Mean-Variance Optimization: Modeling an Optimal Investment Portfolio in the U.S. Tech Sector

Abstract

Introduction. Modern Portfolio Theory (MPT) provides a quantitative framework for making informed investment decisions. The highly variable and uncertain U.S. technology sector challenges traditional investment approaches, necessitating methods that better address its unique risk-return trade-offs.
Problem Statement. Traditional investment strategies frequently fail to capture the dynamic and volatile
nature of the tech market. They rely on limited data and inefficient calculation processes, resulting in suboptimal
asset allocation. One of the advanced methods for refining portfolio formation strategies tailored to the tech market is the mean-variance optimization (MVO) method.
Purpose. To optimize mean-variance optimization (MVO) to construct optimal portfolios for the U.S. tech sector, leveraging contributions from MPT, Sharpe’s optimization techniques, and Tobin’s asset allocation model.
Materials and Methods. Historical stock data serves as the basis for implementing MVO with Python to construct portfolios that include a risk-free asset, enabling the calculation of the Capital Allocation Line (CAL) and the upper Efficient Frontier. The geometric mean evaluates expected returns, improving long-term predictability
and portfolio comparability, while daily returns enhance the model’s sensitivity.
Results. The study has demonstrated that optimized portfolios achieve higher Sharpe ratios and superior riskreturn characteristics, outperforming benchmarks through effi cient computation.
Conclusions. The MVO is an effective investment tool for the tech sector, enabling informed asset selection and
portfolio construction. This study has highlighted the importance of integrating iterative calculation processes and advanced computational techniques to adapt traditional investment strategies to the extensive data requirements of today’s markets.