Font Size: a A A

Research On Forecasting Methods Of China Stock Market

Posted on:2004-08-08Degree:DoctorType:Dissertation
Country:ChinaCandidate:G W YinFull Text:PDF
GTID:1116360122982148Subject:Systems Engineering
Abstract/Summary:PDF Full Text Request
In this dissertation, the forecasting methods of China stock market are researched systematically on the basis of wavelet analysis theory, chaotic theory and wavelet packet analysis technique. The main contents are as follows.The chaotic theory and its analysis technique are applied to the research into Shanghai Composite Index (SHCI) and Shenzhen Composite Index (SZCI) daily gain rate series of China stock market, and the evident chaotic phenomenon is confirmed to exist in the two series.Regarding the chaotic theory as foundation, the part forecasting method is applied to establish two part forecasting models separately, and forecasts of the two series are made. And the complexity of stock market is discussed further. Then, the new thinking of China stock market modeling and forecasting is advanced.The wavelet transformation theory is applied to the forecasting of the China stock market, and a wavelet-chaotic integration-based forecasting method for China stock market is advanced. Furthermore, the method of improving forecasting precision further is advanced. The concrete research of SHCI and SZCI daily gain rate series is done. And the reason of improving forecasting precision is explained from the point of view of the construction of attractors.The wavelet package transformation theory is applied to the forecasting of China stock market, and a wavelet packet-chaotic integration-based forecasting method for China stock market is advanced. The concrete research of SHCI and SZCI daily gain rate series is done. And the reason that the forecasting precision of this method is better than that of wavelet transformation alone is explained from the viewpoint of the construction of attractors.Two methods, the wavelet-chaotic integration-based forecasting method and the wavelet packet-chaotic integration-based forecasting method, are applied to the forecasting of stock market in more than one step. The two methods are applied separately to establish forecasting models for SHCI and SZCI daily gain rate series in more than one step and the concrete research is done. It is confirmed further that the chaotic phenomenon exists in China stock market. The complexity and difficulty in long-term forecasting of stock market are indicated. The time series forecasting methods for stock market are analysed and compared. The conclusion is that the wavelet-chaotic integration-based forecasting method and the wavelet packet-chaotic integration-based forecasting method are good methods to forecast stock series.
Keywords/Search Tags:stock market, forecasting, chaotic theory, wavelet transformation, wavelet package analysis technique
PDF Full Text Request
Related items