Analysis Of Shanghai Index Fluctuation Based On EEMD

After20years of development, China’s stock market has become a crucial part ofChina’s economy. The healthy development and the standardization of the stock marketis the important foundation for the steady development of our country’s economy. Manyscholars in the economics circles believe thatthere is close relationship between theprice volatility of the stock market and macro economy.There is necessity to haveanembedded theoretical research and empirical test of therelationship between economyand stock market, which is advantageous to the standardization anddevelopment ofthecapital market, help improve and maturate thestock market, has a very important rolein promoting the socialistreform and the development of economy.Therefore, in recentyears,the economic and financial scholars give more and more concern about therelationship between macroeconomicvolatility and price volatility of the stock market.Member of American engineering Dr.Huang put forward a new method of signaldecomposition which is called Empirical Mode Decomposition in1998. The essence ofthis method isto obtain eigen wave patterns through the time scale of data and decom-posethe original data.Empirical Mode Decomposition(EMD)is a time-frequencyanalysis method to process non-1inear and non-stationary signals．According to thecharacteristics of input signals, this method can decompose the original signal to a sumof some intrinsic mode functions adaptively without any prior knowledge．It isconsidered to be a great breakthrough of traditional time-frequency analysis methodssuch asFourier analysis and wavelet analysis which are based on the hypothesis of linearand stationary．In the course of development after many years of EMD method，itgradually shows the unique advantage in processing non-stationary signals．EMDmethod not only has the important theoretical research value but also haswideapplication prospect.EEMD (Ensemble EMD) is put forward to overcome thedefect of the frequent appearance of mode mixing in2005by addingwhite noise whichmay help extract the true signal in the data.In this paper, we firstly use the EEMD method to decompose the price of theShanghai index (SHI) sequence, and evaluate the IMFs by average period, thecontribution of variance and Pearson correlation coefficient. Further, the IMFs areclassified into three components which named high frequency, low frequency and thetrend. And we give them appropriateeconomicinterpretations. The trend componentcan represent a long-running trend of Shanghai stock indexwhich is rising, buthas sloweddown now in our investigation period. And, the low frequency component reflectslong-term volatility of the Shanghai indexwhich can be considered as thereflect ofmacro-economy. It is not difficult to observe thatIMF6is very close to the shape of theCPI and IMF5to the added value of industrial. The average cycle of high frequencycomponent is short, so it represents the short-term volatility of Shanghai stock index. Itmainly influenced by the factors such as investor psychology, short-term stimulus andvolatility spillover effect of other markets. Then we set up corresponding models toanalyzehigh and low frequency component.In this paper, the low-frequency component which represents SHI’s the volatility ofthe medium and long termextracted separately. Then the VAR（3）model is built toevaluatelong-term equilibrium relationship between CPI, the added value of industrial,money supply (M1) and the low-frequency component. We come to the conclusion thatinflation is negatively related to the low frequency component and industrial addedvalue is positive through the cointegration test and variance decomposition. It meansindustrial added value growth stimulates the medium and long term performance of theShanghai composite index but inflation contains the upward trend.Finally, we inspectthe correlation of high frequency fluctuations between SHI and other stock indexes. Wefind that the high frequency wave of SHI has a very weak correlation to the Hang Sengindex (HSI) and the Dow Jones industrial average (DJIA) in the first stage (1997-2006)by Granger causality analysis. But there is a significant correlation between SHI andDJIA in the secondstage. The main reason lies in the exchange rate reform and theopenness of capital accountin our country. Moreover, the impact of capital market inAmerica contaminate to others is more easily because of the subprime crisis.In conclusion, we use the EEMD method combined with traditional measuringmethod to study the influence factors of different frequency fluctuations of SHI in thispaper.It helps tounderstand the influence of different factors comprehensively anddeeply. And, it may alsoprovide advice toinvestorsin the stock market to pursue the maxprofits.