Affinit, an AI-driven financial services firm, said on the 16th that a joint industry-academic research paper written with economists at Northwestern University has been accepted for presentation at the 2026 FMA Annual Meeting, hosted by the Financial Management Association.
The FMA, founded in 1970, is an international finance association that connects researchers, practitioners and students worldwide with the aim of advancing and disseminating knowledge on financial decision-making. Its annual meeting is an academic event where new research in finance is presented and key issues in financial management and related fields are discussed.
The accepted paper, titled "The Distributional Effects of Alternative Data in Credit Scoring," analyzes the effects of applying unstructured alternative data such as text messages to AI credit-scoring models in place of conventional financial history, along with how the benefits are distributed across income groups. The study found that an alternative credit scoring system combining text-message data and other alternative data with machine learning improved predictive power by about 33% compared with traditional credit scores. Text-message data made the largest contribution among the alternative data sources. The system also demonstrated an expansion of financial inclusion by extending lending opportunities to new borrowers who had struggled to obtain loans from established lenders because they lacked credit histories.
The researchers also identified a distributional effect in which the gains from the technology accrue more heavily to groups with relatively abundant financial information. That stems not from algorithmic bias but from differences in the volume of information that can be extracted across groups.
Affinit has operated its own alternative credit scoring system in India for more than a decade, accumulating behavioral data on borrowers with mid- to low-tier credit and on people excluded from the formal financial system. "This FMA presentation shows that the alternative data and AI credit-scoring capabilities Affinit has built up over 10 years in India continue to be recognized by the global finance academic community," said Lee Chul-won, chief executive of Affinit. "We will move faster to build AI financial infrastructure for the financially underserved around the world, beyond India."