Modelling the Volatility of Maize Prices Using Autoregressive Integrated Moving Average Model

Abstract

This paper examined the monthly prices of maize using Autoregressive Integrated Moving Average (ARIMA) models so as to determine the most efficient and adequate model for analyzing the volatility of the maize prices in Kenya. An exploratory research design and purposive sampling method was adopted for a sample of 55 observations. The monthly maize price data for 90kg bag of maize for a period of five years obtained from Kenya National Bureau of Statistics and National Cereals and Produce Board archives. Time series Analysis was done using R-Gui software.  The results indicate that Autoregressive Integrated Moving Average models ARIMA (1,2,2) is the most adequate and efficient model. This was ascertained by comparing the various model selection criterion and the diagnostic tests for various models among them Akaike Information Criteria A better understanding of a country's maize price situation and future prices will facilitate users to make appropriate decisions regarding buying and selling patterns hence adequate policy for maintain stable maize prices. The forecasted results suggest that there are expectations of increasing maize prices in the next five months. This requires the government to take appropriate measures to ensure that this trend of increasing prices is regulated

Key words: ARIMA Model, Volatility, Akaike Information Criteria, Maize Prices