Research On Stock Selecting Model In Style-Strategy Based On Genetic Algorithm (GA)

In nowadays, security investment is becoming a core topic with no doubt, While how to select the promising stock from numerous species is always a tough puzzle for investors. Throughout the history of security development, there were many outstanding investment methods which were famous for different stages respectively. Today, modern quantitative investment is leading investment industry to a new round of revolution. Quantitative investment has been very popular among developed countries, since the stock markets of developed countries are relatively mature and the relevant theories are fairly sophisticated, which provides certain references and experience to Chinese stock markets. The style stock selecting model established by this thesis is a branch of quantitative multi-factor stock selecting model. This thesis mainly studies the optimized style stock selecting model by generating the optimal market style through genetic algorithm (GA) method, besides this thesis has researches on return、risk、stability、selecting power of the model to evaluate the effectiveness of style strategy method.The innovation in this thesis mainly embodies in three parts:the first part is applying the weighted market style to stock selecting models; the second part is using the GA method to calculate the accurate quantity of market style; the third part is utilizing horizontal and vertical methods to process data respectively for the purpose of standardization and comparability. Through detailed theoretical derivation and rigorous mathematical calculation, we can draw conclusion as follows:the modified style stock selecting model based on GA method perform far more excellent than the traditional stock selecting model and beat the market in bear market、bull market and volatile market respectively.