TIME SERIES ANALYSIS AND QUARTERLY FORECAST OF NIGERIA’S EXTERNAL RESERVE USING BOX-JENKINS ARIMA MODEL BETWEEN THE YEARS 1981 – 2012

This research work involves the use of Box-Jenkins time series ARMA model to forecast quarterly external reserve of Nigeria. Forecasting is a necessary tool for any business establishment or organization for the purpose of budgeting of time and resources whether human or financial. It provides a guide for businesses in order to achieve set goals and objectives. The data span for the study is from 1981 to 2012 largely obtained from CBN statistical bulletins (1980-2012). Box-Jenkins ARMA model was used for model fit and forecasting. Necessary conditions of using the method such as ensuring stationary were met before application. The data used was found to be stationary at first difference and after several trials, the process was found to follow an autoregressive integrated moving average model: ARIMA (2, 1, 2) using the minimum Bayesian Information Criterion (BIC) values. The parameters of the model were estimated and short term forecast were obtained with the use of Eviews software and result were used to make reasonable conclusions.