Modelling consumer price index inflation in Ghana

Abstract

This paper aims to develop a model for CPI inflation of Ghana using data on CPI inflation obtain from the Bank of Ghana database within the period 2000 to 2011. Based on the method of maximum likelihood and the lower lags of the ACF and PACF, ARIMA (0, 1, 1) X (0, 1, 1)12 was identified as the tentative model for the CPI inflation data in Ghana. The diagnostic check revealed that the residuals of the fitted model have zero mean, constant variance, and free from higher-order serial correlation. The Ljung-Box statistics and the time series plot of the model residuals clearly indicated no significant departure from white noise. The model was found to be free from conditional heteroscedasticity following the ARCH-LM test