FORECASTING NIGERIA’S INFLATION USING SARIMA MODELING

Abstract

Inflation series exhibit trend or seasonality that makes it difficult to analyze the inflationary pressures for monetary policy decision making. It is imperative to note that few empirical studies have tilted towards addressing the seasonality issues to track the sources responsible for these fluctuations. It is against this background that the study aims at developing a model of inflation with higher data points and taking into cognizance its periodic seasonal component and use the estimated model to make forecast. The study used monthly data sourced from the Central Bank of Nigeria (CBN) Statistical Bulletin. Data was analysed using seasonal ARIMA model (SARIMA) which is an extension of autoregressive (AR) and moving average (MA) process in the popular Box-Jenkins methodology.  With 300 data points, the study developed SARIMA (1,0,0) x (1,1,0)12 from among the competing models based on its AIC and BIC values. The estimated model is found to be adequate in making forecast using a sample data for 2019. The study thus recommends that monetary authorities should consider the seasonal component in designing monetary policies targeted at inflation to stabilize the economy.