COMPSTAT 2018: Start Registration
View Submission - COMPSTAT2018
A0432
Title: Conditional skewness and kurtosis in the good and bad times Authors:  Lukas Fryd - University of Economics, Prague (Czech Republic) [presenting]
Abstract: The main contribution is to the broader understanding of risk premium in the PX50 index before, during and after the crisis in 2008. The most common framework in financial economics take the variance as the main source of risk, and so most of the papers contain the models from the GARCH family. The GARCH helps to reduce the leptokurtosis of standardized returns but not eliminate it. We could find additional information in the high--order moments. If we assume that higher moments represents the additional risk premium, then we could get a better understanding of financial markets. We will assume the data generation process of returns follows the skewed generalized\textit{t} distribution with time-varying parameters. The GARCH methodology will be utilized to the modeling of high-order moments as autoregressive processes.