Portfolio Optimization Based On Generalized ES

The measure of ES is one of the modern method of risk measure. It first proposed by Artzne, etc.The generalized ES is much more extensive. It consider both the cases of continuous and discrete of the Loss function. That is to say, when calculating ES, it's not must to consider if the loss function is continuous or not. The focus of this article is to do Portfolio optimization based on the generalized ES. Thinking about the applications of generalized ES is more wide, it is called ES in this article excepted it need to be distincted in some special situation. The main work in this article as follows.First, we introduced the risk of portfolio. It included the background and development of the portfolio.Second, we introduced the research of ES, including its backgroung, definition ,the characteristics and the applications.Third, we established the optimization Model of ES and did empirical analysis, receiving the optimum portfolio when giving the earnings.At last, we compared the results calculated by mean-GES with the results calculated by mean-variance.In this paper, the innovations are: the use of ES is a more generalized approach, which applies to more normal circumstances as a basis for the establishment of the planning model. Due to the complexity of the original formula, we derived the formula, when done for a number of conversion, making the formula to deal with more user-friendly planning. When doing empirical analysis, selected four funds on a certain period of time as the sample. Using the mean-GES model to return the portfolio in the case of giving the expected earning.