We examine the viability of quantitative techniques for investing in individual stocks in the South African market. We measure the information in various firm specific attributes in forming portfolio strategies designed to outperform standard benchmarks. This market is particularly challenging given the extremely short sample by which we can calibrate our methods. Nevertheless, our out-of-sample analysis suggests that we can achieve up to 9% out-performance of standard benchmarks with our buy portfolio. Further, the difference between our recommended stocks and those we suggest avoiding is over 24% per year.
The behaviour of the South African rand is modeled using a monetary model of the exchange rate. The assumptions underlying the monetary approach are discussed in the context of the conditions prevailing in the South African economy. A cointegrating relationship involving the exchange rate, money stocks, industrial production and inflation rates is identified for the 1980M01-1998M11 period. In a later sub-period spanning 1988M01-1997M12, a sticky-price monetary model appears to fit the data better. The model also explains movements in the Rand during a sample encompassing the Russian financial crisis, as long as a dummy variable is included in the regression.
A variety of measures to slow down international capital flows are in place somewhere in the world, and some additional measures have been proposed by academic observers. The debate over the desirability of such measures has been sterile however. It generally consists of arguments for and against the virtues of free and unfettered capital markets. On the one hand, proponents of introducing "sand in the wheels" of international financial markets point to evidence of inefficiencies, anomalies, bubbles, speculative attacks and crashes.