The economic crises in Iraq’s economy have led to the deterioration of the quality of assets of banks operating in Iraq over the years, so that banks operating in Iraq suffer from a large number of non-performing loans; Due to the succession of these crises, perhaps the most recent of which is the COVID-19 crisis and the discontinuation of most economic activities, which caused many borrowers to delay repayment of their obligations and consequently deteriorate the quality of assets and this has significantly eroded the capital of many banks operating in Iraq, A dramatic rise in non-performing loan assets has weakened banks’ performance in Iraq in particular and economic growth in general. Therefore, the importance of conducting asset quality forecasting for banks operating in Iraq through the use of ARIMA models to analyze their performance and predict their quality as this quality is linked to financial stability and the ability to deliver banking services effectively, the need to enhance public and depository confidence in the financial system, as well as to enhance transparency and control of banks’ financial risks, By understanding the factors affecting asset quality and using appropriate statistical models to predict them banking “, banks can provide better banking services and achieve greater stability and success in today’s business environment. So the paper aims to select the best model among time series prediction models (such as the auto regressive integrated moving average model, ARIMA, and other models) by employing measures of predictive ability (such as the Akaike information criterion, AIC, the Schwarz Bayesian information criterion, SBIC, and the Hannan-Quinn criterion, HQC) to compare these different methods in predicting the monthly quality of assets for banks operating in Iraq until 2030. The results show that the best model for monthly prediction of non-performing loans to total capital ratio for operating banks in Iraq over the next (108) months is the ARIMA (1,0,6) model. Additionally, The best model for monthly prediction of non-performing loans to total loans ratio for operating banks in Iraq over the next (108) months is the ARIMA (1,1,1) model. The main recommendation is to use the appropriate quantitative ARIMA model for monthly asset quality prediction and to utilize statistical measures of predictive ability for continuous monitoring of the model’s performance.