|
| |
|
|
Prediction of Multiple Time Series at Stock Market Trading
Palček, Peter ; Zbořil, František (referee) ; Rozman, Jaroslav (advisor)
The diploma thesis comprises of a general approach used to predict the time series, their categorization, basic characteristics and basic statistical methods for their prediction. Neural networks are also mentioned and their categorization with regards to the suitability for prediction of time series. A program for the prediction of the progress of multiple time series in stock market is designed and implemented, and it's based on a model of flexible neuron tree, whose structure is optimized using immune programming and parameters using a modified version of simulated annealing or particle swarm optimization. Firstly, the program is tested on its ability to predict simple time series and then on its ability to predict multiple time series.
|
|
|
Backtesting of Time Series Models
Stroukalová, Marika ; Houfková, Lucia (advisor) ; Zichová, Jitka (referee)
Title: Backtesting of Time Series Models Author: Marika Stroukalová Department: Department of Probability and Mathematical Statistics Supervisor: Mgr. Lucia Jarešová Supervisor's e-mail address: lucia.jaresova@centrum.cz Abstract: In the present work we study the basic models of financial time series (ARMA, GARCH), we focus on parameter estimation and forecasting in estimated models. We describe the means of estimating parametres and future values in the program R. In the theoretical section we also discuss the features of financial time series, define simple returns and log returns and we introduce the benefits of the log returns. We also apply the white noise model, ARMA(1,1) and GARCH(1,1) on historic time series of logarithmic returns of chosen stock exchange indices, we also backtest 1-step ahead fore- cats and 5-step ahead forecasts and we compare the results of these models. By empirical comparison of real data we also analyze how the models reac- ted on the present financial crisis and evaluate how the normal distribution assumption for the data held up. Keywords: time series, ARMA, GARCH, backtesting. 1
|
|
|
Software products for financial time series analysis
Vlasáková, Romana ; Zichová, Jitka (advisor) ; Cipra, Tomáš (referee)
The present work deals with selected methods suitable to work with financial time series. Firstly, univariate linear models ARMA are introduced, followed by the description of volatility models ARCH and their generalization to GARCH models. There are many modifications of standard GARCH models designed with respect to the nature of financial data, some of which are presented. Another part of the work dealing with multiple time series focuses on VAR models and bivariate GARCH models. The most important part of the work are practical examples of building the theoretically described models in various types of software with built-in procedures for time series analysis. We apply five different types of commercial and non-commercial software, namely EViews, Mathematica, R, S-PLUS and XploRe. The used software products are presented and compared in terms of their capabilities and the results obtained for particular methods.
|
|
|
Parameter estimating in time series models
Kostárová, Aneta ; Zichová, Jitka (advisor) ; Prášková, Zuzana (referee)
This bachelor thesis deals with some methods of parameter estimating in linear time series models. The most used approach in software products is the maximum likelihood estimation. The theoretical part explains the parameter estimation of the ARMA model by conditional and unconditional maximum likelihood estimation and demonstrates both methods for lower order models. The practical part examines and describes the imple- mentation of parameter estimating in Mathematica and R software. The comparison of the quality of the estimates calculated by various procedures of the chosen software is included. Finally, the acquired findings is used in a simulation study. 1
|
|
|
Econometric Modelling and Forecasting of Natural Gas Spot Prices
Kubišová, Barbora ; Hendrych, Radek (advisor) ; Hudecová, Šárka (referee)
The thesis deals with modeling and forecasting of natural gas spot prices, con- sumption of natural gas and average daily temperature. We assume that these three variables are influenced by each other, because as temperature decreases, consumption increases, which in turn increases the price with the increasing de- mand. Therefore, we propose to model these variables by vector autoregression. We compare this model with one-dimensional models where for each one we build a model from the ARMA-GARCH class. Models are estimated using historic va- lues and then designed models are used to simulate scenarios. Analysis of scenarios provides information to gas supply companies estimates of portfolio consumption and financial flows related to the purchase concerning natural gas. 1
|
|
|
Linear and bilinear models for time series from economics and finance
Kotrbová, Anežka ; Zichová, Jitka (advisor) ; Prášková, Zuzana (referee)
This bachelor thesis deals with linear and bilinear models used for modelling time series data applicable in economy and finance. The thesis consists of a theoretical and a practical part. The theoretical part briefly describes ARMA and bilinear process, issues of linear model identification, estimation of the parameters and moment properties of ARMA(1, 1) a BL(1, 0, 1, 1). The typical characteristics of bilinear models and the quality of the estimated parameters are examined by the simulation study in software Mathematica 10. The acquired findings are applied in search for a suitable model for time series of share prices of the company ČEZ. Powered by TCPDF (www.tcpdf.org)
|
|
|
Unit root testing with applications to financial time series
Pechmanová, Kateřina ; Zichová, Jitka (advisor) ; Hendrych, Radek (referee)
This work deals with linear ARMA processes, which are intended to describe the behavior of time series, and also with analysis of selected time series. First, the basic concepts are introduced together with the descriptions of the ARMA models. Further, the Dickey-Fuller test for a unit root, as an approach to the verification of nonstationary time series, is introduced. An important part is the practical application of these models and tests on simulated and real data. Real analyzed data capture developments in the exchange rate of Czech crown against Euro. All calculations were performed in the Mathematica software. Powered by TCPDF (www.tcpdf.org)
|
|
|
Software products for financial time series analysis
Vlasáková, Romana ; Zichová, Jitka (advisor) ; Cipra, Tomáš (referee)
The present work deals with selected methods suitable to work with financial time series. Firstly, univariate linear models ARMA are introduced, followed by the description of volatility models ARCH and their generalization to GARCH models. There are many modifications of standard GARCH models designed with respect to the nature of financial data, some of which are presented. Another part of the work dealing with multiple time series focuses on VAR models and bivariate GARCH models. The most important part of the work are practical examples of building the theoretically described models in various types of software with built-in procedures for time series analysis. We apply five different types of commercial and non-commercial software, namely EViews, Mathematica, R, S-PLUS and XploRe. The used software products are presented and compared in terms of their capabilities and the results obtained for particular methods.
|
|
|
Behaviour of Stocks on the Prague Stock Exchange During the Financial Crisis: Evidence from Empirical Research
Koza, Oldřich ; Teplý, Petr (advisor) ; Krištoufek, Ladislav (referee)
This work studies the behaviour of the four most traded stocks on the Prague Stock Exchange from January 2007 to July 2010. Its main goal is to describe how the financial crisis influenced the Prague Stock Exchange. Employing standard statistical methods, ARMA, GARCH, and VAR models I examine on daily data the following phenomena: volatility, price jumps, the day of the week effect, validity of the efficient market hypothesis, and information flow between the stocks. The results imply that the financial crisis had stronger impact on the banking sector stocks than on other stocks. The crisis was mainly characterized by rapid growth in volatility and correlation between the stocks. It also influenced the information flow and the day of the week effect. However, the crisis did not trigger growth in the number of extreme price movements, and it did not cause the market to be less information efficient.
|
guest ::
