FinTech-Enabled Personalized Credit Scoring for Underserved Groups: A Case Study of Alternative Data Application in the U.S. Microfinance Market

Abstract

This paper looks at the issue of financial exclusion present among underserved segments of society like the low income and
non-traditionally creditworthy populations who lack a credit history needed to obtain microloans in the US microfinance market. The potential solutions proposed by FinTech include the usage of nontraditional data sources to generate customized credit score models, allowing this
population group the chance to obtain credits. For this purpose we employ anonymized data from two typical US small microfinance institutions as the experimental data source.
In regards to microfinance organizations (Kiva and Affirm), covering from 2022 to 2023, the paper shows the research about the process of data integration and AI-driven score model and deduces their effects, which indicates that personalized credit scores increase credit
approval rates for underserved groups by more than 30% and decrease default rates more than 15% compared with traditional score methods,
and also indicates that personalized credit scoring has obvious positive impact on default rate.
Creates concerns over the issue of data privacy protection (including CCPA requirements) and model bias. Theoretically, it expands
the related research in inclusive finance and Fintech credit innovation; practically, it offers actionable insights for micro-finance institutionsto
improve the application of alternative data and policy makers to improve the regulation frame-work.