Study On Quantifying Of The Selection Of Funds

Since the first raised funds was set up in March1998,and ended March13,2012, the number of funds in China has reached1119. The large number of fund products brought a huge amount of market data, which made it difficult for the individual fund investors to analyze the information,resulting in demand by professional investment institutions on behalf of the Fund’s investments.However, it is hard for both individual and institutional investors to do a detailed analysis of each fund, and forecast its future performance, thus making investment decisions time-consuming. How to transfer the huge amount of market data into useful information to guide the selection of the fund becomes one of the challenges faced by many personal and professional investment institutions.Although many data providers and fund rating agencies have opened fund performance evaluation or the rating formula, and some seller Institute also have developed a quantitative selection method,it did have some flaws in the data acquisition cost, objectivity,etc.In order to overcome these challenges,this essay was to build one model, which was based on open market, simple logistic, easy to achieve.The main contribution and innovation was: Firstly,we analyzed products revenue sources, made objective position of the role of quantifying selection of funds in the FOF investment decisions and its expectation.Because we wanted to provide this paper to actively managed funds investors, so we started from the perspective of the most typical fund investment entities-FOF (Fund of Fund), analyzing the sources of income of FOF products. Compared to inferior FOF, we found that the superior ones might have a better performance,whether in bull or bear market.Further more,the asset allocation strategy was more in line with the investment logic, and was in relatively fixed positions in actively managed funds investment, maintaining at about35%. Therefore, this article was subject to provide quantitative recommendations for the selection of actively managed fund of35%of the fixed positions of FOF products. Of course, the conclusions obtained for individual fund investors were also applicable. More specific, this article sought to quantify the selection-based model,which was able to build a portfolio that had a yield ranked in the top35%of its investment target.Secondly, we built a quantifying selection model, which constructed quantitative indicators of innovative management capabilities for the fund company (investment and capacity), abandonning the traditional qualitative analysis and subjective scoring rules.Taken income characteristics and risk characteristics as the basic framework, we included fund company management capacity into this model. Based on"Momentum Effect (Momentum Effect)",we used those three key indicators to predict the Fund’s future operating performance, and thus achieved good results.Thirdly,we analyzed the simulation investment result of model,and reviewed the different effect under variant market environments, in order to make the the model user more aware of the use of model notices and risks that may arise.Based on the comparation of the investment results and reference standard, we determined the validity of the model and stability, not only achieved the expected results of the constructed model, but also made the charateristics in different period consistent. Unfortunately, the investment performance of the model in the environment of the great bull market did not reach the preset target, yet we had their causes analyzed.Forthly,compared with the real market compared, a model showed an available value.We evaluated the pratical value of the model after the contrast of the simulation investment results and the real FOF products performance.Fithly, we made a correction program for the defect of the model during big bull, and discussed ways to improve.The paper proposed two amendments to the program, and discussed the feasibility. In addition, we thought up a way to improve the model,which was to add time-seleting indicators.Model users could choose to use the modified program or to develop the model.