Lellep, Jaan, juhendajaKantšukov, Mark, juhendajaSetihe, OmarTartu Ülikool. Loodus- ja täppisteaduste valdkondTartu Ülikool. Matemaatika ja statistika instituut2020-09-022020-09-022020http://hdl.handle.net/10062/69403The objective of the thesis is to use optimal control theory in order to optimize portfolios. More precisely, using principles from calculus of variation in order to define the portfolio problem with reasonable constraints to maximize the profit while minimizing the risk or vice versa. Theoretical cases would be solved with simple constrains, and real application part would be made in Tallinn stock market. The latter is still in development with sixteen companies listed, fourteen which are taken in the analysis. The Values at Risk (VaR) method was the most successful in generating profit but really affected by the randomness of the solution and the nature of the market. The most stable method was the Conditional Values at Risk (CVaR) growing the portfolio slowly but surely. The whole market seems to be suffering from the COVID-19 pandemic resulting in an sharp drop in the stocks making the future returns negative.engopenAccessAttribution-NonCommercial-NoDerivatives 4.0 InternationalEuler’s equationstochastic optimal controlMarkowitz portfolioValue at Risk (VaR) modelConditional Value at Risk (CVaR) modelAuto Regressive Integrated Moving Average (ARIMA) modelEuleri võrrandstohhastiline optimaalne kontrollARIMA mudeltinglik VaR mudelVaR mudelMarkowitzi portfooliodynamic programmingdünaamiline programmeerimineinfo:eu-repo/semantics/masterThesis