","journal-title":"The Review of Financial Studies"},{"key":"ref3","doi-asserted-by":"crossref","first-page":"929","DOI":"10.1111\/jofi.12021","article-title":"Value and momentum everywhere","volume":"68","author":"C S Asness","year":"2013","journal-title":"The Journal of Finance"},{"key":"ref4","doi-asserted-by":"crossref","first-page":"2587","DOI":"10.1287\/mnsc.2022.4449","article-title":"Machine learning vs. economic restrictions: Evidence from stock return predictability","volume":"69","author":"D Avramov","year":"2023","journal-title":"Management Science"},{"key":"ref5","doi-asserted-by":"crossref","first-page":"419","DOI":"10.1057\/s41260-023-00318-z","article-title":"Stock market anomalies and machine learning across the globe","volume":"24","author":"V Azevedo","year":"2023","journal-title":"Journal of Asset Management"},{"key":"ref6","doi-asserted-by":"crossref","first-page":"3548","DOI":"10.1093\/rfs\/hhad017","article-title":"Option return predictability with machine learning and big data","volume":"36","author":"T G Bali","year":"2023","journal-title":"The Review of Financial Studies"},{"key":"ref7","author":"T G Bali","year":"2026","journal-title":"Machine forecast disagreement. The Review of Financial Studies, forthcoming"},{"key":"ref8","doi-asserted-by":"crossref","first-page":"1090","DOI":"10.1093\/rfs\/hhaa098","article-title":"2021a. Corrigendum: Bond risk premiums with machine learning","volume":"34","author":"D Bianchi","journal-title":"The Review of Financial Studies"},{"key":"ref9","doi-asserted-by":"crossref","first-page":"1046","DOI":"10.1093\/rfs\/hhaa062","article-title":"Bond risk premiums with machine learning","volume":"34","author":"D Bianchi","year":"2021","journal-title":"The Review of Financial Studies"},{"key":"ref10","doi-asserted-by":"crossref","first-page":"40","DOI":"10.3905\/jfds.2023.1.135","article-title":"The term structure of machine learning alpha","volume":"5","author":"D Blitz","year":"2023","journal-title":"Journal of Financial Data Science"},{"key":"ref11","doi-asserted-by":"crossref","first-page":"5","DOI":"10.1023\/A:1010933404324","article-title":"Random forests","volume":"45","author":"L Breiman","year":"2001","journal-title":"Machine Learning"},{"key":"ref12","doi-asserted-by":"crossref","first-page":"803","DOI":"10.1093\/rfs\/hhae036","article-title":"Missing financial data","volume":"38","author":"S Bryzgalova","year":"2025","journal-title":"The Review of Financial Studies"},{"key":"ref13","doi-asserted-by":"crossref","DOI":"10.1016\/j.jedc.2023.104725","article-title":"Machine learning goes global: Crosssectional return predictability in international stock markets","volume":"155","author":"N Cakici","year":"2023","journal-title":"Journal of Economic Dynamics and Control"},{"key":"ref14","doi-asserted-by":"crossref","DOI":"10.1016\/j.irfa.2024.103244","article-title":"Machine learning and the cross-section of cryptocurrency returns","volume":"94","author":"N Cakici","year":"2024","journal-title":"International Review of Financial Analysis"},{"key":"ref15","doi-asserted-by":"crossref","DOI":"10.1016\/j.irfa.2024.103569","article-title":"What drives stock returns across countries? Insights from machine learning models","volume":"96","author":"N Cakici","year":"2024","journal-title":"International Review of Financial Analysis"},{"key":"ref16","author":"N Cakici","year":"2026","journal-title":"Getting the target right in return prediction"},{"key":"ref17","doi-asserted-by":"crossref","DOI":"10.1016\/j.jbankfin.2026.107658","article-title":"The more, the better? Predicting stock returns with local and global data","volume":"186","author":"N Cakici","year":"2026","journal-title":"Journal of Banking & Finance"},{"key":"ref18","doi-asserted-by":"crossref","first-page":"57","DOI":"10.1111\/j.1540-6261.1997.tb03808.x","article-title":"On persistence in mutual fund performance","volume":"52","author":"M M Carhart","year":"1997","journal-title":"The Journal of Finance"},{"key":"ref19","doi-asserted-by":"crossref","first-page":"207","DOI":"10.1561\/104.00000112","article-title":"Open source cross-sectional asset pricing","volume":"11","author":"A Y Chen","year":"2022","journal-title":"Critical Finance Review"},{"key":"ref20","doi-asserted-by":"crossref","first-page":"714","DOI":"10.1287\/mnsc.2023.4695","article-title":"Deep learning in asset pricing","volume":"70","author":"L Chen","year":"2024","journal-title":"Management Science"},{"key":"ref21","doi-asserted-by":"crossref","DOI":"10.1016\/j.econlet.2025.112641","article-title":"Model complexity and the performance of global versus regional models","volume":"257","author":"M Chen","year":"2025","journal-title":"Economics Letters"},{"key":"ref22","doi-asserted-by":"crossref","first-page":"1047","DOI":"10.1111\/j.1540-6261.2011.01671.x","article-title":"Presidential address: Discount rates","volume":"66","author":"J H Cochrane","year":"2011","journal-title":"The Journal of Finance"},{"key":"ref23","author":"G Coqueret","year":"2026","journal-title":"Label alchemy: Target engineering for improved stock selection"},{"key":"ref24","doi-asserted-by":"crossref","first-page":"1074","DOI":"10.1017\/S0022109024000036","article-title":"Estimating stock market betas via machine learning","volume":"60","author":"W Drobetz","year":"2025","journal-title":"Journal of Financial and Quantitative Analysis"},{"key":"ref25","author":"T L Dudda","year":"2025","journal-title":"Evaluating machine learning in business and economics research: Transparency, predictive power, and practical costs"},{"key":"ref26","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1016\/j.jfineco.2014.10.010","article-title":"A five-factor asset pricing model","volume":"116","author":"E F Fama","year":"2015","journal-title":"Journal of Financial Economics"},{"key":"ref27","doi-asserted-by":"crossref","DOI":"10.1016\/j.jbankfin.2024.107372","article-title":"Predicting individual corporate bond returns","volume":"171","author":"G Feng","year":"2025","journal-title":"Journal of Banking & Finance"},{"key":"ref28","doi-asserted-by":"crossref","DOI":"10.1016\/j.irfa.2024.103106","article-title":"Non-standard errors in the cryptocurrency world","volume":"92","author":"C Fieberg","year":"2024","journal-title":"International Review of Financial Analysis"},{"key":"ref29","doi-asserted-by":"crossref","first-page":"760","DOI":"10.1093\/rfs\/hhae003","article-title":"Missing data in asset pricing panels","volume":"38","author":"J Freyberger","year":"2025","journal-title":"The Review of Financial Studies"},{"key":"ref30","doi-asserted-by":"crossref","first-page":"2326","DOI":"10.1093\/rfs\/hhz123","article-title":"Dissecting Characteristics Nonparametrically","volume":"33","author":"J Freyberger","year":"2020","journal-title":"The Review of Financial Studies"},{"key":"ref31","author":"R Gao","year":"2026","journal-title":"Nonstandard errors in AI agents"},{"key":"ref32","doi-asserted-by":"crossref","first-page":"2223","DOI":"10.1093\/rfs\/hhaa009","article-title":"Empirical asset pricing via machine learning","volume":"33","author":"S Gu","year":"2020","journal-title":"The Review of Financial Studies"},{"key":"ref33","article-title":"Cross-cryptocurrency return predictability","volume":"163","author":"L Guo","year":"2024","journal-title":"Journal of Economic Dynamics and Control"},{"key":"ref34","doi-asserted-by":"crossref","first-page":"13","DOI":"10.3905\/jpm.2019.45.3.013","article-title":"Factor momentum everywhere","volume":"45","author":"T Gupta","year":"2019","journal-title":"The Journal of Portfolio Management"},{"key":"ref35","doi-asserted-by":"crossref","DOI":"10.1016\/j.ememar.2023.101022","article-title":"Machine learning and the cross-section of emerging market stock returns","volume":"55","author":"M X Hanauer","year":"2023","journal-title":"Emerging Markets Review"},{"key":"ref36","doi-asserted-by":"crossref","DOI":"10.1016\/j.frl.2022.102856","article-title":"Boosting agnostic fundamental analysis: Using machine learning to identify mispricing in European stock markets","volume":"48","author":"M X Hanauer","year":"2022","journal-title":"Finance Research Letters"},{"key":"ref37","doi-asserted-by":"crossref","first-page":"993","DOI":"10.1109\/34.58871","article-title":"Neural network ensembles","volume":"12","author":"L Hansen","year":"1990","journal-title":"IEEE Transactions on Pattern Analysis and Machine Intelligence"},{"key":"ref38","doi-asserted-by":"crossref","first-page":"5","DOI":"10.1093\/rfs\/hhv059","article-title":"and the cross-section of expected returns","volume":"29","author":"C R Harvey","year":"2016","journal-title":"The Review of Financial Studies"},{"key":"ref39","author":"O Hellum","year":"2025","journal-title":"How global is predictability? The power of financial transfer learning"},{"key":"ref40","doi-asserted-by":"crossref","first-page":"2019","DOI":"10.1093\/rfs\/hhy131","article-title":"Replicating anomalies","volume":"33","author":"K Hou","year":"2020","journal-title":"The Review of Financial Studies"},{"key":"ref41","doi-asserted-by":"crossref","first-page":"81","DOI":"10.1080\/0015198X.2024.2388024","article-title":"Choices matter when training machine learning models for return prediction","volume":"80","author":"C Howard","year":"2024","journal-title":"Financial Analysts Journal"},{"key":"ref42","author":"W Huang","year":"2026","journal-title":"AI \"errors"},{"key":"ref43","doi-asserted-by":"crossref","first-page":"463","DOI":"10.1111\/j.1475-6803.2006.00189.x","article-title":"Individual equity return data from Thomson Datastream: Handle with care","volume":"29","author":"O S Ince","year":"2006","journal-title":"Journal of Financial Research"},{"key":"ref44","author":"S Ioffe","year":"2015","journal-title":"Batch normalization: Accelerating deep network training by reducing internal covariate shift"},{"key":"ref45","doi-asserted-by":"crossref","first-page":"65","DOI":"10.1111\/j.1540-6261.1993.tb04702.x","article-title":"Returns to buying winners and selling losers: Implications for stock market efficiency","volume":"48","author":"N Jegadeesh","year":"1993","journal-title":"The Journal of Finance"},{"key":"ref46","doi-asserted-by":"crossref","first-page":"3193","DOI":"10.1111\/jofi.13268","article-title":"2023. (re-)imag(in)ing price trends","volume":"78","author":"J Jiang","journal-title":"The Journal of Finance"},{"key":"ref47","author":"S Y Kalfa","year":"2025","journal-title":"Overhyped? Can machine learning models reliably predict stock returns? Working Paper"},{"key":"ref48","doi-asserted-by":"crossref","first-page":"83","DOI":"10.1080\/0015198X.2021.1954377","article-title":"Boosting the equity momentum factor in credit","volume":"77","author":"H Kaufmann","year":"2021","journal-title":"Financial Analysts Journal"},{"key":"ref49","author":"D P Kingma","year":"2017","journal-title":"A method for stochastic optimization"},{"key":"ref50","doi-asserted-by":"crossref","first-page":"816","DOI":"10.1111\/eufm.70033","article-title":"Empirical asset pricing via machine learning: The role of research design choices","volume":"32","author":"V Lalwani","year":"2026","journal-title":"European Financial Management"},{"key":"ref51","doi-asserted-by":"crossref","first-page":"64","DOI":"10.1016\/j.jfineco.2021.08.017","article-title":"Machine learning in the Chinese stock market","volume":"145","author":"M Leippold","year":"2022","journal-title":"Journal of Financial Economics"},{"key":"ref52","first-page":"4768","article-title":"A unified approach to interpreting model predictions","volume":"30","author":"S M Lundberg","year":"2017","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref53","author":"T Masters","year":"1993","journal-title":"Practical neural network recipes in C++"},{"key":"ref54","doi-asserted-by":"crossref","first-page":"5","DOI":"10.1111\/jofi.12365","article-title":"Does academic research destroy stock return predictability?","volume":"71","author":"R D Mclean","year":"2016","journal-title":"The Journal of Finance"},{"key":"ref55","doi-asserted-by":"crossref","first-page":"2339","DOI":"10.1111\/jofi.13337","article-title":"Nonstandard errors","volume":"79","author":"A J Menkveld","year":"2024","journal-title":"The Journal of Finance"},{"key":"ref56","doi-asserted-by":"crossref","first-page":"1249","DOI":"10.1111\/0022-1082.00146","article-title":"Do industries explain momentum?","volume":"54","author":"T J Moskowitz","year":"1999","journal-title":"The Journal of Finance"},{"key":"ref57","doi-asserted-by":"crossref","first-page":"703","DOI":"10.2307\/1913610","article-title":"A simple, positive semi-definite, heteroskedasticity and autocorrelation consistent covariance matrix","volume":"55","author":"W K Newey","year":"1987","journal-title":"Econometrica"},{"key":"ref58","doi-asserted-by":"crossref","first-page":"9","DOI":"10.3905\/jfds.2019.1.3.009","article-title":"Industry return predictability: A machine learning approach","volume":"1","author":"D E Rapach","year":"2019","journal-title":"The Journal of Financial Data Science"},{"key":"ref59","doi-asserted-by":"crossref","first-page":"70","DOI":"10.1080\/0015198X.2019.1596678","article-title":"Machine learning for stock selection","volume":"75","author":"K C Rasekhschaffe","year":"2019","journal-title":"Financial Analysts Journal"},{"key":"ref60","doi-asserted-by":"crossref","first-page":"267","DOI":"10.1111\/0022-1082.95722","article-title":"International momentum strategies","volume":"53","author":"K G Rouwenhorst","year":"1998","journal-title":"The Journal of Finance"},{"key":"ref61","first-page":"918","article-title":"K-sample Anderson-Darling tests","volume":"82","author":"F W Scholz","year":"1987","journal-title":"Journal of the American Statistical Association"},{"key":"ref62","first-page":"307","article-title":"A value for n-person games","volume":"II","author":"L S Shapley","year":"1953","journal-title":"Contributions to the Theory of Games"},{"key":"ref63","doi-asserted-by":"crossref","DOI":"10.1016\/j.jempfin.2024.101517","article-title":"Non-standard errors in asset pricing: Mind your sorts","volume":"78","author":"A Soebhag","year":"2024","journal-title":"Journal of Empirical Finance"},{"key":"ref64","doi-asserted-by":"crossref","first-page":"702","DOI":"10.1177\/1745691616658637","article-title":"Increasing transparency through a multiverse analysis","volume":"11","author":"S Steegen","year":"2016","journal-title":"Perspectives on Psychological Science"},{"key":"ref65","doi-asserted-by":"crossref","first-page":"86","DOI":"10.3905\/jfds.2020.1.034","article-title":"The cross section of commodity returns: A nonparametric approach","volume":"2","author":"C Struck","year":"2020","journal-title":"The Journal of Financial Data Science"},{"key":"ref66","first-page":"11","article-title":"Factor zoo (.zip)","volume":"50","author":"A Swade","year":"2024","journal-title":"Journal of Portfolio Management"},{"key":"ref67","doi-asserted-by":"crossref","first-page":"135","DOI":"10.1016\/S1574-0706(05)01004-9","article-title":"Forecast combinations","volume":"1","author":"A Timmermann","year":"2006","journal-title":"Handbook of Economic Forecasting"},{"key":"ref68","doi-asserted-by":"crossref","DOI":"10.1016\/j.finmar.2020.100588","article-title":"Does it pay to follow anomalies research? Machine learning approach with international evidence","volume":"56","author":"O Tobek","year":"2021","journal-title":"Journal of Financial Markets"},{"key":"ref69","author":"D Walter","year":"2024","journal-title":"Methodological uncertainty in portfolio sorts"},{"key":"ref70","doi-asserted-by":"crossref","first-page":"87","DOI":"10.1257\/jep.15.4.87","article-title":"Applications of generalized method of moments estimation","volume":"15","author":"J M Wooldridge","year":"2001","journal-title":"Journal of Economic Perspectives"}],"container-title":[],"original-title":[],"deposited":{"date-parts":[[2026,10,23]],"date-time":"2026-10-23T19:30:06Z","timestamp":1792783806000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.ssrn.com\/abstract=5031755"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024]]},"references-count":70,"URL":"https:\/\/doi.org\/10.2139\/ssrn.5031755","relation":{},"subject":[],"published":{"date-parts":[[2024]]},"subtype":"preprint"}}