Modeling of Petroleum Prices in Kenya Using Autoregressive Integrated Moving Average and Vector Autoregressive Models

Abstract

The demand for crude oil and petroleum products in Kenya has been increasing very fast over the past twenty years. This is mainly because this particular commodity is used in many sectors of the country’s economy. Ever changing prices affect the exchange rates which also affects industrial production of goods in Kenya. The oil production sector has a crucial impact on the other industries. Any change in the price of petroleum products has a great impact on the prices of other goods produced and even the growth of the economy. This is mainly due to the transport cost involved in transporting the goods. The major aim of this research is to model petroleum products prices in Kenya using Autoregressive Integrated Moving Average(ARIMA) and Vector Autoregression (VAR) models; the models are then compared to determine which of them predicts better the prices in Kenya. Modeling of the prices will greatly guide the government and investors in the energy sector so that they can accurately forecast the future prices. The main sources of data in this research were secondary petroleum pump prices data from the Energy and Petroleum Regulatory Authority (EPRA) of Kenya, exchange rates, inflation rates and the crude oil prices in the world market. Petroleum products prices data from January, 2011 to December, 2018 was used for the modeling process. Comparing several ARIMA candidate models using model criterion, ARIMA (1,1,0) emerged the best model was used to check how international oil prices, the exchange rate of Kenyan shilling against the dollar and inflation rate of the Kenyan shilling affect the petroleum prices in Kenya. Comparison of forecasting ability for the ARIMA and VAR models was done using the mean absolute percentage error (MAPE) mean absolute error (MAE) and the root mean squared error (RMSE). The results showed that VAR was better for forecasting the petroleum prices in Kenya as compared to ARIMA