Whether or not currency markets may be regarded as efficient or not has been a hotly debated issue in the academic literature over recent decades. Economic theory would suggest that these markets should be efficient because they are apparently good examples of a perfectly competitive market structure. On the other hand, empirical tests of the efficient market hypothesis within these currency markets unequivocally find them to be inefficient. There is still no good explanation for this conundrum and as a result a fair amount of effort is still expended on refining the empirical studies of market efficiency, a task which is taken up in the four empirical studies that comprise this thesis. Within efficient markets, prices are predicted to respond &quotquickly" with the arrival of new information and the empirical work in the thesis focuses on these issues by identifying three key areas for research, namely, price adjustment and volatility, volatility and the &quotnews", and the speed of price adjustment. In essence, the studies examine whether there is inefficient adjustment to news in terms of excessive volatility, whether or not news is actually the main driver of exchange rate volatility and whether or not &quotquickly" can be measured empirically. The empirical results reported within this thesis confirm that the Australian dollar has not been an excessively volatile currency, even though the level of volatility has been increasing; that the pattern of information flow explains a significant degree of the non constant variance in the returns of the world's most actively traded currencies, (i.e. information explains price innovation); that the reaction time to macroeconomic news occurs within seconds of a pre-scheduled announcement, and that the bulk of adjustment to fundamental value occurs within the hour. These findings are consistent with what would be expected within an efficient market. The results reported within this thesis therefore suggest that the currency markets studied are efficient, at least for the sample periods of the data used in the studies. Exchange rates adjust rapidly with information arrival albeit not completely. It is also the case that a number of additional research questions emerge from this research. For example we know that volatility is not excessive and that it is increasing. What we do not know is the point at which increasing volatility becomes excessive. We know that exchange rates react quickly with the arrival of macroeconomic news, but we do not know precisely how long it takes for volatility to return to preannouncement levels, or why the reaction to news is inconsistent. We also do not know what type of information best explains volatility above that which is explained by the systematic dissemination of information or why full adjustment to fundamental value does not occur? Answers to these questions provide a future research agenda. Answers may provide insight that will help financial economists explain the apparent failure of the speculative efficient hypothesis.

With respect to foreign currency exchange markets, most governments would favour a stable exchange rate over a volatile exchange rate. This is also true for South Africa where volatile movements in the South African Rand pose challenges to industries, businesses and Government alike. There are a multitude of factors that affect the volatility of the South African Rand. These factors are difficult to identify, manage individually and measure. There are several tools that are available to manipulate and control foreign exchange rates in an attempt to reduce volatility; one such tool is the currency transaction tax. Spahn tax is one such form of currency transaction tax. The precursor to Spahn tax is Tobin tax. As a result of many criticisms levelled against Tobin tax, Spahn expanded on this original idea and made a few modifications to address some of the concerns. Spahn focused on creating a two-tier tax base where transactions falling within a normal and reasonable trading range would be taxed at a nominal amount and transactions that fall outside of the band would be taxed at a higher punitive rate. The trading band, or range, would be adjustable though market dynamics on a daily basis using a moving average. No country has implemented Spahn tax yet. The implementation of such a tax would have strong revenue-generating potential. A modification of such a tax with only a punitive rate and a wide trading band could be considered for South Africa. However, in being prudent South Africa does not appear to be in a position to be the first country to implement Spahn tax. There are too many market risks associated with the introduction Spahn tax that cannot be ignored. / Dissertation (MCom)--University of Pretoria, 2012. / Taxation / unrestricted