Time series analysis of the consumer price index in Ghana (1964-2009)

Abstract

The dissertation was on Time Series Analysis of the Consumer Price Index (CPI) in Ghana for the period 1964 to 2009. The main objective of the study was to determine the general trend pattern and to develop a model for forecasting the Consumer Price Index (CPI) for the old and new series. In order to achieve these objectives, CPI figures for the period were obtained from the prices section of Ghana Statistical Service.The major statistical software used to analyze the data was the MINITAB and SPSS.The study revealed for the old series that the index has increase to 72488.38 percentage points from 1965 to 1997. Also the highest inflation rate of 121.35 percent was recorded in the year 1983 and the lowest inflation rate of - 87.09 percent occurred in the year 1978. For the new series it revealed that the index has increase to 165.15 percentage points from 1998 to 2009 and the highest inflation rate of 33.59 percent was recorded in the year 2001 while the lowest inflation rate of 10.73 percent occurred in the year 2007. The yearly CPI shows an increasing trend for both the old and new series.The trend analysis for the old series shows that the average change of CPI from one year to the other is 43.89 percent and that for the new series was 139.882 percent. Four main smoothing techniques were used for forecasting the yearly CPI figures and from the empirical evidence gathered the Holt’s exponential smoothing is most significant for the old series and AR(1) of the ARIMA model was the most appropriate technique for the forecast of the new series.