Abstract
The purpose of this paper is to use calibrated univariate GARCH family
models to forecast volatility and value at risk (VaR) of the CARBS indices and a
global minimum variance portfolio (GMVP) constructed using the CARBS equity
indices. the reliability of the different volatility forecasts are tested using the mean
absolute error (MAE) and the mean squared error (MSE). The rolling forecast of
VaR is tested using a back-testing procedure. The results indicate that the use of a
rolling forecast from a GARCH model when estimating VaR for the CARBS indices
and the GMVP is not a reliable method.