Research On Risk Diversification Strategies Of Multiple-Factor Quantitative Investment

With the development of China's stock market,investment information,price information and so on are exponential growth,so how to sort out those stocks with investment value in massive amounts of information and disperse portfolio risk reasonably are hot issues in investment research field.Multi-factor quantitative investment strategy has become one of the investment strategies widely recognized and used by its effectiveness and stability.However,due to the diversity of factors,investors often have difficulties in determining the effective factors,and their way of combination of factors is also too rough.In addition,the focus of the traditional investment strategy is to achieve high excess returns,for the portfolio risk,lack of sufficient attention and the way is simple.This thesis discusses the question of selecting effective factors in Chinese stock market and the multi-factor quantitative stock selection and portfolio optimization.This thesis utilizes stock data of the Hushen300 Index constituent stocks from January 2006 to December 2015 as the study sample.At first,according to the criterion of factor effectiveness,seven effective factors are selected from the fifteen candidate factors.Then using the entropy TOPSIS method to synthesize these effective factors to construct multi-factor quantitative stock selection model,and this model is compared with the single factor model and multi factor stock selection model based on factor scoring method.Finally,based on the TOPSIS multi-factor quantitative stock selection model,this thesis put the TOPSIS value of the stock into the mean-CVaR model to construct portfolio optimization model,and compares the performance of the optimized portfolio with the equal weight of the portfolio to evaluate the effectiveness of the portfolio optimization model.The empirical results show that the multi-factor quantitative stock selection model based on TOPSIS has higher stock picking ability than the single factor model and the multi factor stock selection model based on factor scoring method,also can obtain significant excess returns.On this basis,the portfolio optimization model can make the weight distribution of portfolio more reasonable,the return and risk-benefit ratio of portfolio have been improved,the asset allocation have been further optimized.This thesis uses TOPSIS method to integrate effective factors of stock market to select stocks with investment value,and then uses the mean-CVaR model to optimize the portfolio to reduce risk.This first choose re-optimization investment strategy optimization not only achieve significant excess returns,but also significantly reduces the risk of portfolio,as well as provide support for the theory and method of effective capital market investment decision-making and risk management.