AI fintech company Affinit has released research findings showing that incorporating unstructured data such as text messages (SMS) into credit assessment improves prediction accuracy by approximately 33% compared to conventional methods. The study was conducted jointly with economics researchers at Northwestern University and has been accepted for presentation at the 2026 FMA Annual Meeting hosted by the Financial Management Association (FMA).
The FMA is an international finance association founded in 1970. It connects financial researchers, practitioners, and students worldwide with the goal of advancing and disseminating knowledge related to financial decision-making. The FMA Annual Meeting is an annual academic event where new research findings in finance are shared and key issues in financial management are discussed.
The accepted paper is titled “The Distributional Effects of Alternative Data on Credit Assessment.” It analyzes the effects and distributional impacts across borrower segments when unstructured alternative data such as SMS is applied to AI credit scoring models, replacing traditional financial history data.
The research found that the Alternative Credit Scoring System (ACS), which combines alternative data such as SMS with machine learning, improved prediction accuracy by approximately 33% compared to traditional credit scores. SMS data in particular was analyzed as contributing the most among alternative data sources. The system demonstrated its effectiveness in expanding financial inclusion by providing lending opportunities to new borrowers who previously struggled to obtain loans from traditional financial institutions due to a lack of credit history.
The research team also identified a “distributional effect” whereby the benefits of this technological innovation accrue more substantially to segments with relatively abundant financial information. This was explained as stemming not from algorithmic bias, but from differences in the amount of information that can be extracted across different borrower segments.
Affinit has operated its own ACS in the Indian market for over a decade, accumulating behavioral data on low- and mid-credit borrowers and financially underserved populations. This research is being evaluated as a case demonstrating technological credibility by combining this accumulated data with academic methodology.
Lee Cheol-won, CEO of Affinit, said, “This FMA presentation is an achievement demonstrating that Affinit’s alternative data accumulated over 10 years in India and its AI credit assessment capabilities continue to gain recognition from the global financial academic community.” He added, “Following ACM ICAIF and APAD, we are now sharing our research findings at FMA, and we will accelerate our efforts to build AI financial infrastructure for financially underserved populations beyond India on a global scale.”
Building on this achievement, the company plans to accelerate the advancement of its “AI Finance Decisioning Platform,” which spans all areas of financial decision-making, and its push into new Southeast Asian markets.
This research is notable for demonstrating that alternative data can go beyond simply improving credit assessment accuracy to substantively improve access for populations previously excluded from the financial system. At the same time, by precisely identifying the distributional effects arising from information gaps, it also presents policy implications for designing inclusive financial systems.
The news and data on this website are for reference only and do not constitute investment advice or an offer to buy or sell. Information is sourced from exchanges and public sources, and may be delayed, interrupted, or updated. While we strive for accuracy, we do not guarantee timeliness, correctness, or completeness. Content may include external links for which we are not responsible. By using this site, you agree that we and our partners are not liable for any losses. Investment carries full responsibility; please carefully assess risks and consult professionals. If there are errors in the content, please contact us for correction.

source