Corporate Credit Scoring Model of Banking Sector in Indonesia

Using loan level data of monthly financial information services system database from January 2019 to June 2021, this study applies simple logistic regression to develop credit scoring model for corporate bank loans in Indonesia. Variables used in the model consist of borrower characteristic and loan characteristic. We find that days past due and past due are the strongest predictors to determine loan quality. Further evidence shows restructuring policy during COVID-19 period makes the loan qualities stay in collectability 1. Consequently, days to restructuring and frequency of restructuring variables may impact the loan quality looks better than its real condition.