We are indebted to the Editor Stefan Nagel, an anonymous AE, and two anonymous referees. We thank seminar and conference participants at the University of Houston, AEA Meetings, Dartmouth HF Seminar, IPA Financial Inclusion Gathering, SITE, CFPB Research Conference, FIRS, MFA, CEPR HF seminar, Australian National University (ANU), Shanghai Advanced Institute of Finance, SJTU (SAIF), and Lone Star Conference, as well as Taha Choukhmane (discussant), James Choi (discussant), John Beshears, Doug Bernheim, David Laibson, Antonio Gargano, John Gathergood, Zack Liu, Anton Lines, Catherine Evans, Nhan Le, Erkki Vihriälä (discussant) and Han Xia (discussant) for providing valuable comments. We also thank Jose Antonio Murillo for invaluable support and access to the data, Isidoro Garcia for helping with preliminary analysis, and Sofia Terrazas for outstanding research assistance. Portions of this research were conducted with computing resources provided by the Research Computing Data Core at the University of Houston and Texas A&M High Performance Research Computing Cluster. The experiment was pre-registered under RCT number AEARCTR-0003047. IRB ID:IRB2018-0597. Authors do not have direct conflicts of interest. At the time of the experiment, Medina was a full-time employee of Banorte; however, the bank does not reserve the contractual right to review or withhold any of the research results. The views expressed herein are those of the authors and do not necessarily reflect the views of the National Bureau of Economic Research.