|
|
Non-Linear Classification as a Tool for Predicting Tennis Matches
Hostačný, Jakub ; Baniar, Matúš (advisor) ; Krištoufek, Ladislav (referee)
Charles University Faculty of Social Sciences Institute of Economic Studies MASTER'S THESIS Non-Linear Classification as a Tool for Predicting Tennis Matches Author: Be. Jakub Hostacny Supervisor: RNDr. Matus Baniar Academic Year: 2017/2018 Abstract In this thesis, we examine the prediction accuracy and the betting performance of four machine learning algorithms applied to men tennis matches - penalized logistic regression, random forest, boosted trees, and artificial neural networks. To do so, we employ 40 310 ATP matches played during 1/2001-10/2016 and 342 input features. As for the prediction accuracy, our models outperform current state-of-art models for both non-grand-slam (69%) and grand slam matches (79%). Concerning the overall accuracy rate, all model specifications beat backing a better-ranked player, while the majority also surpasses backing a bookmaker's favourite. As far as the betting performance is concerned, we develop six profitable betting strategies for betting on favourites applied to non-grand-slam with ROI ranging from 0.8% to 6.5%. Also, we identify ten profitable betting strategies for betting on favourites applied to grand slam matches with ROI fluctuating between 0.7% and 9.3%. We beat both bench mark rules - backing a better-ranked player as well as backing a bookmaker's...
|
|
|
Extending volatility models with market sentiment indicators
Röhryová, Lenka ; Krištoufek, Ladislav (advisor) ; Jakubík, Petr (referee)
In this thesis, we aim to improve forecast accuracy of a heterogenous au- toregressive model (HAR) by including market sentiment indicators based on Google search volume and Twitter sentiment. We have analysed 30 com- panies of the Dow Jones index for a period of 15 months. We have performed out-of-sample forecast and compiled a ranking of the extended models based on their relative performance. We have identified three relevant variables: daily negative tweets, daily Google search volume and weekly Google search volume. These variables improve forecast accuracy of the HAR model se- parately or in a Twitter-Google combination. Some specifications improve forecast accuracy by up to 22% for particular stocks, others impair forecast accuracy by up to 24%. The combination of daily negative tweets and weekly search volume is a superior model to the basic HAR for 17 stocks according to RMSE and for 16 stocks according to MAE and MASE. The daily nega- tive tweets specification outperforms the basic HAR for 17 and 19 stocks, respectively. And, the combination of daily negative tweets and daily search volume outpaces the basic HAR for 15 and 18 stocks, respectively. Based on the average MASE improvement, the combination of daily negative tweets and weekly search volume is a clear winner as it lowers the...
|
|
|
The impact of renewable resources on price volatility in the European power markets
Líšková, Katarína ; Krištoufek, Ladislav (advisor) ; Luňáčková, Petra (referee)
Integration of renewable energy sources impacts electricity spot price and its variation. Remaining open question is, in which direction. Volatility fluctuations threaten secur- ity of electricity supply, influence trading strategies and create uncertainty in optimal installed capacity planning. In this thesis, drivers of price volatility in Czech and Ger- man day-ahead power market are analysed with an emphasis on penetration of renewable energy sources. To the best of our knowledge, this is the first study focused on this issue in Czech electricity market. We apply recently developed approach of quadratic variation theory with an adjustment for electricity prices. Realised volatility is divided into its continuous and jump component. The continuous part is modelled by three het- erogeneous autoregressive models, differing in complexity and inclusion of market-specific fundamental variables. Amendments to each model for the particular market are proposed and the models are evaluated both in-sample and out-of-sample. Addition of exogenous variables − commodity prices, weather conditions and seasonal variables − to simpler heterogeneous autoregressive model is found to improve volatility forecast accuracy. The results suggest higher continuous volatility due to increased penetration of power from wind...
|
|
| |
|
|
Impact of Czech intraday market on the electricity prices
Béreš, Samuel ; Krištoufek, Ladislav (advisor) ; Valíčková, Petra (referee)
We analyse Czech intraday market for electricity and its impact on day- ahead prices. We inspect effect of fundamental drivers of price deviation between intraday and day-ahead market in form of positive and negative forecast errors and examine intraday price's role in explaining next trading period's day-ahead price. Our findings suggest photovoltaic and load fore- cast errors to be most statistically significant fundamental factors, together with autoregressive term and day-ahead price, determining intraday market price deviation from day-ahead. Variables' influences on intraday market are in accordance with hypothesised expectations, except for the effect of export and excessive import of electricity to and from German TSO, 50 Hertz, and extreme day-ahead prices. We confirmed symmetric effects of forecast errors on intraday price for all observed variables. In the second part, intraday prices are found to be statistically significant factor affecting next day's day-ahead market price. The results support the conclusion that Czech spot market for electricity possesses mean-reverting properties. Keywords electricity, intraday market for electricity, price modeling
|
|
|
Portfolio selection in factor investing
Hronec, Martin ; Baruník, Jozef (advisor) ; Krištoufek, Ladislav (referee)
This thesis empirically examines the role of advanced portfolio selection methods in factor investing. These methods provide more efficient exposure to underlying risk sources in factor portfolios. Their performance is evaluated across number of prominent factors and compared with more naive equal- and value- weighting, typically used in asset pricing literature as well commercial investment vehicles. The most diversified portfolio consistently achieves the highest returns, while having only moderate volatility and one of the lowest tail risk exposure. On the other hand, the diversified risk parity portfolio suffers high volatility as well as the greatest tail risk exposure, while achieving only comparable average returns with other strategies. 1
|
|
|
Comparison of different models for forecasting of Czech electricity market
Kunc, Vladimír ; Krištoufek, Ladislav (advisor) ; Kopečná, Vědunka (referee)
There is a demand for decision support tools that can model the electricity markets and allows to forecast the hourly electricity price. Many different ap- proach such as artificial neural network or support vector regression are used in the literature. This thesis provides comparison of several different estima- tors under one settings using available data from Czech electricity market. The resulting comparison of over 5000 different estimators led to a selection of several best performing models. The role of historical weather data (temper- ature, dew point and humidity) is also assesed within the comparison and it was found that while the inclusion of weather data might lead to overfitting, it is beneficial under the right circumstances. The best performing approach was the Lasso regression estimated using modified Lars. 1
|
|
|
Prediction of Stock Return Volatility Using Internet Data
Juchelka, Tomáš ; Krištoufek, Ladislav (advisor) ; Novák, Jiří (referee)
The thesis investigates relationship between daily stock return volatility of Dow Jones Industrial Average stocks and data obtained on Twitter, the social media network. The Twitter data set contains a number of tweets, categorized according to their polarity, i.e. positive, negative and neutral sentiment of tweets. We construct two classes of models, GARCH and ARFIMA, where for either of them we research basic model setting and setting with additional Twitter variables. Our goal is to compare, which of them predicts the one day ahead volatility most precisely. Besides, we provide commentary regarding the effects of Twitter volume variables on future stock volatility. The analysis has revealed that the best performing model, given the length and structure of our data set, is the ARFIMA model augmented on Twitter volume residuals. In the context of the thesis, Twitter volume residuals represent unexpected activity on the social media network and are obtained as residuals from Twitter volume autoregression. Plain ARFIMA model was the second best and plain volume augmented ARFIMA was in third place. This means that all three ARFIMA models outperformed all three GARCH models in our research. Regarding the Twitter estimation parameters, we found that higher the activity the higher tomorrow's stock...
|
|
|
Portfolio Construction Using Hierarchical Clustering
Fučík, Vojtěch ; Krištoufek, Ladislav (advisor) ; Baruník, Jozef (referee)
Hlavním cílem této práce je vyložit a zejména propojit existující metodologii filtrování korelačních matic, grafových algoritmů aplikovaných na minimální kostry grafu, hierarchického shlukování a analýzy hlavních komponent, pro vytvoření kvantitativních investičních strategií. Namísto tradičního použití časových řad akciových výnosů je užito reziduí z faktorových modelů. Tato rezidua jsou klíčovým vstupem pro všechny používané algoritmy k výpočtu pravděpodobnosti středovosti dané akcie. Pravděpodobnost středovosti je nekonvenční ukazatel pravděpodobnosti, kde hodnota blízko 1 značí vysokou pravděpodobnost středovosti dané akcie v dané ekonomické síti. Na základě této míry pravděpodobnosti je vybudováno několik investičních strategií, které jsou dále testován hlavních amerických akciových indexů. Nemůže být generalizováno, že periferní strategie dosahují konzistentně lepších výsledků než středové strategie. Zatímco při použití klasického Markowitzova optimalizačního procesu jsou zisky stabilní a potenciál průměrný, oba typy vybudovaných strategií (středové i periferní) sdílí vysoký potenciál zisku, který je ovšem vykoupen vysokou volatilitou.
|
|
|
Spillovers between low and high risk assets during business cycle
Matyáš, Jan ; Krištoufek, Ladislav (advisor) ; Kukačka, Jiří (referee)
1 Abstract This master thesis examines linkages among bond and stock markets in Ger- many, Austria and Italy. For the purpose of analysis of return spillovers, we use Spillover index framework which enables us to describe development of inter- market linkages over time. The data used in the study includes the period from January 2nd, 1998 to May 23rd, 2017 which allows us to estimate long- term development of spillovers among markets. We find unequal link between stocks and bonds and increase in co-integration of markets during the financial crisis of 2007-2008 with significant persistence after the crisis. Mechanism of transmission of financial shocks among European countries is affected by eco- nomic and political integration of countries. We identify strong interlinkages of markets with substantial influence of Italian assets in transmitting shocks to German and Austrian assets, especially during periods of economic distress. On the other hand, Germany represents an open economy that is increasingly integrated to other markets. Scale of return spillovers is highly dependent on economic situation which is evident from clustering of high spillovers during recessions and a great deal of persistence of these interdependencies. JEL Classification G01, G12, G15, C63, C67 Keywords return spillovers, asset...
|
guest ::
RSS feed.