Demand For Electricity In Ethiopia As Related To Consumption Population And Gross Domestic Product

Multivariate time series (MVTS) models are helpful to examine thernrelationships among many macroeconomic variables of a country. Theyrngive us information about the effect of time on some variables and therneffect of one variable on another at the same time. The main objective ofrnthis study is to develop a MVTS model which can be used to describe andrnforecast demand (DMD) and consumption (CON) of electricity, populationrnsize (POP), and gross domestic product (GDP) of Ethiopia simultaneouslyrnbased on forty one years national time series data. Different vectorrnautoregressive (VAR) and vector error correction (VEC) models arernapplied to the data together with unit root test for stationarity andrnJohansen test for cointegration. Unit root test reveals that all the seriesrnare not stationary and the possibility of expressing all the variables usingrntwo cointegration equations is proved by cointegration analysis. The finalrnresult shows that a VEC model of lag one with two cointegrationrnequations best fits the data among various statistical model selectionrncriteria perspective. The accuracy of forecasts using this model is foundrnto be high with mean absolute percentage error (MAPE) less than 5%.rnThe forecasting ability of one variable as compared to another wasrnverified by Granger causality tests. Short-tem and long-termrninterpretations from the VEC model were also given for the differentrneffects among variables under study.

Get Full Work

Report copyright infringement or plagiarism