A comparative study between ARIMA and ARIMAX in Forecasting Gross Domestic Product (GDP) in Nigeria
pdf

Keywords

GDP, ARIMA, ARIMAX, Root Mean Square Error, Mean Square Error

How to Cite

A comparative study between ARIMA and ARIMAX in Forecasting Gross Domestic Product (GDP) in Nigeria. (2025). KASU JOURNAL OF MATHEMATICAL SCIENCE (Maths Access), 2(2), Page: 30-42. https://mathsaccess.org.ng/index.php/kjms/article/view/47

Abstract

The Gross Domestic Product (GDP) is one of the primary indicators used to measure the strength of a country’s economy. The Linear class of time series models have received much attention over the years both in theory and applied research. However, the linear class of models usually do not capture vital features of most macro-economic and financial data. Since the business cycle for most economies is bound to go through either an expansion or recession, it is reasonable to assume that each kind of data set may require a different type of time series model. The data used for this work is an annual data collected for the period 1981-2020 from the Central Bank of Nigeria. The analysis was carried out with the aid of E-Views software. The Akaike Information Criterion (AIC) and Schwarz Information Criterion (SIC) serve as the accuracy criteria in evaluating the performance of the models. The Root Mean Square Error (RMSE) and Mean Square Error (MSE) serve as the error matrices in evaluating the accuracy of the out-sample forecast of the models. Based on the analysis, ARIMAX model is found to perform well with least the Akaike Information Criterion (AIC) and Schwarz Information Criterion (SIC) of 16.64667 and 16.81729 respectively as compared to the ARIMA models with Akaike Information Criterion (AIC) and Schwarz Information Criterion (SIC) of 17.72631 and 17.89869 respectively. Also the ARIMAX model have the best out-sample forecast accuracy with the least Root Mean Square Error (RMSE) and Mean Square Error (MSE) of 899.4937 and 567.9662 respectively as compared to the ARIMA models with  Root Mean Square Error (RMSE) and Mean Square Error (MSE) of 1662.352 and 1160.719 respectively. The results show that ARIMAX model is absolutely suitable for forecasting the data. It is recommended that the policy makers should apply ARIMAX model in forecasting gross domestic product (GDP) in Nigeria. Government should give more attention to Agriculture, Industrial and Services sectors because of their contribution to Nigeria economy especially at this time of economic recovery.

pdf