|
|
Capital Market Hypotheses and Their Statistical Implications: A Comparative Study
Petras, Petr ; Krištoufek, Ladislav (advisor) ; Křehlík, Tomáš (referee)
In this bachelor thesis we focus on different Market Hypotheses. Specifically on Efficient Market Hypothesis, Fractal Market Hypothesis and Coherent Market Hypothesis. In the first part of the work we provide description of researched hypotheses and methods used for testing. In the second part of the work we run test on time series of share markets, gold markets and currency markets and test if our hypotheses can provide explanation about price changes on those markets. For Efficient Market Hypothesis we wonder if prices are following random walk (via augmented Dickey-Fuller test), if residuals are normally distributed (via Shapiro-Wilk and Jarque-Bera tests) and if residuals are uncorrelated (via Box-Pierce test). For Fractal Market Hypothesis we are trying to find value of Hurst exponent via Rescaled Range analysis. This exponent describes if time series are persistent or not. And for Coherent Market Hypothesis we develop simple method for testing if some time periods can yield above-average revenues, thanks to increased mean and decreased standard deviation. After that we find out what are consequences of short time series and different frequencies for obtaining data points and we learn that some hypotheses describes different time periods or lengths better and are not so good for different ones. Powered...
|
|
|
Are More Liquid Stocks Also More Efficient?
Kupka, Petr ; Krištoufek, Ladislav (advisor) ; Benčík, Daniel (referee)
Liquidity and informational efficiency are closely watched features of financial markets. Together with stock exchange size effect, captured by market capitalization, this thesis examines the triple of relationships among these three stock market properties. Applying methods of sequences and reversals ratio test, autocorrelation coefficient test and variance ratio test provided us with 14 proxy measures of efficiency for each stock. Daily prices and volumes traded for period 2003 - 2013 of 206 stocks sampled from 22 stock exchanges were used. The same data were used for Amihud illiquidity measure. The positive relationship between stock efficiency and liquidity was not strongly supported neither rejected. It turned out that stock liquidity is very strongly positively dependent on size of stock exchange where is that particular stock listed. It was also concluded that there are more efficient stocks listed in larger stock exchanged. JEL Classification: G12, G14, G15 Keywords: stock liquidity, stock exchange size, stock efficiency Author's e-mail: kupkapetr@gmail.com Supervisor's e-mail: kristoufek@ies-prague.org
|
|
|
Financial earthquakes: Are volatility correlations related to Omori's law?
Bureš, Vlastimil ; Krištoufek, Ladislav (advisor) ; Žigraiová, Diana (referee)
The purpose of this thesis is to analyze the market dynamics in periods following a large financial shock. In order to do so, we compute the cumulative number of times the volatility is greater than a given threshold. Such a method is analogous to Omori's law from geophysics. We draw statistical evidence from three different events. The first one is concerned with the death of Steve Jobs and how it affected the evolution of Apple's share price. The second one focuses on the Flash Crash of 2010 when the Dow Jones Industrial Average experienced the largest drop of 900 points. And the last one is when IBM announced its 2013Q1 earnings which were significantly below expectations. By employing two different approaches to volatility calculation, we are able to compare the obtained results and thus draw a more definite conclusion. Our findings suggest that the decay rate of after-shocks for the considered earthquakes is well described by a power-law which is analogous to Omori's law. Powered by TCPDF (www.tcpdf.org)
|
|
|
The fractal dimension and forecasting of financial time series
Kaplan, Robert ; Krištoufek, Ladislav (advisor) ; Džmuráňová, Hana (referee)
In this thesis, we strive to build on the fractal market hypothesis and to develop two methods which aim to reveal whether the fractal dimension, as a property of the short memory, can be applied for forecasting of financial time series. In the first one, we use ten world market indices and repeatedly estimate the fractal dimension by boxcount, Hall-Wood, and Genton estimators on fixed number of returns and make one step ahead forecasts by AR(1) and ARMA(1,1) models; then, we look whether forecast errors from realized returns are lower when the fractal dimension is estimated lower. The second method incorporates only the fractal dimension and studies, if the sign of return persists in next period more likely with lower fractal dimension. The results indicate that the short memory is truly present in the markets and the fractal dimension may be potentially useful for prediction and increased profit for investors. However, the significance of our results is not strong. We recommend more sophisticated methods and models for further research.
|
|
|
Google Econometrics: Unemployment in Visegrad Countries
Pavlíček, Jaroslav ; Krištoufek, Ladislav (advisor) ; Zeynalov, Ayaz (referee)
This thesis examines the relationship between job-related Google search query indices and unemployment rate in Visegrad countries. We found that the unemployment rate generally moves in the same direction as the search volume index for the job-related term. The series of Google search query indices also proved useful for prediction-making. Models with Google series showed lower MAE and RMSE of static forecast compared to base models in all four countries. However, only models for Poland and Slovakia showed potential for nowcasting. Powered by TCPDF (www.tcpdf.org)
|
|
|
Dissection of Bornholdt's model: examination of inner dynamics and effect of parameter change
Chrz, Štěpán ; Krištoufek, Ladislav (advisor) ; Vácha, Lukáš (referee)
Dissection of Bornholdt's model - Analysis of Inner Dynamics and Effect of Parameter Change Mgr. Štěpán Chrz Abstract In this work we thoroughly analyze Bornholdt's version of Ising model of ferro- magnetism, with emphasis on its ability to mimic some basic stylized facts of financial series. Initially, we provide a breakdown of model definition and anal- ysis of underlying dynamics. Subsequently, we examine and confirm model's ability to mimic stylized facts of financial series. To examine robustness of this ability to parameter change, we conduct simulations over a set of parameter combinations. We conclude that there is a wide set of combinations that yields acceptable simulation results. We also note that the seemingly best results are obtained at parameter values close to border of this set. 1
|
|
|
Relationship between liquidity and volatility of selected exchange rate pairs
Kotek, Martin ; Krištoufek, Ladislav (advisor) ; Benčík, Daniel (referee)
The thesis explores relationship between volatility and liquidity of ten selected exchange rate pairs. Several volatility and liquidity measures are computed and the relationship between volatility-liquidity pairs is tested for cointegration and Granger causality; impulse response functions are computed as well. We find that volatility measures provide similar information (are cointegrated), while volatility measures differ to a large extent. A few cointegrating relationships between volatility and liquidity are found, but they are specific to only some currency pairs. Granger causality tests give different results for different currency pairs, but in general, the relationship between volatility and liquidity is two-way (feedback). Shocks in volatility or liquidity have little impact on the other and quickly fade away, usually within one or two days. Powered by TCPDF (www.tcpdf.org)
|
|
|
Modeling Conditional Quantiles of Central European Stock Market Returns
Burdová, Diana ; Baruník, Jozef (advisor) ; Krištoufek, Ladislav (referee)
Most of the literature on Value at Risk concentrates on the unconditional nonparametric or parametric approach to VaR estimation and much less on the direct modeling of conditional quantiles. This thesis focuses on the direct conditional VaR modeling, using the flexible quantile regression and hence imposing no restrictions on the return distribution. We apply semiparamet- ric Conditional Autoregressive Value at Risk (CAViaR) models that allow time-variation of the conditional distribution of returns and also different time-variation for different quantiles on four stock price indices: Czech PX, Hungarian BUX, German DAX and U.S. S&P 500. The objective is to inves- tigate how the introduction of dynamics impacts VaR accuracy. The main contribution lies firstly in the primary application of this approach on Cen- tral European stock market and secondly in the fact that we investigate the impact on VaR accuracy during the pre-crisis period and also the period covering the global financial crisis. Our results show that CAViaR models perform very well in describing the evolution of the quantiles, both in abso- lute terms and relative to the benchmark parametric models. Not only do they provide generally a better fit, they are also able to produce accurate forecasts. CAViaR models may be therefore used as a...
|
|
|
How do the efficient portfolios at various investment horizons differ?
Růžek, Pavel ; Krištoufek, Ladislav (advisor) ; Křehlík, Tomáš (referee)
The Efficient Market Theory that assumes the homogeneity of investors' ex- pectations has several shortcomings and has failed to predict development of fi- nancial markets many times, recently. Previous research, therefore, has focused more intensively on incorporation of some aspects from Behavioural Finance to their models. This thesis implements another form of heterogeneity coming from different investment horizon preferences, and investigates the impacts on the selection of the efficient portfolios compared to the original Markowitz's framework. We employed the mean-variance model adjusted for the purpose of the work, and, additionally, suggested extensions that assure robustness of the model and the highest possible objectivity of the empirical results inde- pendently on the choice of data sets. The findings from our research strongly confirmed proposed hypotheses that the efficient portfolios do differ at the var- ious investment horizons and that the efficient portfolios for long investment horizons are less risky. JEL Classification G10, G11 Keywords portfolio selection, mean-variance, optimization, investment horizons, Dow Jones Index Author's e-mail pavel.ruzek.ies@gmail.com Supervisor's e-mail kristoufek@ies-prague.org
|
|
|
Applicability of online sentiment analysis for stock market prediction
Rýgr, Petr ; Krištoufek, Ladislav (advisor) ; Křehlík, Tomáš (referee)
The purpose of this thesis is to explore various possibilities of performing online sentiment analysis and utilizing obtained information in stock market prediction. Firstly, several tools and sources available for sentiment analysis are presented and brief history of research related to each tool is provided. Additionally, Google Trend model is designed to evaluate whether information about searching volume of selected terms can be used to predict future movements of S&P 500 index. Strategy based on such model is implemented on historical data and its cumulative return is compared to classical buy and hold strategy. Furthermore, hypothesis whether it is possible to utilize publicly released news as a leading indicator for future stock returns is tested. Lastly, process of algorithmic sentiment analysis is described and its strengths and weaknesses are assessed.
|
guest ::
