
Can Model Combination Improve Volatility Forecasting?
Tyuleubekov, Sabyrzhan ; Baruník, Jozef (advisor) ; Červinka, Michal (referee)
Nowadays, there is a wide range of forecasting methods and forecasters encounter several challenges during selection of an optimal method for volatility forecasting. In order to make use of wide selection of forecasts, this thesis tests multiple forecast combination methods. Notwithstanding, there exists a plethora of forecast combination literature, combination of traditional methods with machine learning methods is relatively rare. We implement the following combination techniques: (1) simple mean forecast combination, (2) OLS combination, (3) ARIMA on OLS combined fit, (4) NNAR on OLS combined fit and (5) KNN regression on OLS combined fit. To our best knowledge, the latter two combination techniques are not yet researched in academic literature. Additionally, this thesis should help a forecaster with three choice complication causes: (1) choice of volatility proxy, (2) choice of forecast accuracy measure and (3) choice of training sample length. We found that squared and absolute return volatility proxies are much less efficient than Parkinson and GarmanKlass volatility proxies. Likewise, we show that forecast accuracy measure (RMSE, MAE or MAPE) influences optimal forecasts ranking. Finally, we found that though forecast quality does not depend on training sample length, we see that forecast...


Pricing Options Using Monte Carlo Simulation
Dutton, Ryan ; Dědek, Oldřich (advisor) ; Červinka, Michal (referee)
Monte Carlo simulation is a valuable tool in computational finance. It is widely used to evaluate portfolio management rules, to price derivatives, to simulate hedging strategies, and to estimate Value at Risk. The purpose of this thesis is to develop the mathematical foundation and an algorithmic structure to carry out Monte Carlo simulation to price a European call option, investigate BlackScholes model to look into the parallel between Monte Carlo simulation and BlackScholes model, provide a solution for BlackScholes model using Lognormal distribution of a stock price rather than solving BlackScholes original partial differential equation, and finally compare the results of Monte Carlo simulation with Black Scholes closedform formula. Author's contribution can be best described as developing the mathematical foundation and the algorithm for Monte Carlo simulation, comparing the simulation results with the BlackScholes model, and investigating how pathdependent options can be implemented using simulation when closedform formulas may not be available. JEL Classification C02, C6, G12, G17 Keywords Monte Carlo simulation, Option pricing, BlackScholes model Author's email ryandutton4@gmail.com Supervisor's email oldrich.dedek@fsv.cuni.cz


At the right time, in the right factor. Can factors be timed?
Nosek, Jiří ; Hronec, Martin (advisor) ; Červinka, Michal (referee)
This thesis examines the controversial prospect of Factor timing. We use Thompson Reuters data that allow us to construct international riskfactors and respective predictive signals and we test the capacity of these signals to time factors using the Kelly Criterion formula to determine the optimal fraction of capital to invest. Concerning the United States market, we showed that among all signals that we used only the Value Spread seems to contain some predictive power for all the factors in the study. All other timing signals were almost uniformly disappointing and were unable to time any of the factors. We further showed that timing strategies performed much better in the intentional setting, often outperforming the passive buyand hold approach. JEL Classification G12, G14, G17, G19 Keywords factors, factor timing, timeseries, Kelly crite rion, empirical analysis Title At the right time, in the right factor. Can factors be timed? Author's email jirkaa.nosek@seznam.cz Supervisor's email martin.hronec@fsv.cuni.cz 1


Trading volume and expected stock returns: a metaanalysis
Bajzík, Josef ; Havránek, Tomáš (advisor) ; Červinka, Michal (referee)
I investigate the relationship between expected stock returns and trading volume. I collect together 522 estimates from 46 studies and conduct the first metaanalysis in this field. Use of Bayesian model averaging and Frequentist model averaging help me to discover the most influential factors that affect the returnvolume relationship, since I control for more than 50 differences among primary articles such as midyear and type of data, length of the primary dataset, size of market, or model employed. In the end, I find out that the relation between expected stock returns and trading volume is rather negligible. On the other hand, the contemporaneous relation between returns and volume is positive. These two findings cut the mixed results from previously written studies. Moreover, the investigated relationship is influenced by the size of country of interest and the level of its development. Besides the primary studies that employ higher data frequency provide substantially larger estimates than the studies with data from longer time periods. On the contrary, there is no difference among different estimation methodologies used. Finally, I employ classical and modern techniques such as stembased methodology for publication bias detection, and I find evidence for it in this field. 1


Relationship of Economic Growth and Pollution in the Czech Republic
Moldan, Martin ; Červinka, Michal (advisor) ; Valíčková, Petra (referee)
The Environmental Kuznets Curve (EKC) is a hypothesized relationship between GDP per capita and pollution. It suggests that the relationship has a shape of a concave quadratic functioni.e. that firstly, with increasing GDP per capita, levels of pollution increase. And then, from some level of GDP per capita, as GDP per capita rises, levels of pollution decrease. This bachelor thesis examines whether the EKC holds for the Czech Republic or not. It uses panel data on air pollution for the period 19952017, in particular concentrations SO2 and NOx. This analysis is conducted using the fixed effects method. The results of this bachelor thesis suggest that for the case of SO2, there is a relationship between GDP per capita and the pollutant's concentrations. However, this relationship does not change over time significantly. Moreover, for the case of NOx, the relationship between the pollutant's concentrations and GDP per capita is not significant, hence, the EKC hypothesis can be rejected for both examined pollutants.


Rivals as Allies: Combining Fundamental and Technical Analysis for Stock Investing
Buinevici, Igor ; Červinka, Michal (advisor) ; Čech, František (referee)
The main aim of this thesis is to perform a detailed investigation of cer tain investment strategies based on European stock data. There are four investment strategies overall that are examined from the performance per spective: momentum strategy, momentum strategy with BOS ratio filtering, fundamental buy and hold strategy using F SCORE and a key combined strategy incorporating all the methods mentioned above. After the esti mation of Fama and French threefactor model for the combined strategy using the aggregated group of stocks, it can be inferred that in the case of a monthly rebalancing this strategy generates statistically significant monthly riskadjusted return of 0.938%. For the threemonth, sixmonth and nine month holding periods the conclusion for the aggregated group of stocks is similar  in all of these cases the combined strategy also generates statisti cally significant riskadjusted returns. Based on further comparative testing of strategies for the aggregated group of stocks, it can be stated that the combined investment strategy significantly outperforms all other strategies in terms of returns, especially in the case of a onemonth holding period. Keywords: portfolio analysis, fundamental analysis, tech nical analysis, stock investing, empirical testing JEL...


Impact of Terrorism on Economic Growth
Siegl, Jakub ; Červinka, Michal (advisor) ; Kolcunová, Dominika (referee)
The negative market atmosphere resulting from terrorism may potentially affect key macroeconomic variables and be reflected in economic growth both immediately and with time lags. This thesis utilizes quarterly data on variables related to terrorism and key macroeconomic metrics for the time period 19702017 and establishes the effect of terrorism on economic growth. Furthermore, it elaborates on the change of general perception of terrorism after the 9/11 2001 attack and assesses the difference of its effect before and after this key violent act. In general, it has been found that the deaths and wounds resulting from terrorism affect economic growth with lags. Further more, following the 9/11 2001 terrorist attack, the time layout of the effect of deaths resulting from terrorism has changed. Keywords terrorism, economic growth, panel data analysis, fixed effects model, mac roeconomic metrics


Artificial Neural Networks in Option Pricing
Vach, Dominik ; Gapko, Petr (advisor) ; Červinka, Michal (referee)
This thesis examines the application of neural networks in the context of option pricing. Throughout the thesis, different architecture choices and prediction parameters are tested and compared in order to achieve better performance and higher accuracy in option valuation. Two different volatility forecast mechanisms are used to compare neural networks performance with Black Scholes parametric model. Moreover, the performance of a neural network is compared also to more advanced modular neural networks. A new technique of adding rational prediction assumptions to neural network prediction is tested and the thesis shows the importance of adding virtual options fulfilling these assumptions in order to achieve better training of the neural network. This method comes out to increase the prediction power of the network significantly. The thesis also shows the neural network prediction outperforms the traditional parametric methods. The size and number of hidden layers in a neural network is tested with an emphasis to provide a benchmark and a structured way how to choose neural network parameters for future applications in option pricing. JEL Classification C13, C14, G13 Keywords Option pricing, Neural networks, Modular neu ral networks, S&P500 index options Author's email vach.dominik@gmail.com...


Analysis of Term Structures in High Frequencies
Nedvěd, Adam ; Baruník, Jozef (advisor) ; Červinka, Michal (referee)
This thesis represents an indepth empirical study of the dependence structures within the term structure of interest rates. Firstly, a comprehensive overview of term structure modelling literature and methods is provided together with a summary of theoretical notions regarding the use of highfrequency data and spectral analysis. Contrary to most studies, the frequencydomain approach is employed, with a special focus on dependency across various quantiles of the joint distribution of the term structure. The main results are obtained using the quantile crossspectral analysis, a new robust and nonparametric method allowing to uncover dependence structures in quantiles of the joint distribution of multivariate time series. The results are estimated using a dataset consisting of 15 years worth of highfrequency tickbytick time series of US Treasury futures. Complex dependence structures are revealed showing signs of both cyclicity and dependence in various parts of the joint distribution of the term structure in the frequency domain. JEL Classification C49, C55, C58, E43, G12, G13 Keywords term structure of interest rates, yield curves, highfrequency analysis, spectral analysis, inter est rate futures Author's email adam.nedved@fsv.cuni.cz Supervisor's email barunik@fsv.cuni.cz


The Facility Location Problem for an Alternative Delivery Service
Vaněk, Petr ; Červinka, Michal (advisor) ; Habibi, Arash (referee)
The main intention of the bachelor thesis is to introduce all necessary theory to under stand and apply the facility location problem for an alternative delivery service. Solving the facility location problem we are able to find optimal warehouse locations to minimize transportation costs from a firm to customers. Our attention is focused on an alternative delivery service which is slowly becoming to be taken seriously in logistic disciplines be cause many companies are trying to apply drones or robots for shipping goods. The first part of this thesis deals with an important theoretical background needed to understand the facility location problem. Microeconomic view of the problem is presented, basic knowledge of mathematical programming with algorithms is explained and the facility location problem itself is defined. Moreover, each step of the branch & bound algorithm for the facility location problem is explained on an example from Prague. In the practical part of the thesis, we solve a real problem for the Czech company which would like to im plement a drone delivery service in Prague. Finally, we discuss future technical progress and its impacts on autonomous delivery service, legislative and privacy problems with a drone delivery system in Prague and other possible usage of autonomous vehicles in...
