TOTAL NEW HIV INFECTIONS IN ESWATINI A BOX-JENKINS ARIMA APPROACH

Abstract

Using annual time series data on the total number of new HIV infections in Eswatini from 1990 – 2018, the study makes predictions for the period 2019 – 2030. The study applies the Box-Jenkins ARIMA methodology. The diagnostic ADF tests show that, A, the series under consideration is an I (0) variable. Based on the AIC, the study presents the optimal model, the ARIMA (5, 0, 2) model, also known as the ARMA (5, 2) model. The diagnostic tests further reveal that the presented model is stable and its residuals are not serially correlated. The results of the study indicate that the total number of new HIV infections in Eswatini will continue on an upwards trajectory from 7970 new infections in 2019 to approximately 12350 new infections by 2026, after which new infections might start declining slowly and gradually to almost 11203 new infections by 2030. This projected increase in new HIV infections in the country calls for immediate drastic action to be taken by responsible authorities with regards to HIV prevention and control in the country.