Abstract:

In developing economies such as Turkey, the estimation of the return of shareholders is among important research issues in financial studies. It is also important for investors to predict the return of the stock. In this study, the aim is to estimate the return of the stock by means of investor sensitivity. In the theoretical part of the study, a detailed literary study was carried out. In the application section, the Models developed for prediction have used Consumer Trust Index and sub-indices that represent investor sensitivity as an independent variable. In this framework, a total of 5 models established with 5 variables have been analyzed using support vector machines methods. All analysed models used monthly data between 2007 and 2018. 70% of the 144 monthly data collected on variables were used as educational data, while 30% of the logs were used to measure the predictive success of mathematical models. In the findings obtained, the forecasts made using the support vector machine method have shown successful results.

Keywords: