风险条件下基于收益视角的最优投资决策研究

THE OPTIMAL PORTFOLIO DECISION UNDER RISK BASED ON RETURN

给出了风险条件下基于收益视角的损失最小化投资组合模型.数值试验表明,当投资者预期收益较高时,该模型等价于条件在险价值模型,当投资者满足于较低收益时,该模型优于条件在险价值模型.提出的基于主成分分析的情景生成方法避免了过度分散投资,投资组合模型的最优目标值随情景数目的增加快速趋于收敛;该情景生成方法无需假定随机收益的分布,融合了收益分布的非对称及尖峰等统计特征,主成分分析法的降维优势使该方法也适合于资产数目较大时的情景生成.

A portfolio selection model to minimize the expected loss under risk is proposed based on the perspective of return.Numerical analysis shows that this model is equivalent to conditional value-at-risk（CVaR） model when higher target return is required but superior to CVaR with lower target return.A scenario generating approach based on the principle component analysis effectively avoids over-diversified investment and the optimal decision quickly tends to converge along with the increasing scenarios.Without assuming the distribution of the portfolio return,the scenario generating approach considers the asymmetric and lepokurtic characters of the return distribution while the advantage of the principle component analysis to decrease the dimensions makes it possible to generate scenarios when a large number of risky assets are involved in the portfolio.