Font Size: a A A

Nonlinear Analysis And Forecasting In China Stock Market

Posted on:2007-01-04Degree:DoctorType:Dissertation
Country:ChinaCandidate:W L HanFull Text:PDF
GTID:1119360212467709Subject:Systems Engineering
Abstract/Summary:PDF Full Text Request
With the rapid development of nonlinear theory and tool, many fields have been deeply affected. One of them is stock market. The study about stock market in nonlinear theory include fractal market, chaos market, forecasting stock index using nonlinear tool. There have been much literatures about fractal market, chaos detection as well as forecasting stock index using nonlinear toll in China. However, the results of those fields are different because the experts study them in various methods. In the discourse nonlinear theory and nonlinear tool are used to analyze Chinese stock market and forecasting the change direction of stock index. There are four sections in the discourse. First section is the theoretic platform of whole research and actual base. In the part summarizing and analyzing the main literatures of stock market using nonlinear theory and describe nonlinear characteristic of Chinese stock market. It is showed that some problems of research in the field, such as some improper methods are used to test chaos in financial time series. The second section is the core of the discourse. Depending on summarizing and observing the main method of chaos detection in financial time series, the discourse point out the view that chaos detection in Chinese stock market should apply the method basing on small size time series. In addition to the discourse introduce a new method for chaos detection: Gencay-Dechert method and test chaos in Chinese stock market in Gencay-Dechert method. The results of detection are that there isn't chaos in both shanghai stock market and Shenzhen stock market. The third section is another core of research: forecasting stock index in China. The discourse forecast stock index using Probabilistic Neural Networks. It is showed that the method of PNN is feasible. The last section of the discourse is summary as well as put forward weak point of research. It is showed that research of stock market with Gencay-Dechert method can be done further step if using some statistics, in addition to, the discourse point out a view that as the development of related basic research, the accuracy of...
Keywords/Search Tags:nonlinear, time series, chaos detection, stock market, prediction, Lyapunov exponent, probabilistic neural network
PDF Full Text Request
Related items