We study discrimination towards Venezuelan migrants in a large commercial bank in Peru. Bank executives (including branch and risk managers) participate in a comprehensive survey experiment, including vignettes and sociodemographic and professional information about executives. Participants are randomly assigned to one condition in each vignette experiment, so we can establish causal relations between the nationality (Peruvian or Venezuelan) and gender (men or women) of clients described in the vignettes and the quality of the service provided. Our results strongly suggests that executives are blind to the discrimination of migrants in Peru, outside their offices. However, the discrimination of migrants in substantial, as the likelihood of accessing basic financial services (opening a bank account, getting a loan application approved) is significantly lower for Venezuelan than for Peruvian clients. Discrimination is largely preference-based and based in a profound mistrust in migrants. Executives see Venezuelan and Peruvian clients as equally skilled and literate financially and equally prone to relevant behavioral drivers (risk tolerance, present bias). We document the existence of intersectional discrimination, as Venezuelan women have a probability of getting a loan approved that is half the probability Peruvian women have (and sixty percent the probability of a male Venezuelan applying to the same loan). Exposure to migrants in their daily professional lives has very limited positive effects. We develop an Intersectional Discrimination Dual Index (IDDI), measuring supply and demand intersectional discrimination drivers, and document how the experiment triggered a large positive change in the bank’s policies.