Quantitative Investment Product Design Based On Logistic Regression Stock Picking And Stock Index Volatility Timing

Quantitative investment has attracted widespread investor attention due to its extraordinary performance.Over the past 40 years,it has subverted the traditional investment philosophy and has been hailed as "the revolution in the investment community." The author tried to introduce machine learning methods on the basis of predecessors,add some variables and indicators,and try to find a strategy that can overcome the market and defeat some quantitative products in the market.After continuous trials,this paper designed a quantitative investment product based on the timing of stock selection and stock index volatility.In the stock selection strategy,first classify the data of the constituent stocks of the Shanghai and Shenzhen 300 Index into five categories: Quality Factor,Growth Factor,Technical Indicator,Momentum Factor,and Emotion Factor,and then rank according to the rate of return.The first 30% or the last 30% sets the dependent variables of the two categories.Finally,through the logistic regression model,stocks are selected.In the timing strategy,when the broader market style is attributed to "inversion" and the smoothed one-way volatility is negative,the stock is held;when the broader market style is attributed to "trend",and the smoothed one-way volatility is positive At that time,holding the stock;when the broader market style attributed to "shock",also choose to hold the stock.In all other cases,the stock is sold.Through empirical research,the portfolio products constructed through logistic regression design choices of stock selection and stock index volatility are significantly better than the Shanghai and Shenzhen 300 Index both in terms of cumulative yield and annualized yield,demonstrating the strategy The validity of the product and the excellence of the product confirm the value of this article.From the perspective of cumulative returns,the cumulative yield of portfolio products constructed in this paper was 386.6%,and the cumulative return rate of the CSI 300 Index over the same period was 59.6%.During the period of backtesting,the annual backtest yields were all positive,and all of them surpassed the CSI 300 index gains except for 2017.In terms of risk control,this paper designs innovative time-series indicators based on the volatility of the Shanghai-Shenzhen 300 Index.Through empirical evidence,it can be found that when extreme market conditions are encountered,it can effectively signal,reduce investor losses,and control maximum back-testing.At the same time,products that can be found through Brinson analysis have strong stock selection ability in the sector.Therefore,we can significantly improve the product's profitability by studying industry asset allocation.Through the regression analysis of the net value,we can see that the product style of this article has not changed during the six-year period.Maintaining investment in high-value and high-liquidity stocks can be seen in this paper.Quantitative products have a certain degree of practicality.