Abstract:

The city index calculated by the Borsa Istanbul (BIST) is an important indicator of regional development. The city index, which brings light on the financial situation of the area in which it is calculated, has the characteristic of being a very useful guide for investors making investment decisions under uncertainty. Therefore, drawing a map of the financial performance of these cities in recent years has attracted the attention of both investors and researchers. Based on this importance, the study calculated between February 2009 and March 2019 in the 9 city index daily return series volatility, variable variance models with different symmetrical and asymmetrical conditions were predicted. The thick corner structure of the financial time series has been taken into account and the conditional distribution of the error terms has not only been used normal distribution but also Student-t and Generalized Error Distribution (GED). In the selection of the suitable model, the various model selection criteria in addition to the return series are divided into two as training and test, the predictive performance criteria of the models are calculated. One of the remarkable results is that in all index returns series except the Antalya city index, asymmetric models have produced successful results. In addition, in all indicators except the Tekirdağ city index, negative shocks have caused greater volatility compared to positive shocks of the same size.

Keywords: