A New Method For Analysing Funds Performance

Mutual fund is the new kind of investing instrument in our country's financial market. It is very popular in western financial market and it had over 100 years' history. It became a very important financial instrument in financial market. As the scale and classification is more complicated, researching, analyzing and evaluating the mutual fund performance is becoming more and more important. Because the investment strategy and operation method are different between different funds, their performance and risk are also different. So it is necessary to evaluate the actual result of the mutual fund investment operation. How to evaluate the funds performance is always be an important problem in modern financial theories. From the early Markowitz portfolio theory (1952), the CAPM theory of Sharp to Treynor Index, Jensen Index, and also the latest APT model of Ross (1978), we can all see that fund performance evaluation is a special application of the pricing risk asset. In this paper, we first classify and summarize the performance evaluation methods, and then pay more attention to risk adjusted fund performance evaluation method which was based on Markowitz portfolio theory and market timing and stock picking performance evaluation method which reflects the investment manager's investment ability. Secondly, we introduce the new method—Stochastic Discount Factor (SDF) performance evaluation method. The assumption of this method is no arbitrage; the price of any asset is the expectation value of future return discounted by Stochastic Discount Factors. This paper choose return data of new types of ETF and LOF funds which are lasting for more than 2 years, and use Stochastic Discount Factors as variables to explain the information which the investment managers use, and then find out the performance index under condition, and compare the method with other classic performance evaluation methods. On one hand it demonstrats that it is a good method to evaluate fund performance, on the other hand, this model still has potential on information variables and primitive assets choosing.