Bank Loan Default Risk Model

Background

Since the U.S. Small Business Association (SBA) loans only guarantee a portion of the entire loan balance, banks will incur some losses if a small business defaults on its SBA-guaranteed loan. Therefore, banks are still faced with a difficult choice as to whether they should grant such a loan because of the probability of default. The dataset consists of 899,164 rows and 27 columns of historical information spanning multiple decades. The primary objective is to assist stakeholders in making informed decisions about loan approvals by maximizing profits and minimizing losses through data-driven insights. By analyzing loan repayment probabilities, the project seeks to identify the most effective classification methods for predicting loan outcomes, incorporating key profitability metrics and risk considerations. A Logistic Regression model was used and evaluated based on performance metrics such as accuracy, precision, recall, and F1 score. However, profit metrics will also be considered to ensure that the models contribute to maximizing profitability.