| Abstract [eng] |
The aim of this MBA thesis is to develop and assess an optimization model for investment portfolios that integrate psychological risk tolerance evaluation, behavioral bias mitigation, and CVaR-based decision-making automation. The paper addresses the gap between classical optimization models, which often ignore behavioral aspects and extreme risk exposure, and the real needs of individual investors operating under uncertainty and emotional pressure. The research applies a methodological framework grounded in behavioral finance theory, conditional value-at-risk (CVaR) modeling, and Monte Carlo simulation. The model incorporates risk tolerance assessment based on the Grable and Lytton (1999) test and includes a dynamic hedging mechanism using S&P 500 futures to ensure portfolio compliance with defined CVaR limits. Optimization is performed using linear programming techniques based on Rockafellar and Uryasev’s (2000) CVaR minimization approach. The practical part presents a simulation experiment where real historical data from Yahoo Finance is used to demonstrate the model's functionality. The results show that the proposed model effectively aligns portfolio structure with psychological risk constraints, while selective hedging further stabilizes CVaR in adverse market scenarios. The developed system is implemented as a web-based application using NestJS, Next.js and MySQL. |
