An Analysis on Autoregressive Conditional Variance of Real Exchange Rate Index: Modelling with Two-Threshold Arch Method

Time series has an auto regressive conditionally heteroskedasticity state virtually having been ignored in traditional regression models nevertheless if the serie has a volatility, this will be seriously effects the cosistency of the analysis. The ARCH and GARCH methods are used in order to determine the potential volatility in time series. In this study initially the real exchange rate index has been tested with ARCH method after to be achieved its stability and had been marked with ARIMA models. The volatility has been found and it has been modelled by using two thresholded TARCH (1,1) method. The new model reexamined whether it has a volatility or not. After this step there is no voaltility.