Econometric modelling of the volatility of ise index and the relationships between ISE and some International Stock Indexes

This study has two aims. The first aim of the study is to model the volatility structure of the ISE index?s returns. In addition to this, the two crisis periods (1999-2001 crisis period and global economic crisis in 2008) which were experienced in Turkey has been taken individually and the volatility of the ISE index?s returns have been examined in these periods. For the main related period and also for the two sub-periods, the volatility of ISE index?s returns have been examined by the conditionally heteroscedastic variance models.The second aim of the study is to find out the relationships between ISE index and some international stock indexes in a certain time period. In this respect various econometric analysis techniques have been used. To design a common scale due to the different scaling of the indexes, the logarithmic forms of indexes have been used. First of all, the stationary of the indexes has been examined by the unit root tests and the results of the tests show that all of the indexes are stationary at a level of first order (I(1)). Then, Engle-Granger two step cointegration test has been used and it has been found two cointegration relationships as ISE-Bovespa and ISE-Merval. The error correction models have been designed for the two models in which the cointegration relationships have been established with the help of Engle-Granger method and it has been identified that how many days will it take to balance the system. Accordingly, it has been determined that in the ISE-Bovespa error correction model, the error correction parameter is negative and statistically significant. On the other hand, the ISE-Merval error correction model has been found insignificant.Moreover, Granger Causality Test is used to determine the relationships between indexes and Vector Autoregressive (VAR) model is used to examine the effects of the indexes to each other.