IDENTIFYING KEY FACTORS OF MICROLOAN DEFAULTS IN PNG USING PREDICTIVE MODELLING TECHNIQUES

Abstract

One of the key issues faced by financial institutions in regards to their clients, is the problem of loan defaults by their clients, and this greatly affects the sustainability of various financial institutions in PNG. This paper investigates the causes, consequences, and potential solutions to this problem. The study aims to identify the factors that cause the development of such a problem and contribute to it growing. Some of the most common factors involved include age, income, employment status, past repayment records, the reason for taking the loan, the amount borrowed, the terms of the loan agreement, the interest rate on the loan, and the level of education of the borrower. This paper addresses problems that arise as a result of high loan default rates in terms of losses for financial institutions and other lending organizations, lack of access to credit for low-income earners, and issues that are related to financial stability. The methodology implements logistic regression, random forest, and XGBoost models, which are all coded and simulated using RStudio. The research paper suggests practical solutions which include better screening of borrowers through predictive models, better credit monitoring, and financial literacy programs. In PNG, there is a need for all stakeholders involved to come up with a solution that addresses this issue. Other key areas discussed in this paper include credit risk management, which involves data analytics that aim to reduce the chances of loan defaults and improve portfolio performance, ensuring that financial services are available to all our people.