Who Benefits from Alternative Data for Credit Scoring? Evidence from Peru

Abstract

The World Bank estimates that 1.4 billion individuals worldwide remain unbanked, lacking access to traditional credit due to the absence of credit scores. In this paper, we demonstrate how retail transaction data can be used to construct an alternative credit score, potentially expanding credit access for these individuals. Our study utilizes a unique dataset obtained through a partnership with a Peruvian conglomerate. We merge customer loyalty data and credit card repayment data with administrative records from the Peruvian financial system, which provide detailed financial histories of individuals. This comprehensive dataset allows us to construct credit scores for individuals both with and without a credit history. Through simulations of credit card approval decisions, we find that incorporating retail data increases approval rates for individuals without a credit history, from 15% to between 31% and 47%. In contrast, for those with an established credit history, approval rates remain largely unchanged at around 87%. We investigate why retail data particularly benefits individuals without a credit history and discuss the broader implications of this credit scoring methodology for consumers, firms, and policymakers. Our findings highlight its potential to transform credit access for millions of previously unbanked individuals.

Keywords: Alternative data, Financial inclusion, Consumer finance, Credit cards market. JEL Codes: D19

JEL Classification: D19, G21, G23, G51, I39, O16