Finansal piyasalarda birlikte hareket üzerine üç makale: Fraktal davranış, bilgi akışı, nedensellik ve tahmin

The behavior of financial time series, their interactions with each other, and analyzes of forecasts especially in crisis and shock periods, have gained an important place in the literature. In the coming years, such analyzes will find a place in many studies. The relationships between financial time series are a very important indicator in analyzing the economic shocks encountered. There are many methodologies and analyses on this subject in the literature. From this point of view, the aim of the thesis is to analyze the co-movement of financial time series, fractal behavior, the measure and direction of information flow, and future price forecasting, and to develop new analysis methods. The fact that all these methods have not been studied together is another starting point of the thesis. In the first part of the thesis, the co-movement of financial time series was analyzed by Wavelet coherence (WTC) method, crisis and shock periods were determined by Multiple wavelet coherence (MWC) method, then fractal behaviors were examined with Multifractal de-trended fluctuation analysis (MFDFA) method and for the post-crisis periods, daily price range estimations for the future were made by using the Vector autoregressive fractionally integrated moving average (VARFIMA) method, in comparison with real data. In the second part, a new method called Wavelet transform guided transfer entropy method (WTGTEM) has been proposed. In this new method, the co-movement of exchange rate time series has been examined with the Wavelet coherence (WTC) method, and the measurement and direction of the information flow between the financial series under the guidance of WTC have been analyzed by the Transfer entropy (TE) method. In the last part of the thesis, the new methodology presented in the second part has been applied to major stock indices. Especially during the COVID period, the behavior of the series has been analyzed, and it would be helpful for investors in portfolio diversification, especially in times of crisis. We believe that this thesis will guide future studies of this kind and that the methods used will be preferred by researchers.