Institutional Repository of Vilnius University
Investicinio portfelio optimizavimo vertinimas /
Abstract
dc:descriptionThe main purpose of this work is to empirically compare two investment portfolio optimization methods - Mean-Variance Optimization (MVO) and Conditional Value-at-Risk (CVaR) minimization - and to evaluate their advantages and disadvantages under different market conditions. The work consists of three main parts: the analysis of scientific literature, the research methodology and empirical results, and the conclusions and recommendations. The literature analysis reviewed the principles of investment portfolio construction, introduced the main quantitative optimization methods - Markowitz's mean-variance model, the Black-Litterman model and the Rockafellar and Uryasev conditional value-at-risk minimization approach - and discussed the application possibilities of machine learning and the integration of ESG criteria. The review revealed that although the MVO model is the theoretical foundation of portfolio theory, it has significant practical drawbacks, which is why alternative methods that better manage tail risk have been developed. After the literature analysis, the author carried out an empirical study using historical data of the 30 largest companies by market capitalization in the S&P 500 index for the period 2015-2024. Four optimized portfolios were constructed: MVO with Sharpe ratio maximization, MVO with standard deviation minimization, CVaR with modified Sharpe ratio maximization, and CVaR with conditional value-at-risk minimization. This dual research design made it possible to separate the effect of the optimization method from the effect of the optimization objective on the final results. Portfolio performance was tested on 2025 data, which was not known at the time of portfolio construction, and compared with each other and with an equal-weighted benchmark portfolio. The performed research revealed that the choice of optimization objective has a greater impact on the final portfolio behaviour than the choice of optimization method. Sharpe maximization strategies generated higher returns but also larger drawdowns, while risk minimization strategies produced portfolios with lower returns but approximately half the maximum drawdowns. Risk minimization strategies also demonstrated greater out-of-sample robustness, and paradoxically the highest tail risk was reached by the CVaR portfolio with the Sharpe maximization objective, even though it was constructed using the CVaR risk measure. In the conclusions and recommendations, after summarizing the main concepts from the literature review and the results of the empirical research, recommendations are provided for investors regarding the choice of optimization method depending on their investment goals and risk tolerance. The author believes that the results of the study could be useful for both individual and institutional investors when forming their investment portfolios and choosing an appropriate risk management strategy under different market conditions.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2026
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Dovydaitis, Arnedas,
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:ELABAETD308322649&prefLang=en_US