Modelling and Forecasting Realized Volatility with Semiparametric Diffusion Models

Download or read book Modelling and Forecasting Realized Volatility with Semiparametric Diffusion Models written by Maxim Fedotov and published by . This book was released on 2016 with total page pages. Available in PDF, EPUB and Kindle. Book excerpt: The thesis deals with application of semiparametric diffusion models for Realized Variance process. Theoretical foundations of the work are based on two papers: Ait-Sahalia 1995 who introduced semiparametric modeling for interest rate processes and its applications for derivatives pricing and Koo and Linton 2012 who relaxed assumption of stationarity of the process and introduced more flexibility into the model and its estimation. The thesis has two main goals - first, try to apply semiparametric diffusion models on Realized Variance time series, emerged recently thanks to high- frequency trading and gaining attention from many researchers today. This includes estimation of the model and behavior of the estimates with time and state. Second, in this thesis we are going beyond purely analytical applications of estimated model by trying to build simulations framework around it and in particular study performance of risk forcast in this framework. This, infact, is the first attempt to use this framework for forecasting Realized Variance series, in general, and by utilizing simulations, in particular. It is established in the thesis that the estimation of the process parameters gives results in line with expectations and these results indeed appear to capture true observed dynamics of the process. However, though some of the results look promising, in the course of numerical part of the thesis, a number of complications revealed themselves, which in this work were attempted to be addressed in a simplified fashion, and, undoubtedly, form an interesting direction for further research.