MODELLING PRICE VOLATILITY OF THREE MAJOR CEREALS IN THE NORTHERN REGION OF GHANA

Cereals are important crops that feed over billions of households worldwide. They have been used extensively for both human consumption and feeding of livestock. In Ghana, cereals such as rice, maize and millet are staple food of great socio-economic importance and they contribute significantly to agriculture Gross Domestic Product (GDP) and the economy of the country. In this study, we developed an ARIMA (p, q)-GARCH (m, s) model to model the volatility of the returns of rice, maize and millet in the Northern region of Ghana. Data on monthly returns of rice, maize and millet from the Ministry of Food and Agriculture were used for the modeling. The results revealed that ARIMA (0, 1)-GARCH (1, 0) was the best model for modeling the volatility of rice returns. Also, ARIMA (0, 1)-GARCH (1, 0) emerged as the best model for modeling the volatility of millet returns. Furthermore, ARIMA (1, 1)-GARCH (1, 0) was the best model for modeling the volatility of maize returns. Diagnostic checks of the three models with the Ljung-Box test and ARCH-LM test revealed that all the models were free from higher-order serial correlation and conditional heteroscedasticity respectively. The dynamic relationship between the returns of the cereals was also investigated using Vector Autoregressive model. VAR (2) and VAR (3) models were fitted to the data. Base on the Likelihood Ratio Test, VAR (3) model was the best for modeling the dynamic relationship between the returns of the cereals. The diagnostic checks revealed that VAR (3) model was adequate. The VAR (3) model was then used to make inference about the relationship between the returns of these cereals. The Granger causality test revealed a bilateral relationship between the returns of rice and that of millet whiles the returns of maize was independent of the returns of rice and millet. The IRF and FEVD analysis both affirm that there exists a dynamic relationship between the returns of the three cereals.