Market integration and price transmission in a selected area of the South African market

Abstract

This study analyses the performance of a part of the South African market by studying the transmission mechanism of some selected market prices, price spreads and market volatility in the market. A comparative assessment of future forecasts for import and export prices is also conducted using three forecasting models - double exponential smoothing, Winters' multiplicative method and the Box-Jenkins forecasting method. The analyses reveal that a cointegrating relationship exists between producer and consumer prices of goods, producer prices Granger-cause consumer prices but not the reverse, the transmission between producer and consumer prices of goods is asymmetrical, and price transmission is elastic and perfect in the mark-up model but inelastic and imperfect in the mark-down price transmission model. The producer price shock effect is observed with a total consumer price increase after 4 periods (1 year), exchange rates significantly explain import prices (but only a small part of oil prices), and low standard deviations of the cyclical components of the import and export prices are observed suggesting high volatility of the prices. The best forecasting model for both import and export prices is produced by the Box-Jenkins method as ARIMA (1,1,1) model.