| Stock market is a product of highly developed market economy, the total market value of China's stock market has risen from 1048 billion yuan in 1992 to 20.2039 trillion yuan in July 2010.Establishment and development of the stock market played a positive role in the economic development of the market, giving a powerful impetus to further development of China's economic reform.Among the world stock markets, the bull market in Shanghai and Shenzhen stock market in October 2007 has surpassed the total market value of Nasdaq Stock Exchange as the world's second-largest stock market. Study on the relationship between the Chinese stock market and macroeconomic become great significant. However, unlike in mature capital market, in developing countries, the link between the two is very weak, lack of statistical significance.China's stock market, with its high Volatility history, occurred at some stage, showed a greater statistical variance, while in other relatively stable phase, showing a smaller variance. This shows that the variance of broader market gains changes over time, which means time-varying volatility. By the volatility clustering, the price fluctuations within a period of time has specific similarities, and different periods of volatility rules are different. The results of short-term data is not representative; while time scales of the stock market data and macroeconomic data varying, the stock market data exists heteroscedasticity; Jenson-Shannon divergence based on Entropy could composite Shanghai Stock Index in April 1992 to March 2010 into five stages by measuring the volatility of Shanghai Stock Index. In this paper, the main contents and conclusions are:The first part reviewed existing literature on the domestic research. Foreign mature capital markets reflect macroeconomic fundamentals of operation, and the development of China's stock market imperfections caused by the stock market and economic operation deviates from serious. In particular, under the sample period and indicators, indexes, and out of coordination between economic performances, even some degree of predictability of economic operation, but does not have a promotion and universal.The second part analysis of J-S programming information entropy theory and application of algorithms, J-S got the stock market data in China 1992-2009 J-S entropy segmentation results. Entropy algorithm in accordance with the entire range has been divided into five small fragments of the results, these five fragments and the corresponding compare the macroeconomic situation, basically the stock market and economic departure from the current results, and each stage Stock Market described in more detail.The third part contains the variable selection and the empirical method of choice. In many of the macroeconomic variables reflect the availability and openness in line with the principle of selection on behalf of the six major aspects of the macroeconomic indicators of the monthly data of seven, but did not pass multicollinearity test, using backward stepwise regression France retained only the most representative of the added value of gross national product, consumption levels and on behalf of the consumer price index inflation, the opportunity cost of holding money on behalf of interest rates, inflation levels representative of the broad money supply is explained on the composition of Card KLCI independent variables vector. These variables were seasonally adjusted and smoothed index.PartⅣempirical results and analysis. From the theoretical analysis of the relationship between them and the use of vector auto-regression, Johansen Cointegration Test, Granger causality test, vector error correction model methods of an empirical analysis. The results show that Granger causality test of the basic judgments under different lag all variables and VAR causality conditions established endogenous and exogenous variables. VAR model, been in the index and economic variables such as the consumer price index, interest rates, lagged one month on the interaction in the relationship.. The unit root test are all variables are integrated of order one sequence, with the Johansen test of my stock index in the long term is consistent with the macroeconomic development, the existence of a long-term equilibrium relationship. However, in some stage time, there is no cointegration relationship between the two. In a January 1997 to December 1999 as a fragment and in January 2000 to March 2006, as a fragment consisting of the first sample, established by judging the internal and external health variable VAR model, the Shanghai index and macro-economic variables, no long-term equilibrium relationship, but the inter-stage VAR model results than the ideal impulse response function can explain the stock market and the economy deviate from delay and causality; the second is from 2000 to 2006 and 2006, a fragment By 2008 the composition of the second sample, as is the smooth sequence, only the establishment of VAR model, been three high goodness of fit equation, the first phase of the industrial added value can explain the volatility of the stock market next performance, the contrary is also true.In summary, this paper innovations are:1, for the first time biomedical common Jenson-Shannon entropy segmentation algorithm is used in stock market data, obtained in accordance with the fluctuation variance is divided into 5 different fragments, and fragments of the border sound; 2, for the full sample and the phases were considered, especially in stages during the study selected two typical deviation from the stock market and economic fluctuations causal role as cross-phase made the application through the introduction of a larger lag VAR model validation; 3, between the regression model with stage shows that the previous phase of economic prosperity and a booming stock market to become the next phase of the reasons that we should maintain the stock market long period, not only the current economic performance, but the light of the currency investment channel and Considering the potential economic strength. |
PDF Full Text Request