An Estimation of Fractional Integration Parameter in Time Series of Prices of Basic Metals Group by Using Modern Econometric Techniques in Tehran Stock Exchange

In this study, we examine the existence of long range dependence using an estimation of fractional integration parameter in Tehran Stock Exchange by daily data prices of basic metals group between the 2006 to 2011. To calculate the “d” parameter (the Fractional Integration Parameter) we use analysis of the rescaled range (r/s) and analysis of modified rescaled range (MRS). But before that calculation, short-term memory in prices was removed through the regression of ARMA model and the residuals of these models have been used to test hypotheses and the existence of long range dependence. The results of testing in relation to the prices of basic metals time series show that the parameter of fractional differencing in R/S method and MRS method respectively were equal to 0.34 and 0.11 that confirms the existence of long-memory of the aforementioned index. Thus the hypothesis based on the existence of long range dependence of this variable will also be accepted.