CREDIT MANAGEMENT ANALYSIS TO PREDICT EARLY NON-PERFORMING LOANS AT PT. BANK NEGARA INDONESIA (PERSERO) TBK

Hardiansyah
Nazaruddin
Muhammad Anggia Muchtar

International Journal of Educational Review, Law And Social Sciences (IJERLAS)
Vol. 5 No. 4 (2025): July

The purpose of this study is to improve credit quality in designing a customer loan default prediction system so that it can reduce the possibility of huge losses in banking in order to produce a model that has high accuracy and recall rate at PT. Bank Negara Indonesia (Persero) Tbk. The entire population in this study was sampled as many as 467 customers. Using the random forest method, the results obtained are: 1) descriptive analysis shows that the variables of customer credit history, payment ratio, age, collateral value, and affiliate balances influence the occurrence of bad debt, 2) the prediction model identifies patterns of customers at risk of bad debt, 3) the application of the credit model appropriately can reduce the level of bad debt.