存在高阶矩风险偏好条件下,组合投资选择必须考虑最大化收益、偏度和最小化方差、峰度四个相互冲突的目标。同时,考虑到高阶矩风险的时变特征,应建立高阶矩动态组合投资模型。基于多目标优化技术和效用理论,讨论了高阶矩动态组合投资模型的构建,并利用MATLAB软件中的非线性优化函数fmincon对模型进行了求解,从理论和实证两个层面对两类模型进行对比。
Higher moments risk cannot be neglected in portfolio analysis. The optimal portfolio must address the trade-off between conflicting objectives, such as maximizing the expected rate of return and positive skewness, and simultaneously minimizing the variance and kurtosis. Furthermore, since the higher moments risk is time-variant, it is necessary to consider a dynamic portfolio model. Based on multi-objective programming technique and utility function theory, this paper establishes the dynamic portfolio model under higher moments and solves it using a nonlinear programming function "fmincon" in MATLAB. We also compare the proposed two models from the view of both theory and empiric.