基于波動(dòng)源模型的我國(guó)股市股價(jià)指數(shù)波動(dòng)研究.pdf

1、This thesis takes a close and deep look at volatility of China's Stock Market Price. Based on the study of China's Stock Price volatility, this thesis analyses the cointegration between macroeconomic varibles and the ind

2、ex of Shanghai stock market. It has certain theoretical meaning and realistic meaning. Firstly, this thesis starts with probeing into characteristic of China's stock pricevolatility. And according to the range of the

3、 share index return,the quantitative definition of abnormal fluctuation was proposed. Secondly, theoretical models on the stock price are introduced, including random walk model, lognormality model and fluctuating source

4、s model. GARCH and EGARCH class of models for time series and the method of parameter estimate are introduced. Then, the models of GARCH (1,1)and EGARCH ( 1, 1 ) are used to test the volatility of the Shanghai stock log

5、arithmic return time series .At last, In chapter 5,this section study the cointegration between the index of Shanghai stock market and 4 macroeconomic factors by using co-integrate technology. Our empirical results a

6、re as follows: Firstly, the phenomena of thick tails,volatility clustering, leverage effects, are exist in Shanghai stock market. Bad news affectiong is higher than good news;Secondly, It is found from the comparison tha

7、t the ARCH type models with student't innovation is more capable to capture charateristics of logarithmic return time series ;Thirdly, Fluctuating sources model can describe the stock market more accurately than the tra

8、ditional models. Fourthly, the composite index of Shanghai stock market has a long run balanced relationship with the 4macroeconomic varibles, such as gross domestic product,money supply, inflation rate,and interest,thro

9、ugh which a long term trend of the composite index of Shanghai stock market can be forecasted. Fifthly, the results of the estimation of error correction model shows that the 4 macroeconomic varibles have not remarkable