|dc.description.abstract||A considerable share of attention from economists and analysts was focused on the
fluctuations of exchange rates worldwide, including South Africa, over the last few years.
Economists generally believe that there is a long-term equilibrium level to which the
currency will converge, given that the exchange rate fluctuates over the short term. The
aim of this study is to estimate the long term equilibrium exchange rate of the Rand
against the currency of South Africa's main trading partner, the Euro.
The collapse of the Bretton Woods System changed exchange rate determination
significantly. A more volatile international monetary framework followed, both
worldwide and in South Africa. Today, South Africa has a formal inflation targeting
framework and a free floating exchange rate regime in place.
A behavioural approach was followed to determine the equilibrium exchange rate, after
analysing a variety of theories on equilibrium exchange rates that focus on different
aspects of the equilibrium exchange rate. The econometric technique implemented is
direct estimates of a single exchange rate equation. The behavioural, single equation
approach is chosen for its simplicity on the one hand as well as for the fact that current
variables that influence the exchange rate are identified via this approach.
Fundamental variables identified include the real GDP per capita, the real gold price and
gross reserves of the SARB. A vector error-correction mechanism is used in the
estimation of the long-term relationship. Significant values are found up to four lags,
with 1 cointegrating equation at the 1 percent significance level.
The exchange rate is fluctuating around its long-term equilibrium level and is currently
very close to the estimated equilibrium exchange rate. The trend in the data is slightly
negative implying that the equilibrium level has decreased slightly. The closeness of the
exchange rate to the long-term equilibrium is a positive sign for the economy, supporting
the current stable inflation, low unemployment, and growth rate. Unfortunately, the time series of data is very short. The results may therefore be biased as predicted by Maeso-
Fernandez in 2004.||