Hybrid Modeling of Currency Circulation Volatility: Evidence from the Central Bank of Iraq

Abstract

The currency in circulation is a key element of the monetary supply system of the Iraqi economy because itreflects the level of economic activity and the liquidity level in the market. It can be expressed as an important tool when formulating monetary policy. This research aims to analyze and forecast the behavior of the currency in circulation in Iraq using the ARMA-GARCH model for monthly data from 2004 to 2025 to understand the dynamics of monetary liquidity, The sample was divided into two parts: approximately 80% for the training set (2004-2021), and approximately 20% for the testing set (2022-2025). Data were analyzed in Python using many packages. The results showed that the time series was initially non-stationary but became stationary after the first difference. The presence of the ARCH effect, i.e., the unstable variance, justifies the use of the GARCH model to analyze volatility. The ARMA-GARCH model was found to be the most appropriate model for representing the data, achieving the lowest values for the used criteria: RMSE, MAE, BIC, AIC. Furthermore, this study provides a free, open API that allows researchers from around the world to upload their own datasets and obtain instant comprehensive statistical analyses (including tables, diagnostics, and downloadable plots) based on an identical ARMA-GARCH framework. This paper recommends combining the GARCH model with machine learning techniques or Bayesian models, which will increase the accuracy of predictions and the effectiveness of future monetarypolicy choices