基于新型负指数期望效用的投资组合选择模型

Portfolio selection models with new negative exponential expected utilities

结合现有文献对最优投资决策问题的讨论,提出了一类满足单调性和凹性的新型负指数效用函数,并给出数学和经济学上的合理解释.通过多种类型的加权函数以及对尾部的恰当描述,损失分布的厚尾现象得到更加有效地控制.利用L-统计量估计新型期望效用,并说明其合理性.进一步地,构建了兼顾多种市场摩擦因素的实际投资组合选择模型.选用中国和美国股票市场的数据进行实证研究.结果表明了新期望效用的优越性和鲁棒性.

With an overview of the literatures on discussions about optimal investment decision problems,we propose a new class of negative exponential utility function satisfying the monotonicity and concavity.And reasonable explanations are also given from the viewpoints of mathematics and economics.The fat-tail phenomenon of the loss distribution is controlled efficiently by different kinds of weighted function and proper description to the tail.We use L-statistics to estimate the new expected utility and the rationality is also illustrated.Moreover,we construct a realistic portfolio selection model with multiple market frictions.With the real data from Chinese and American stock market,we carry out a series of empirical studies.Empirical results show the superiority and robustness of our new expected utility.