Study On Multi-scale -GARCH Model Of Valatilities In The Financial Capital Gains

The purpose of modeling financial time series is to infer how the real Generation of bata or the probability rules of financial market. The information we obtain can be used to proof-test economic suppose and fiancial theory and explain important economic phenomena.This paper makes full use of multi-resolution analysis and GARCH model,and make models of financial time series with multiscale method.we make empirical Analysis by selecting out proper financial asset samples and get characteristics of waves from differert time scales.Therefore it provides evidence of financial market analysis, forecasting and supervise.The main result of study as follows:As for the problem of change-points detecting in the Garch model of the fluctuations of financial asset, we put forward change-points detecting maximum arithmetic of financial time series basing on the different characteristic between the real information and the noises.We also make the Digital Simulating on GARCH-M model, and the result proves that the model has practicability and validity.The research on Chinese market with this method proves the Chinese stock has obvious characteristic of policy market.As for the characteristic of multiscale in financial time series, we bring forward multiscale Garch model (MRA-GARCH).we get Garch model using multiscale wavelet coefficients of inancial time series, so we can forecast wavelet coefficients well.And then we rebuild the wavelet coefficients using Mallat arithmetic and get the forecasting values.This method improved the conventional optional threshold denoising processing and enhance the precision of forecasting."How to choose the stock when the fluctuation of stock expected income is not obviouse?"To answer this question, we contrive the Dynamic fuzzy clustering arithmetic of the fluctuation of stock expected income with MRA-GARCH and dynamic fuzzy clustering principle. We also simulate on the twelve stocks which have not significant differernces in Chinese market. And from the Plane Graph of the future 21 days, we can observe the best trade scheme.The result of this paper is applied to investment analysis well.