Stock Selection Methods In The China Market

Successful investment, from the fundamentals, it means for an accurate valuation of listed companies; from technical terms, it means we must thoroughly understand market trends. The value of investments and trends in the practice of long-term investment transactions are being proved an effective manner. In recent years, China's stock market performance is more proof of the concept of value investing trends. In this paper, we build A-share stock selection models, based on the value investment philosophy, and simulated stock data for the period of 2004-9-30 to 2009-10-30.Firstly we refer to Canadian Royal Bank's index system of their stock selection models, combined with the actual adjustment of China's stock market, using genetic algorithms to build A-share stock selection model--model 1.Considering stock data sequences are non-stationary time series with irrational noise, we proposed model 2. We firstly pre-process data with SSNF wavelet algorithm to make data more stable, and then use the model 1.Validity of the models are decided by comparing the relative excess positive earnings and negative returns compared to HS300 Index in the event of frequency, accumulated earnings, the mean and the most value (refer to absolute value).Through the experiment, the model 1 has a frequency of 59.68% when it gets the relative excess returns; the cumulative excess return is 105.12%. Model 2 improves the results, respectively to 64.52% and 106.99%. Model 2 reduce the frequency, mean and minimum values (absolute value) when it has a negative excess return, with a better ability to resist risks; increased the frequency and maximum values when it has a positive excess return, with a better ability to obtain positive excess returns.The experiment results show that our A-share stock selection model is Valid.