Abstract:

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.

Key words: negative exponential expected utility, market frictions, portfolio optimization, performance ratio