Search

Rapid economic growth propelled China to become South Africa's largest trade partner, and the recent economic slowdown in China and collapse of commodity prices threatens to reverse this trend. Previous estimations of South Africa's bilateral trade functions document strong income effects on trade, but evidence for the effects of the exchange rate and its volatility are decidedly mixed. In this paper we argue that this ambiguity is due in part to the omission of information about the level and volatility of commodity prices. We note that South Africa's exports to China are heavily concentrated in metal ores, and the practice of deflating the nominal exchange rate by either a consumer price index or deflator is a poor proxy for the commodity price movements that matter for exporters. In this paper we use quarterly data from the period 1993-2014 to estimate South Africa's bilateral export function with China using Chinese GDP, the real bilateral exchange rate, exchange rate volatility, iron ore and precious metals price indices, and their respective volatilities. We specify an error-correction model designed to model departures from, and adjustment to, the long-run equilibrium, and thereby distinguish short-run and long-run effects, and we employ the bounds testing approach of Pesaran, Shin and Smith (2001) to test for long-run cointegration among the variables. The baseline model results in large and statistically significant estimates for income and real exchange rate elasticities that are in the expected direction. The elasticity estimate for exchange-rate volatility suggests a negative but statistically insignificant effect. This ambiguity resonates with much of the literature. In our augmented models we include separate variables to capture the effects of iron ore and precious metals prices and their volatility, resulting in considerable improvement in model fit in both a static and dynamic sense. In the long run, the positive price elasticity is consistent with an upward-sloping export supply curve, and the negative and relatively elastic effect of commodity price volatility suggests an uncertainty effect that reduces exports. Importantly, the inclusion of the commodity price variables improves the precision of other estimates, increasing our confidence in a similar uncertainty effect emanating from volatility in the bilateral exchange rate. Our results highlight South Africa's continued vulnerability to the commodity price cycle. To escape this boom and bust cycle, South Africa's government must emphasize public investment and embrace structural reforms aimed at diversifying its export sector.