SSE 180 Index Enhancement Strategy Study

Since twenty-first Century,the proportion of trading in the financial market has been expanding.In 2017,the share of capital transactions in the US capital market has reached over 70%.Quantitative finance has gradually become an important research direction in finance.Market style switching is often encountered in the transactions of the financial market.For example,in the bull market and the bear market,people understand the behavior of the market in different ways.To grasp the market characteristics of different periods has become an important issue that investors pay attention to.From the perspective of data science,combined with the characteristics of micro transaction in financial market,the basics data of financial market are processed by feature engineering,and intermediate variables are extracted.The Gauss hidden Markov model is used to fit the characteristic variables that reflect the basic data of the financial market,and the hidden characteristics of the market transaction extracted.The corresponding quantitative timing strategy is designed according to the hidden characteristics of the transaction.On this basis,more than 40 characteristic variables are extracted by processing the characteristic engineering of the market basic data which affect the stock and fall.The GCForest cascade depth tree model is used to fit the prediction of Shanghai Stock Exchange 180 constituent stocks.The prediction accuracy of the stock model is generally over 65%,and the prediction accuracy of the overall model is 71.13%.In this paper,we first propose the GCForest cascade depth tree model to design the corresponding quantitative stock selection strategy,and the above mentioned quantitative timing strategy combined with market transaction feature extraction is used to do the simulation transaction return test.