| Resumo: | In this study we examine the long memory behaviour of stock market volatility of the PIIGS major indices: PSI 20, FTSE MIB, ISEQ, FTSE/ATHEX and IBEX 35. In order to conduct our analyses we apply two FIGARCH-type models, one derived by Baillie, Bollerslev and Mikkelsen (1996) and another one developed by Chung’s (1999). In addition the Local Whittle estimator is also computed. A data set comprising the daily closing prices of the PIIGS’ major stock market indices spanning from 1st January 1998 to 8th March 2013 is used. The results suggest that, irrespective of the FIGARCH model adopted, there is evidence of long memory in stock market volatility. However, the Local Whittle Estimator reveals that the data generating process is a combination of long memory and jumps/structural breaks. Therefore, this feature of the data has to be taken into account when constructing models for volatility prediction.
|
|---|