Modeling and Forecasting CPI Inflation in Nigeria: Application of Autoregressive Integrated Moving Average Homoskedastic Model

Abstract This study employs a univariate Autoregressive Integrated Moving Average (ARIMA) homoskedastic model in conjunction with Box and Jenkins modeling procedure to model and forecast annual Consumer Price Index (CPI) data in Nigeria from 1950 to 2014. The annual data on Consumer Price Index is obtained as secondary data from Penn World Table, the National Bureau of Statistics and the Central bank of Nigeria over the period 1950 to 2014. We examine the graphical, statistical, unit root and stationarity properties of the series using time plots, ACF, PACF, Phillips-Perron as well as Dickey-Fuller Generalized Least Squares unit root tests. The results show that the CPI data in Nigeria is non-stationary in level but stationary in logged first difference and thus integrated of order one, I(1). We then applied Box-Jenkins modeling methodology to search for an optimal model and found that ARIMA (3, 1, 0) was the best fitting model to describe CPI data series in Nigeria. The model was validated and found to be adequate and good. Based on this model, we forecast the future annual CPI in Nigeria for a period of 6 years from 2015 to 2020. The forecasts show a steady increase in the annual values of CPI in Nigeria. The study predicts that inflation will increase in Nigeria from 2015 since the confidence intervals of the forecast suggest a consistent increase in annual CPI during the forecasted period of 2015 to 2020.

[File Type:PDF]

Updated: July 22, 2016 — 12:18 pm
© Journal of Scientific and Engineering Research Frontier Theme