Forecasting of Bad and Restructured Loans in Bulgarian Bank System through ARIMA Models
Forecasting of Bad and Restructured Loans in Bulgarian Bank System through ARIMA Models

Summary/Abstract: The monitoring of bad and overdue credits in Bulgaria is an important duty of the National Bank. Rapid increase of these types of credits is a potential menace to the stability of the bank system and thus forecasting of their future values is of great importance. In this work we apply classical linear time-series models and Box-Jenkins approach to bad and overdue credits monthly data. Starting with relatively wide set of initial “candidates” we find appropriate model. Three months ahead forecasts are calculated along with 95 % confidence intervals for the forecasts. The obtained results were generally confirmed by the actual values observed.

• Details
• Contents