133138 20260217205513.0 doi 10.1016/j.eswa.2024.123288 sideral 137820 ART-2024-137820 eng Vilas, Pablo The limited role of sustainability in mutual fund investor decisions: A machine learning approach 2024 Access copy available to the general public Unrestricted Despite the growth in the supply of socially responsible investment products, the weight of environmental, social and governance (ESG) factors in the decisions of mutual fund investors remains under-researched. We conducted a study relating fund flows to past returns, ESG performance and other financial variables using data from 3,767 US mutual funds from 2015 to 2021. First, we aimed to assess the importance of ESG performance in investment decisions. Next, we studied whether ESG performance is increasingly important or has reached its limit. Finally, we developed decisional models to predict the flows raised by each investment fund, its financial return and ESG performance. We used logistic regression, neural networks, random forest and gradient boosting decision trees. We found that the investors consider ESG performance, but the factors that matter most are past growth, mutual fund fees and past returns. Our models predicted the money raised by the funds, obtaining accuracy rates of around 70%. In addition to confirming that “past financial return does not guarantee future financial return,” we found that “past ESG performance guarantees future ESG performance,” which may be of interest to socially responsible investors. info:eu-repo/grantAgreement/ES/AEI/PID2022-136818NB-I00 info:eu-repo/grantAgreement/ES/DGA/S38-23R info:eu-repo/grantAgreement/ES/UZ/JIUZ-2021-SOC-03 info:eu-repo/semantics/openAccess by https://creativecommons.org/licenses/by/4.0/deed.es 7.5 2024 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE 28 / 204 = 0.137 2024 Q1 T1 OPERATIONS RESEARCH & MANAGEMENT SCIENCE 7 / 106 = 0.066 2024 Q1 T1 ENGINEERING, ELECTRICAL & ELECTRONIC 37 / 368 = 0.101 2024 Q1 T1 1.854 2024 Artificial Intelligence 2024 Q1 Engineering (miscellaneous) 2024 Q1 Computer Science Applications 2024 Q1 15.0 2024 info:eu-repo/semantics/article info:eu-repo/semantics/publishedVersion Andreu, Laura Universidad de Zaragoza (orcid)0000-0003-3661-5448 Serrano-Cinca, Carlos Universidad de Zaragoza (orcid)0000-0003-2596-9638 4002 230 Universidad de Zaragoza Dpto. Contabilidad y Finanzas Área Economía Finan. y Contab. 247 (2024), 123288 [16 pp.] Expert syst. appl. Expert Systems with Applications 0957-4174 1401790 http://zaguan.unizar.es/record/133138/files/texto_completo.pdf Versión publicada 2734842 http://zaguan.unizar.es/record/133138/files/texto_completo.jpg?subformat=icon icon Versión publicada oai:zaguan.unizar.es:133138 articulos driver 2026-02-17-20:25:52 ARTICLE