As part of the credit risk management process of financial institutions, the non-performing loans (NPLs) ratio remains one of the essential components that distinguishes the well-managed assets of a bank. In this paper, we aim to empirically forecast the level of non-performing loans (NPL) including afflicted periods like the COVID-19 ­pandemic using a seasonal ARIMA model. Our analysis is based on the NPLs level observed in the Albanian banking system between December 2015 and December 2022. The results indicate that the seasonal ARIMA (0,1,1)x(2,2,2)12 is the appropriate model that can be applied to predict the monthly level of NPLs. The results also reveal that the expected average monthly ratio of NPLs remains stable, with a slight decrease until the end of 2023. Efforts to be proactive rather than reacting post-factum involve using mechanisms and forecasting models to define non-performing loan ratios and better manage them. This paper considers significant implications in credit risk management in terms of developing actions to manage the magnitude of non-performing loans throughout the COVID-19 pandemic.­ Read more
Keywords:
COVID-19, forecasting, SARIMA, non-performing loans

JEL:
E37, G21, C23, C53