Value at risk forecasting of gold price: a comparison between the GARCH and LST-GARCH models

Value at risk is one of the most important measure in finance. This paper evaluates the value at risk forecasting performance of the GARCH and logistic smooth transition GARCH (LST-GARCH) models for the gold markets. The LST-GARCH model is capable to react differently to positive and negative shocks in financial time series. The results show that the LST-GARCH structure provides the more adequate value at risk forecasts relative to the GARCH model.