|
|
Portfolio selection based on hierarchical structure of its components
Ševinský, Robert ; Krištoufek, Ladislav (advisor) ; Rusnák, Marek (referee)
This thesis investigate empirical performance of three portfolio selection and covariance matrix models. The goal is to find a strategy that outperform equally weighted portfolio in the long run and survives even in times of finan- cial distress. Two models based on Markowitz approach absolutely failed in this context, however the last approach based on network analysis indeed out- perform the market even after risk adjustment of returns. Moreover this model have sparse transaction matrix throughout time, therefore exhibit excellent properties even in the presence of transaction costs. Results for network based portfolio were obtained from running a back test on 160 member companies of S&P 500 index for 6'000 trading days. JEL Classification G11, G32, C10 Keywords Portfolio selection, Minimum spanning tree, Transaction costs, Covariance matrix Author's e-mail r.sevinsky@gmail.com Supervisor's e-mail kristoufek@ies-prague.org
|
|
|
Portfolio selection based on hierarchical structure of its components
Ševinský, Robert ; Krištoufek, Ladislav (advisor) ; Rusnák, Marek (referee)
This thesis investigate empirical performance of three portfolio selection and covariance matrix models. The goal is to find a strategy that outperform equally weighted portfolio in the long run and survives even in times of finan- cial distress. Two models based on Markowitz approach absolutely failed in this context, however the last approach based on network analysis indeed out- perform the market even after risk adjustment of returns. Moreover this model have sparse transaction matrix throughout time, therefore exhibit excellent properties even in the presence of transaction costs. Results for network based portfolio were obtained from running a back test on 160 member companies of S&P 500 index for 6'000 trading days. JEL Classification G11, G32, C10 Keywords Portfolio selection, Minimum spanning tree, Transaction costs, Covariance matrix Author's e-mail r.sevinsky@gmail.com Supervisor's e-mail kristoufek@ies-prague.org
|
|
|
Hierarchical Structure Analysis with Applications to the EU Member States Convergence
Fučík, Vojtěch ; Krištoufek, Ladislav (advisor) ; Bobková, Božena (referee)
The main objective of this thesis is to summarize and possibly extend the existing methodology on correlation matrix filtering, hierarchical clustering and topological classification in the economic networks. In the thesis we use classical MST/HT approach supplemented by edges stability analysis and centrality measures analysis. Graphical objects MST and HT enable us to find relations among the elements of the network. Centrality measures anal- ysis helps us to find the hubs in the network and stability analysis determines the reliability of the resulting model. Presented methodology is then utilized for convergence analysis in the EU and for analysis of clusters in the EU's MSTs and HTs. We detected large clusters of former communist countries for every economic indicator, clusters based on geographical location such as Nordic, Baltic, BENELUX or former ECSC countries and a cluster of PIGS countries. We also found that Spain plays a role of a central node in debt/deficit indicator analysis which made us to express our concerns about potential future problems. 1
|
|
|
Food vs. Fuel: The Role of Bioenergy
Filip, Ondřej ; Janda, Karel (advisor) ; Pertold-Gebicka, Barbara (referee)
This thesis studies the relationship between the first generation biofuels and selected commodities and assets in the USA, Europe, and Brazil. It is the first attempt to combine the taxonomy and wavelet analyses in a single research application. Our unique dataset comprises 32 weekly price series covering the 2003--2015 time period. First, we employ a method of minimum spanning trees and hierarchical trees to model a biofuel-related price network. We demonstrate a development phase shift between Brazilian and the US/EU biofuel industries. We reveal a strong and stable connection between Brazilian ethanol and its main production factor, local sugarcane. We further find that US ethanol is closely linked to corn. In the contrary, European biodiesel exhibits only moderate ties to its production factors. Subsequent wavelet analysis scrutinizes the identified price connections both in time and frequency domains. Both Brazilian and US ethanols are found to be positively related to their respective feedstock commodities. In particular, feedstock proves to lead the price of the biofuel and not vice versa. Moreover, the dynamics remains qualitatively unchanged when controlled for the influence of crude oil.
|
|
|
Food vs. Fuel: The Role of Bioenergy
Filip, Ondřej ; Janda, Karel (advisor) ; Pertold-Gebicka, Barbara (referee)
Bibliographic entry Filip, O. (2015): "Food vs. Fuel: The Role of Bioenergy." (Unpublished master's thesis). Charles University in Prague. Supervisor: prof. Ing. Karel Janda M.A., Dr., Ph. D. Length: 121,976 characters Abstract This thesis studies the relationship between the first generation biofuels and selected commodities and assets in the USA, Europe, and Brazil. It is the first attempt to combine the taxonomy and wavelet analyses in a single research application. Our unique dataset comprises 32 weekly price series covering the 2003-2015 time period. First, we employ a method of minimum spanning trees and hierarchical trees to model a biofuel-related price network. We demonstrate a development phase shift between Brazilian and the US/EU biofuel industries. We reveal a strong and stable connection between Brazilian ethanol and its main production factor, local sugarcane. We further find that US ethanol is closely linked to corn. In the contrary, European biodiesel exhibits only moderate ties to its production factors. Subsequent wavelet analysis scrutinizes the identified price connections both in time and frequency domains. Both Brazilian and US ethanols are found to be positively related to their respective feedstock commodities. In particular, feedstock proves to lead the price of the biofuel and not...
|
guest ::
