Advanced Portfolio Optimization Using Copula and GARCH Models: A High-Performance Computing Approach for the European Stock Market

This paper explores advanced portfolio optimization strategies using copula and GARCH models to enhance risk management and profitability in the European stock market. By utilizing high-performance computing (HPC) to conduct extensive simulations on approximately 10,000 portfolios, the study compares the effectiveness of various copula-GARCH models against traditional approaches, such as the mean-variance model. The most effective configuration-identified as a Student’s copula with marginal Student’s distribution and an eGARCH model-was employed to simulate returns and construct optimal portfolios that minimize Conditional Value at Risk (CVaR). The scalability and robustness of this approach offer valuable insights into its practical applications for portfolio management.