Quantitative Investment Strategy For Stock Index Futures Based On Weighted Money Flow Model

The development of the stock index futures market is an important part of the improvement of a country’s capital market structure,and it is of great significance for building a healthy,comprehensive and multi-level financial market system.It is conducive to economic health and safe development,and has a positive effect on preventing and resolving major systemic financial risks.With the process of reform and innovation of China’s financial markets,China’s stock index futures market has just started,for stock index futures,there are still a lot of criticism,but also related systems is constantly improving.However,due to the leverage of the stock index futures market means that investors face greater risks when making investment in stock index futures.The key to investing in stock index futures is to be able to predict the trend of the future returns of stock index futures effectively.For the analysis and research of financial data,traditional forecasting models often need to be based on a series of more stringent assumptions.When faced with the nonlinear characteristics of the capital market,it is often difficult to achieve satisfactory results.However,with the widespread application of computer algorithms in different fields,the application of computer algorithms to the analysis and modeling of financial data has become an important research direction in the financial field.Because machine learning has excellent approximation ability for non-linear data,the investment strategy of the capital market has been further enriched,and machine learning has therefore been favored by quantitative investment research.Quantitative investment in foreign financial fields has matured after decades of development.Nowadays,the quantitative investment field is an important direction of investment methods and academic research.Its application in China’s capital market is gradually developing.With the rapid development of computer algorithms,today’s major financial institutions continue to apply technical methods in the fields of machine learning and data mining to quantitative transactions.This new technology and method also plays an increasingly important role in investment analysis.In recent years,with the acceleration of the Internet,China has entered the era of 5G communications.The use of computers for algorithmic trading has unique advantages in the management of human subjective emotions and the choice of trading timing.Therefore,in the current high-frequency environment,by designing quantitative trading strategies by acquiring high-frequency minute-level data,the computer language can control the subjective behavior of traders,and automatically issue trading instructions,which can effectively improve data acquisition and processing capabilities.At the same time,the best time for trading is effectively grasped,and the ability to accumulate more stable income through the increase of transaction frequency and the ability to control risks have also been improved.This article combines the development history and operating status of China’s stock index futures market,and aims to establish a money flow indicator for the stock index futures market,objectively reflect the market operation,and provide a reference for investment decisions.To promote the more efficient operation of the stock index futures market and make it more perfect for the development of China’s market economy.First of all,due to the long listing time and the large number of data samples,this article selects the main CSI 300 stock index futures contract as the representative of the stock index futures market to conduct the following money flow and investment strategy construction.Due to the short selling in the futures market,this paper finds that the calculation formula of the existing money flow is the same as the calculation method in the stock market.The factor of interests is not included in the calculation.The change in interests can reflect the shorting characteristics of the futures market.Therefore,in this paper,the interests are included in the calculation of money flow,and the weighted effect of changes in price and interests is also considered,and a formula for money flow is established comprehensively.And compared with the money flow calculated based on price changes only,the money flow constructed in this article can more objectively reflect the development of the stock index futures market,and avoid the trend errors of the money flow calculated based on price changes.Then,for the purpose of constructing an investment strategy through money flow indicator in this article,this article uses the VAR model to analyze the return rate and money flow of the stock index futures market.It has a significant positive impact.At the same time,a Granger causality test was performed to further verify the above conclusions.Finally,the impulse response analysis was used to reflect the above conclusions from an intuitive perspective.Next,this paper uses logistic regression,decision tree,random forest,gradientboosting decision tree,naive Bayes,support vector machine and other models commonly used in machine learning to predict revenue trends,and uses the grid search method for different models.The hyper-parameters are adjusted in order to get more accurate results.Later,in the process of strategy construction,in addition to applying the above models individually,this paper also combines five groups of models into a comprehensive model,and selects the optimal model to construct strategy.Different from the previous evaluation methods,this article divides the model’s evaluation index by focusing on the overall accuracy,focusing on the precision of positive examples and the specificity of negative examples.Through comparison,it is found that the random forest model has a better prediction effect on negative examples,and the comprehensive model has a better prediction effect on positive examples.Therefore,these two models are selected.Finally,the tested comparison between the constructed strategy and the ordinary moving average strategy shows that the strategy constructed in this paper has obtained a cumulative return of 45.97%,which is far better than the-12.69% yield of the moving average strategy.At the same time,it is better than the SCI-6.42% cumulative return,and the CSI-3.37% cumulative return,showing the good profitability of this strategy.It shows the good profitability of this strategy.