Multivariate Time Series Analysis on the Prices of Staple Foodstuffs in Kwara State, Nigeria
No Thumbnail Available
Date
2016
Journal Title
Journal ISSN
Volume Title
Publisher
Department of Science Education, Federal University of Technology Minna.
Abstract
Due to supplementary, complementary and substitute relationship between staple foodstuffs, the prices of
one or more staple foodstuffs tend to influence and could be used to predict the prices of some others.
This study was therefore aimed at establishing the co-movement between the prices of some major stable
foodstuffs - Rice, Maize, Garri, Millet, Guinea-Corn and Beans - in Kwara State, Nigeria. Multivariate
time series models were fitted to data on monthly prices of Rice, Maize, Garri, Millet, Guinea-Corn and
Beans over a period of twelve years (from January 2000 to December 2012). The cointegration relations
among the prices were established by applying Johansen’s cointegration tests. As a result, appropriate
Vector Error Correction (VEC) model was fitted to the data. The unit root test for stationarity in the series
reveals that all the series were non-stationary but they were only made to be stationary at first difference.
The results from the analysis showed that there exist short term adjustments and long-term dynamics
among the prices of Rice, Maize, Garri, Millet, Guinea-Corn and Beans in Nigeria over the study period.
Further results showed that a Vector Error Correction (VEC) model of lag two with one cointegration
equation best fits the data. The forecasting accuracy of the fitted model was determined by out-of-sample
forecasts of the future prices of the selected staple foodstuffs. Suitable model’s assessment criteria such as
root mean square error, mean absolute error and the like were employed to determine the efficiency of the
fitted model. The data employed for the study were collected from the Kwara State office of National
Bureau of Statistics, Nigeria. All analyses were performed in the environment of R Statistical package.
Description
Keywords
Vector autoregressive, co-integration, Vector Error correction model and forecasting, Prices