Zrazhevska N. Methods and models of dynamic stock risk measures forecasting

The thesis is the development of a systematic approach for obtaining the forecast estimates of risk measures VaR and CVaR, the popular measures in assessing of financial risks, primarily stock market risks. In the framework of the proposed approach, a systematic analysis of current methods for VaR and CVaR evaluating is carried out, the results of the analysis are formulated in the form of classification schemes. To take into account the properties of volatility and long range dependence, which are typical for financial series, a new method of smoothing of the autocorrelation function is proposed. For volatility forecasting FIGARCH model is reduced to the AR model of infinite order for the squares of the process. The reduced system of Yule-Walker equations is solved to find the autoregression coefficients. The regression equation for the autocorrelation function based on the definition of the long-range dependence is used to get the autocorrelation estimates. An optimization procedure is proposed to specify the estimates of autocorrelation coefficients. All stages of the system approach are applied to the time series of the indeсes of the various stock exchanges.