Econometric Modeling for Consumer Price Inflation in East Africa Community: interdependency assessment and prediction guideline

Abstract

The negative impact of high inflation on the economic health of country has pushed different countries to come up with economic strategies that might keep inflation rate at a low level. However, the positive trend in inflation rate continues to defy solution in Sub-Saharan Africa in recent years; leaving Sub-Saharan countries the option of building up strong predictive skills as they deal with the challenges of economic manipulations of indicators influencing inflation. In East Africa Community, where there is a suggestion of using one currency, inflation might be a big threat, accented with South Sudan a new member of East Africa which is among countries with the highest inflation rate in the world, as of May, 2016. In this research, we assess the interdependency of consumer price inflation rate of elderly members of East African Community, to look at the credibility of one currency in the area in terms of consumer price inflation. We also make an econometric modeling analysis that provides a guideline in consumer price inflation prediction of East Africa Community Member states. The prediction performance of Ordinal Least Square multiple linear model, Support Vector Machine, K Nearest Neighbor, Penalized Linear Models and Autoregressive Integrated Moving Average have been compared to assess the more reliable model in the area. The results of the study show that there is a strong interdependence of consumer price inflation rate in old members of East Africa Community, and there is no difference in the levels of inflation in these countries, giving an edge for one currency suggestion in the area. For the prediction side, Support Vector Machine turned out to be the best model outperforming all other models on accurate prediction in the whole area