An Empirically Study On The Performance Of Chinese Open-End Equity Funds

This paper is an empirical research on the performance of Chinese open-end equity funds. The purpose of this paper is to provide a reference for selecting the most suitable evaluation model for open-end funds. Twenty-eight equity funds that established during July 1, 2004 to September 5, 2005 have been selected for the empirical research. Empirical studies include fund performance measurement and fund performance attribution decomposition.Through fund performance measurement, This paper found that none of the fund's investment is the full diversification portfolio. The performance of 28 funds are ranked by the use of Sharpe Index and Sortino Ratio, then the rank order is tested by Spearman rank correlation coefficient tests, we found that the order of the two index is highly consistency.Each fund's Sharp Index is compared with the corresponding Sharpe Index of the market portfolio, the paper found that a total of 14 funds have done outperform the market portfolio after risk-adjusted return.In this paper, we use timing model mainly for the research of the fund performance due to decomposition. The main models used in this paper includes CAPM-TM, CAPM-HM, CAPM-GII, FF-TM, FF-HM and FF-GII six models, and we use daily, weekly and monthly data as the evaluation cycle. In order to detect the potential serial correlation, we use DW statistics to judge and then use Durbin Two-step method to solve the relevant issues.The analysis based on a total of 448 results after regression demonstrated that, on the whole, 28 funds has shown positive timing ability when using daily data, and has shown negative timing ability when using weekly and monthly data. In addition, the percentage that the timing coefficient can pass the significant test on the 5 percent level is below 50 percent, indicating that a considerable number of funds do not have the timing ability.Under the premise of the weak form efficient of Chinese stock market and the principle that maximum coefficient of determination, We found that, CAPM-HM model is more suitable evaluation model in timing research for the current condition, and the daily data is the best for the evaluation cycle after solving the relevant time serial correlation.