Institutional Repository of Vilnius University
Rizikuojamos vertės (VaR) skaičiavimas naudojant Gauso mišinio modelius /
Abstract
dc:descriptionThe statement that stock returns are normally distributed is one of the basic assumptions of financial mathematics. However, a growing number of empirical studies that have denied this theoretical foundation indicates that it is becoming increasingly significant to examine the suitability of other distributions in financial risk management. In this research, a Gaussian Mixture Model is proposed as a potential alternative to the normal distribution. The main objective of this study is to implement different Value-at-Risk approaches, compare the forecasting possibilities of the Gaussian Mixture Model and other popular VaR approaches, and identify the most appropriate model for the data. This research uses data on the daily returns of ten stocks from 2006 to 2010 and from 2016 to 2021. The study has found that the Gaussian Mixture Models with varying numbers of components are the most suitable VaR and ES approaches, whereas methods based on normality assumptions are considered suboptimal. The results of the statistical and graphical analysis revealed that Gaussian Mixture Models are able to adjust to significant stock market changes quite quickly and adequately while the remaining methods tend to be markedly delayed. The study also found that the ratio of short-term and long-term standard deviations significantly improves the Gaussian Mixture Model's ability to forecast VaR, consequently, this is considered to be the essential assumption of this model.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2022
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Morkūnaitė, Indrė,
Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- lit
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD146238087&prefLang=en_US