Evaluation of Long Memory in the Volatility of Tehran Stock Exchange

According to the growing importance of financial markets, existence of any volatility in the market has dramatic effects on the economy. Therefore, in a dynamic financial markets including stock market, forecasting has improved to one of the most important issues in financial sciences. In this regard, this paper evaluated existence of long memory in returns and volatility of the market price index by using daily prices and returns Tehran Stock Exchange data from 2009/04/04 to 2013/10/22. The results of this study confirm the existence of long memory in both the mean and variance equations. After confirming the existence of long memory in this index, we used GARCH-type models (including non-fractal and fractal models) to track its volatility. Conclusively, comparing the information criteria (Akaike and Schwartz) of various conditional variance heteroskedasticity models, we found the performance of the ARFIMA (1,2)-FIGARCH (BBM) model is desirable.