Research On Quantitative Investment Strategy Based On CSI 300 Stocks

Quantitative investment strategy is the sum of investment methods that use accurate quantitative methods and evaluate with actual backtesting.With the scale and diversity of financial data collected in mathematical theory and practice,statistical models and financials are further expanded.The combination of big data has become a trend.In western developed countries,the practice of quantitative investment has been in existence for more than 30 years,and China has only started in this field in recent years.Therefore,through the use of quantitative investment strategies,China’s A-share market has matured in recent years.The study of CSI 300 Index constituent stocks as the research target to optimize investment strategies is rich in meaning.The quantitative investment strategy studied in this paper uses the factor screening method of machine learning,which is widely used in financial mathematics today.The Support Vector Classifier(SVC)in this method is a machine learning method that can minimize the risk of error and the generalization risk of the model.It can be more reasonably adapted to the applicability of the non-mass financial data of this study.In the evaluation and analysis of strategy,this paper believes that an effective quantitative investment strategy can exert its intrinsic value in actual securities investment.However,a scientific method cannot guarantee the correctness under all conditions.Therefore,this paper has effectively defined the meaning of investment strategy and conducted different evaluations in the bull market of the securities market to enhance the clarity and research results of the research direction.Intelligibility.The research in this paper uses the excellent mine quantitative platform and the rice basket quantitative back-test platform to compile the strategy and back test the final result.The research process is mainly as follows: firstly,collect the yield data and multifactor data of the Shanghai and Shenzhen 300 constituent stocks and Perform numerical conversion,data cleaning and extreme value processing and put them into the data frame.Secondly,after using the SVC machine learning method and improving the kernel function,the effective factor selection process of the CSI 300 stock pool is used to obtain the three prominent stock selection factors of PEG,PE and operating profit growth rate.The IC value determines the final strategic stock selection factor.Finally,after adjusting the portfolio size,trading model and risk control strategy of the portfolio strategy,this paper obtains the overall portfolio strategy that can significantly outperform the benchmark portfolio,and utilizes the decomposition of the sector,revenue and risk in the adjusted strategy.Brinson attribution analysis.In the conclusion part,the paper puts forward two optimal conditions and four most unsuitable conditions for the actual application of the strategy.