BalanceHero to Present ‘AI-Based Alternative Credit Scoring Model Research’ at International Academic Conference

32% Higher Performance Than Traditional Credit Scoring Models
Effective Use of SMS Text Data

BalanceHero announced on the 12th that its large-scale data-driven credit evaluation model research in India has been selected for presentation at the international academic conference KER (Korean Economic Review International Conference). KER, hosted annually by the Korean Economic Association (KEA), is an international conference where domestic and international economic research achievements are shared. This year’s KER, scheduled for the 17th, will feature discussions on the latest economic theories and empirical studies, with BalanceHero’s research being presented in a session themed ‘Inequality and Markets.’

BalanceHero app

This study empirically analyzed how effectively an artificial intelligence (AI)-based Alternative Credit Scoring System (ACS) predicts credit risk compared to traditional methods, specifically targeting individuals for whom conventional credit evaluation is impossible or unreliable. Under the theme of ‘Financial Accessibility for Low-Income and Credit-Invisible Populations,’ researchers from Northwestern University conducted the study using financial history data accumulated from tens of millions of BalanceHero customers. BalanceHero collects and analyzes a wide range of non-traditional data?such as users’ SMS payment records, installed apps, GPS information, and survey responses, which are not available to traditional financial institutions?and transforms them into financial data.

The results showed that BalanceHero’s ACS outperformed traditional credit scoring models by 32.7% in predictive accuracy. Additionally, financial transaction records and repayment information contained in SMS messages played a significant role in assessing credit risk. Shin Jaehyuk, a leader at BalanceHero, stated, “This research has provided new opportunities for the financially illiterate and those without credit information in developing countries like India. It demonstrates that alternative data and AI technology can greatly contribute to global financial inclusion.” He added, “We will continue to seek a balance between personal data protection and resolving information asymmetry, aiming to develop services that contribute to the global financial market.”

This content was produced with the assistance of AI translation services.