VOLATILITY PREDICTION IN STOCK MARKETS: AN APPLICATION ON BORSA ISTANBUL FINANCIAL AND INDUSTRIAL INDEXES

Abstract

In this study, it is aimed to estimate the volatility of BIST Financial and Industrial Indices by considering the weekly logarithmic returns for the period of 07.01.2007-03.02.2019. In the study, symmetric and asymmetric conditionally varying variance models were used. In this context, firstly the most suitable ARMA model for the series was determined and the volatility structure of the indices was investigated by using the symmetrical model GARCH and the asymmetric model APGARCH model. As a result of the study, the most appropriate estimation model for BIST Financial Index is GARCH (1,1) and the most appropriate estimation model for BIST Industrial Index is APGARCH (1,1). The leverage parameter..1 was positive and significant in the APGARCH (1,1) model for BIST industrial index. This situation shows the existence of leverage effect for negative returns in BIST Industrial Index. Therefore, it can be said that the negative shocks in the BIST Industrial Index will have more impact on the index than the positive shocks of the same size.