Empirically Analyzing Mean-Variance Efficiency and Diversification Contradiction by Chinese Stock Markets

Based on the computational extension of portfolio optimization, we take a closer look at mean-variance efficiency and diversification contradiction, that is, high concentration in a few securities by Markowitz´s portfolio selection. Traditional treatments typically include imposing constraints. Unfortunately, the studies basically hinge on simplified portfolio selection models, focus on a few points of an efficient frontier for example global minimum variance portfolio, or proceed in small scale portfolio optimization. We set up portfolio constraints of upper bound, market value, P/E ratio, turnover ratio, and industry and find that basically the installation of portfolio constraints of upper bound has substantial effect to alleviate the mean-variance efficiency and diversification contradiction, while the effect based on portfolio constraints of market value, P/E ratio, turnover ratio, and industry is quite limited.