EVALUATING THE FORECAST PERFORMANCE OF ARMA-GARCH AND ST-GARCH USING NIGERIAN GROSS DOMESTIC PRODUCTS

Financial data must first be evaluated for forecast performance before being deemed appropriate for use in economic planning, according to policymakers, investors, academics, and end users. Failing to do so will undoubtedly result in a futile endeavor. The forecast accuracy of ARMA-GARCH and ST-GARCH is compared in this paper. Three SmoothTransition GARCH (ST-GARCH) models were examined: Logistic Smooth Transition (LST-GARCH), Exponential SmoothTransition (EST-GARCH), and Exponential Transition GARCH (ET-GARCH). After obtaining the relevant theoretical and mathematical frameworks, the gross domestic product of Nigeria was used as an empirical paradigm. Both the stationarity of the data and the descriptive statistics were looked at in order to assess the state of the series used prior to its analysis.Here, the characteristics of the studied series demonstrate that it was stable at the first difference. It was then examined using specially written programs and the statistical analysis tool Statgraphics Centurion. The results showed that hybrids of LST-GARCH outperformed EST-GARCH, ARMA-GARCH, and ET-GARCH with exceptional performance. Thus, potential users are advised to utilize LST-GARCH.