Zobrazit v EDS

Testování symetrie rozdělení finančních dat

Uloženo v:
Podrobná bibliografie
Název: Testování symetrie rozdělení finančních dat
Autoři: Střasák, Cyril
Přispěvatelé: Zichová, Jitka, Cipra, Tomáš
Informace o vydavateli: 2025.
Rok vydání: 2025
Témata: Pearsonovo rozdělení typu IV|testování symetrie|finanční výnosy|skórový test, Pearson type IV distribution|symmetry testing|financial returns|score test
Popis: This bachelor thesis addresses the problem of testing the symmetry of finan- cial data distributions, which often exhibit significant skewness and kurtosis. Standard symmetry tests typically rely on the assumption of normality, which may lead to misleading conclusions. Therefore, this work employs a score test derived from the Pearson type IV distribution, which accounts for potential lep- tokurticity. The theoretical part provides an overview of Pearson distributions and derives the test statistic. A simulation study is performed to investigate the properties of the test, including its level and power for various parameter values of the Pearson distribution and different sample sizes. In the empirical section, the test is applied to real data from the U.S. stock market, successfully identifying asymmetry in the returns of selected companies. The results show that the proposed test offers a reliable alternative to commonly used methods, particularly in cases where the normality assumption does not hold. 1
Druh dokumentu: Bachelor thesis
Jazyk: Czech
Přístupová URL adresa: http://www.nusl.cz/ntk/nusl-686818
Přístupové číslo: edsair.od......2186..496e40f7465ca43f65aca51e607fa87c
Databáze: OpenAIRE
Popis
Abstrakt:This bachelor thesis addresses the problem of testing the symmetry of finan- cial data distributions, which often exhibit significant skewness and kurtosis. Standard symmetry tests typically rely on the assumption of normality, which may lead to misleading conclusions. Therefore, this work employs a score test derived from the Pearson type IV distribution, which accounts for potential lep- tokurticity. The theoretical part provides an overview of Pearson distributions and derives the test statistic. A simulation study is performed to investigate the properties of the test, including its level and power for various parameter values of the Pearson distribution and different sample sizes. In the empirical section, the test is applied to real data from the U.S. stock market, successfully identifying asymmetry in the returns of selected companies. The results show that the proposed test offers a reliable alternative to commonly used methods, particularly in cases where the normality assumption does not hold. 1