Survey of the Existence of Long Term Memory in Tehran Stock Exchange, with Emphasis on Affecting Factors by ARFIMA-GARCH Approach

Over the last decade, processes with long term memory have been based on time series analyses. Existence of long term memory in stock price index has important applications in market efficiency, asset pricing and choice of portfolio. Accordingly in this paper, macroeconomic factors affecting stock price index are specified, and then the long-run vector is used to VAR model for the period (1369-1390). The results indicate that the real exchange has a positive effect on liquidity, interest rate and inflation but a negative effect on stock price index. Then it has been reviewed in the content of ARFIMA-GARCH model, to investigate the long memory. Finally, according to the estimated equation (ARFIMA (0, 0.24, 1)), that implies to the long term memory in stock price index, if there are shocks in stock price index some effects of these shocks will remain for a long period.