In this research, (ARCH) and (GARCH) were used, which are mathematical models for time series, as the study of time series is considered one of the topics of great importance in analyzing the behavior of various phenomena and interpreting them by analyzing their change according to time, adopting different time periods. Among the most prominent time series are series Finance, which is volatile, that is, characterized by the characteristic of instability (volatility), which makes the use of (ARMA) models inaccurate in its predictive results, which work on the assumption that the variance of the random error, which is distributed normally, and which is constant over time. As for time series, this condition is This is not true and it shows variation and fluctuations at different periods of the series.
 This research aims to choose the best model for the conditional variance of the remainder of the time series model in the process of studying the surplus and deficit in the general financial budget of Iraq on a monthly basis for the period from January 2008 to September 2023. Accordingly, we will build a statistical model using the (ARCH) family and also know the risk element or Uncertainty) by using preference criteria to build time series models and then using the estimated model to predict the variance fluctuations of that model, The study showed that the best forecasting model is GARCH (2,0).