|
|
Value-at-risk forecasting with the ARMA-GARCH family of models during the recent financial crisis
Jánský, Ivo ; Rippel, Milan (advisor) ; Seidler, Jakub (referee)
The thesis evaluates several hundred one-day-ahead VaR forecasting models in the time period between the years 2004 and 2009 on data from six world stock indices - DJI, GSPC, IXIC, FTSE, GDAXI and N225. The models model mean using the AR and MA processes with up to two lags and variance with one of GARCH, EGARCH or TARCH processes with up to two lags. The models are estimated on the data from the in-sample period and their forecasting ac- curacy is evaluated on the out-of-sample data, which are more volatile. The main aim of the thesis is to test whether a model estimated on data with lower volatility can be used in periods with higher volatility. The evaluation is based on the conditional coverage test and is performed on each stock index sepa- rately. Unlike other works in this eld of study, the thesis does not assume the log-returns to be normally distributed and does not explicitly select a partic- ular conditional volatility process. Moreover, the thesis takes advantage of a less known conditional coverage framework for the measurement of forecasting accuracy.
|
|
|
Three Essays on Central European Foreign Exchange Markets
Moravcová, Michala ; Horváth, Roman (advisor) ; Komárek, Luboš (referee) ; Baumohl, Eduard (referee) ; Pappas, Vasileios (referee)
This dissertation thesis consists of three essays on new EU foreign exchange markets (FX), i.e. the Czech koruna, Polish zloty and Hungarian forint. In the first two essays, the impact of foreign macroeconomic news announcements and central banks' monetary policy settings on the value and volatility of examined exchange rates is analyzed. In the third chapter, the conditional comovements and volatility spillovers on new EU FX markets is examined. The aim of this thesis is to contribute to the existing empirical literature by providing new evidence of the examined currencies during periods, which have not been examined yet (after the Global financial crisis (GFC), during the EU debt crisis and during currency interventions in the Czech Republic). The first essay (Chapter 2) examines the impact of Eurozone/Germany and US macroeconomic news announcements and monetary policy settings of the ECB and the Fed on the value of new EU member states' currencies. It is a complex analysis of 1-minute intraday dataset performed by event study methodology (ESM). We observe different reactions of exchange rates in pair with the US dollar on the US macroeconomic announcements and Euro-expressed FX rates on Germany macro news during the EU debt crisis and after it. We also provide evidence of leaking news, showing...
|
|
|
Parametric estimation of GARCH model using MATLAB
Dúbravský, Martin ; Tran, Van Quang (advisor) ; Fučík, Vojtěch (referee)
Timely invariant variance is known not to be stylized fact of financial returns data. Motive of this bachelor thesis is to study financial data's typical variability of variance. In theoretical part, assumtions of GARCH models and its extensions, are summarized. GARCH family models' parameters are estimated, using maximum likelihood are estimated in empirical part. These models are estimated and evaluated across five assets, in which stock indicies DAX and SAP 500, FX major EURUSD and commodities natural gas and gold, are represented. In order to make assumptions about robabilistic distribution of data more realistic, not only Gaussian distribution, but also more leptokurtic Student's t-distribution is assumed to be present in data. Estimations are executed using software package MATLAB and EViews environment. For each asset, the best one of estimated GARCH models will be selected. Results suggest, that assumption of leptokurtic distribution generates models that describe volatility of studied assets better. Regarding testing for assymetric effects in volatility and estimation of EGARCH models, leverage effect of financial returns is shown to be present in returns of studied assets.
|
|
|
The Impact of Macroeconomic News on the Price of Financial Assets
Říha, Jakub ; Moravcová, Michala (advisor) ; Džmuráňová, Hana (referee)
This thesis investigates the effect of Czech macroeconomic news announcements and Czech National Bank (CNB) communication on the price of financial assets and its volatility. As the financial assets we selected the EUR/CZK and USD/CZK exchange rates and also the Prague stock PX Index. To analyze the aforesaid effect we employed the GARCH (1,1) and EGARCH (1,1) models, each with Normal and Student's t error distribution. The main results were that the CNB's communication indeed have significant effect on the price of all three examined assets and surprisingly also tend to increase their volatility. Also the macroeconomic announcements significantly influence examined assets however significant macroeconomic indicators differ for each asset. The most influencing ones are: CPI, 1YPRIBOR and the unemployment rate. Another finding of our research was that volatility of examined time series data shows the characteristics of leverage effect, volatility clustering and persistence. Powered by TCPDF (www.tcpdf.org)
|
|
|
Value at Risk: GARCH vs. Stochatistic Volatility Models: Empirical Study
Tesárová, Viktória ; Gapko, Petr (advisor) ; Seidler, Jakub (referee)
The thesis compares GARCH volatility models and Stochastic Volatility (SV) models with Student's t distributed errors and its empirical forecasting per- formance of Value at Risk on five stock price indices: S&P, NASDAQ Com- posite, CAC, DAX and FTSE. It introduces in details the problem of SV models Maximum Likelihood examinations and suggests the newly devel- oped approach of Efficient Importance Sampling (EIS). EIS is a procedure that provides an accurate Monte Carlo evaluation of likelihood function which depends upon high-dimensional numerical integrals. Comparison analysis is divided into in-sample and out-of-sample forecast- ing performance and evaluated using standard statistical probability back- testig methods as conditional and unconditional coverage. Based on empirical analysis thesis shows that SV models can perform at least as good as GARCH models if not superior in forecasting volatility and parametric VaR. 1
|
|
|
Value-at-risk forecasting with the ARMA-GARCH family of models during the recent financial crisis
Jánský, Ivo ; Rippel, Milan (advisor) ; Seidler, Jakub (referee)
The thesis evaluates several hundred one-day-ahead VaR forecasting models in the time period between the years 2004 and 2009 on data from six world stock indices - DJI, GSPC, IXIC, FTSE, GDAXI and N225. The models model mean using the AR and MA processes with up to two lags and variance with one of GARCH, EGARCH or TARCH processes with up to two lags. The models are estimated on the data from the in-sample period and their forecasting accuracy is evaluated on the out-of-sample data, which are more volatile. The main aim of the thesis is to test whether a model estimated on data with lower volatility can be used in periods with higher volatility. The evaluation is based on the conditional coverage test and is performed on each stock index separately. Unlike other works in this field of study, the thesis does not assume the log-returns to be normally distributed and does not explicitly select a particular conditional volatility process. Moreover, the thesis takes advantage of a less known conditional coverage framework for the measurement of forecasting accuracy.
|
|
|
Value-at-risk forecasting with the ARMA-GARCH family of models during the recent financial crisis
Jánský, Ivo ; Rippel, Milan (advisor) ; Seidler, Jakub (referee)
The thesis evaluates several hundred one-day-ahead VaR forecasting models in the time period between the years 2004 and 2009 on data from six world stock indices - DJI, GSPC, IXIC, FTSE, GDAXI and N225. The models model mean using the AR and MA processes with up to two lags and variance with one of GARCH, EGARCH or TARCH processes with up to two lags. The models are estimated on the data from the in-sample period and their forecasting ac- curacy is evaluated on the out-of-sample data, which are more volatile. The main aim of the thesis is to test whether a model estimated on data with lower volatility can be used in periods with higher volatility. The evaluation is based on the conditional coverage test and is performed on each stock index sepa- rately. Unlike other works in this eld of study, the thesis does not assume the log-returns to be normally distributed and does not explicitly select a partic- ular conditional volatility process. Moreover, the thesis takes advantage of a less known conditional coverage framework for the measurement of forecasting accuracy.
|
|
|
Volatility Modeling of the PX Index
Dvořáčková, Anna ; Borovička, Adam (advisor) ; Zouhar, Jan (referee)
This thesis is focused on modeling of the real financial time series of the PX Index using linear and nonlinear volatility models. In the theoretical part the major terms and typical properties of the financial time series are presented and it is followed by the theoretical description of the linear and nonlinear volatility models including a general volatility model building. The key part of this thesis is the practical application of chosen linear and nonlinear volatility models on the time series of log returns of the PX Index. By using the real data set we verify if the volatility models are really capable of explaining the theoretical properties of the financial time series, such as volatility clustering, leptokurtic distribution and leverage effect.
|
guest ::
