Author
Abstract
This study looks at how Artificial Intelligence can improve credit scoring for borrowers in Pakistan, especially those who struggle to get loans because they lack formal credit histories. Traditional lending still depends on collateral and old repayment records, which leaves out a large part of the population. With the rise of mobile banking and digital payments, new digital footprints such as telecom usage, mobile-wallet activity, and online transactions can now offer useful signals about a borrower’s behaviour. The research tests whether AI can turn this data into fair and accurate credit decisions. Using a dataset of 10,000 borrowers that reflects real microfinance clients in Islamabad, three models, Logistic Regression, Random Forest, and XG-Boost were developed and compared. The models were trained on both traditional variables (income, loan history, debt ratio) and alternative data (telecom activity, wallet usage, digital purchase patterns). Their performance was checked through accuracy, calibration, and fairness metrics. The results show that AI-based models clearly outperform the traditional approach. XG-Boost achieved the highest accuracy (AUC 0.943), followed by Random Forest, while Logistic Regression showed the weakest performance. Adding alternative data improved prediction accuracy by nearly 18%, making it easier to identify reliable borrowers who would otherwise remain “credit invisible.” Fairness tests also showed that AI models, when properly tuned, reduced gender-based bias and produced more balanced decisions. SHAP analysis confirmed that income, credit history, telecom usage, mobile-wallet activity, and loan size were the strongest predictors of default. The study concludes that AI-powered credit scoring can support financial inclusion in Pakistan by making lending decisions more accurate, transparent, and fair. It recommends that financial institutions adopt explainable AI tools, regulators strengthen data privacy frameworks, and pilot programs be used before full-scale deployment. While the study relies on simulated data, it provides a practical pathway for responsible AI adoption in Pakistan’s credit markets.
Suggested Citation
Muhammad Tofeeq, 2024.
"Evaluating The Impact of Machine Learning and Alternative Data on Credit Precision and Fair Lending in Emerging Markets,"
iRASD Journal of Finance, International Research Alliance for Sustainable Development - iRASD, vol. 3(1), pages 01-07, December.
Handle:
RePEc:rad:irdjof:v:3:y:2024:i:1:id:3018
DOI: 10.52131/jof.2024.0301.0011
Download full text from publisher
Corrections
All material on this site has been provided by the respective publishers and authors. You can help correct errors and omissions. When requesting a correction, please mention this item's handle: RePEc:rad:irdjof:v:3:y:2024:i:1:id:3018. See general information about how to correct material in RePEc.
If you have authored this item and are not yet registered with RePEc, we encourage you to do it here. This allows to link your profile to this item. It also allows you to accept potential citations to this item that we are uncertain about.
We have no bibliographic references for this item. You can help adding them by using this form .
If you know of missing items citing this one, you can help us creating those links by adding the relevant references in the same way as above, for each refering item. If you are a registered author of this item, you may also want to check the "citations" tab in your RePEc Author Service profile, as there may be some citations waiting for confirmation.
For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: Malik Shahzad Ahmad (email available below). General contact details of provider: https://journals.internationalrasd.org/index.php/jof .
Please note that corrections may take a couple of weeks to filter through
the various RePEc services.