The Shanghai Composite Index Prediction Method And Empirical Research

Nowadays, Chinese stock market is being in the process of becoming more canonical. At the same time, its size is getting larger and larger. The number of the institution investors in the Chinese stock market who often use portfolio method is also increasing quickly. And there also emerged some financial goods relating to the stock index. So, it becomes more important than ever to do indicate work about the stock index. For the reason, the research work of this article is useful. Furthermore, either the basic analysis method, or the numerical analysis method has its advantages and disadvantages. We can get a more powerful method by combining them. Take the Shanghai Stock Exchange Comprehensive index as example index, we use both basic analysis and numerical analysis to forecast the stock index. The result of our work shows it is useful. In this article, we used some new analytical methods to do our work. Firstly, we used the Orthogonal Polynomial Distributed Lag Model to analyze the short and long relationship between the macroeconomic factors and the index. Secondly, we used the neural network to analysis the nonlinear relationship between them. Thirdly, we judged the quality of the forecasting method by both absolute quality and relative quality. Fourthly, we used the nonparametric model based on the estimating method of orthogonal series. Lastly, we used the network with two hidden layers to do combining forecast work. The result of empirical research shows that such methods get good effect.