Abstract
In conventional mortgage lending process, bank manager decides whether to approve the mortgage applications based on the credit risk of applicants. However, the evaluation result may be distorted due to operational error such as lack of experience, manager’s subjective decision, or business competition. Therefore, it is essential to build a criterion to help bank manager to make right decisions. Credit scoring comprises a set of statistical models that quantify credit risks of loan applicants. It is also a process of assigning a single quantitative measure, the score, to represent the borrower’s probable future loan performance. Once credit scoring model has been set up, bank manager can deal with large volumes of lending decisions in an efficient, consistent, and controlled manner. The intent of this article is to develop a numerical credit evaluation model to quantify the applications’ financial status, and personal characteristic. With the help of the credit scoring model, bank manager can distinguish between “good” applicants and “bad” applicants in an efficient, stable, and accurate manner.