|
|
Quantitative Methods of Risk Control
Marcinek, Daniel ; Hurt, Jan (advisor) ; Hendrych, Radek (referee)
This thesis deals with stock modelling using ARCH and GARCH time series. Important aspect of stock modelling is to capture volatility correctly. Volatility in finance is usually defined as a standard deviation of asset returns. Many different models, which are summarized in the first part of this thesis, are used to model volatility. This thesis focus on multivariate volatility models including multivariate GARCH models. An approach to constructing a conditional maximum likelihood estimate to these methods is given. Discussed theory is applied on real financial data. In numeric application there is a construction of a volatility estimates for two specific stocks using models described in the first part of this thesis. Using the same financial data various bivariate models are compared. Based on comparison using maximum likelihood a specific model for these stocks is recommended. Powered by TCPDF (www.tcpdf.org)
|
|
|
Quantitative Methods of Risk Control
Marcinek, Daniel ; Hurt, Jan (advisor) ; Hendrych, Radek (referee)
This thesis deals with stock modelling using ARCH and GARCH time series. Important aspect of stock modelling is to capture volatility correctly. Volatility in finance is usually defined as a standard deviation of asset returns. Many different models, which are summarized in the first part of this thesis, are used to model volatility. This thesis focus on multivariate volatility models including multivariate GARCH models. An approach to constructing a conditional maximum likelihood estimate to these methods is given. Discussed theory is applied on real financial data. In numeric application there is a construction of a volatility estimates for two specific stocks using models described in the first part of this thesis. Using the same financial data various bivariate models are compared. Based on comparison using maximum likelihood a specific model for these stocks is recommended. Powered by TCPDF (www.tcpdf.org)
|
|
|
Data Envelopment Analysis with financial application
Marcinek, Daniel ; Branda, Martin (advisor) ; Zichová, Jitka (referee)
This thesis deals with various methods of Data Envelopment Analysis and their use in finance. Efficiency is measured by a ratio of weighted outputs to weighted inputs. From this model, a fractional programming problem is formed, which is then transformed into a linear programming problem. We derive a dual problem for that one. We also introduce another methods of Data Envelopment Analysis. We explain difference between a constant return to scale and a variable return to scale. We deal with a risk measures, which are considered as the inputs together with the management fees. We use gross returns as the single input. We apply these models to 15 mutual funds, determine efficiency of these mutual funds and compare these methods with another one. At the end we determine how the efficiency changes if we use only the risk measures as the inputs.
|
guest ::
