CREDIT RISK SCORING MODEL FOR CONSUMER FINANCING USING LOGISTIC REGRESSION METHOD: CASE STUDY OF PT BRI MULTIFINANCE INDONESIA

• Main Author: Yuli Ismawati, Isti
• Format: Theses
• Language: Indonesia
• Subjects: Manajemen umum
• Online Access: https://digilib.itb.ac.id/gdl/view/62809
• Institution: Institut Teknologi Bandung

The credit risk scoring model is an important tool for evaluating the credit risk associated with the customers. The development of this model is based on variables that affect the default. This research was conducted on a finance company which is a subsidiary of one of the commercial banks in Indonesia, to answer the challenge of determining the feasibility of financing quickly and accurately. Previously this finance company faced credit risk where the increase in financing was accompanied by an increase in non-performing financing (overdue) and the financing evaluation process was carried out by manual calculations with a relatively long processing time. This model was constructed using the logistic regression method based on historical customer data in the form of demographic characteristics, assets, employment, and financing payments indicators. There are 21 independent variables in the study and one dependent variable (default). The results of this research identify nine variables that can be used as predictors of the probability of default: job type, work experience, net finance value, financing period, car brand, asset price, percentage of down payment (DP), interest, and income. This research ends with the establishment of a risk-scoring model with the identification of variables that have a significant effect on the model. The model was tested by a validation test using the ROC and AUC curves, with the AUC value for the model with 15 variables 0.691 and the model with 9 variables being 0.688.