| With the deepening of economic globalization, China’s also standing while internationaleconomic turmoil, China’s capital in the proportion of the world economy is more and morehigh, and has become an important and integral part of the world economy, and the status quoof China’s financial system has seriously affected world economic landscape. China’s futuresmarket is emergence and development under the background of this. Although current futuresmarket from the scale to varieties have limitations, it is not hard to see from active tradingmarket prospects that the development of China’s futures market is not limited, and Chinafutures have begun to be the function of hedging and price discovery. In order to China’seconomy more comprehensive and better development, we must go with the world developedcountry standards from the aspects of financial markets, it is necessary for us to deeplydiscuss the development of China’s financial futures trend and influence factors. Through theanalysis of existing trading information, we must find rules of available, providing the futureinformation available for the transaction of financial futures market, in order to make China’sfinancial futures trading system more perfect.Because that most financial data is high frequency, nonlinear and non-stationary, and thedata analysis method which can accuratly analysis and deal with such data in the process isvery limited, and now digital processing method we can use is mostly to linear or nonlinearnon-stationary process smoothly process, such as wavelet analysis, Wagner-Ville, Fourierspectral analysis, a variety of phase plane representation and time delay embedding method.For detailed investigation, the data generated by nonlinear, non-stationary random processfrom the real world, we urgently need new methods. HHT (Hilbert Huang Transform) is anew type of digital signal processing method to most non-linear, non-stationary processintroduced by the national academy of engineering academician Huang E et al. in1998.Stock index futures data analysis using Hilbert-huang transform roughly divided into thefollowing several steps: first of all, do normalization processing to the data to be analysed andeliminate unreasonable data in trading, then carry on the empirical mode decomposition togenerate multiple intrinsic mode function, finally, do Hilbert transform to generate the Hilbertspectrum and Hilbert marginal spectrum based on intrinsic mode function. In this paper, westudy the effectiveness and completeness of the EMD to high frequency financial data ofChina as well as intraday characteristics analysis to the stock index futures based on HHT,intraday characteristics analysis is respectively researched and analysised based on absolutereturns and trading volume of stock index.In his paper, do EMD decomposition to stock index futures returns, because that the MATLAB program of the existing EMD decomposition is not fully applicable to a largenumber of high frequency financial data, so the thesis has carried on the experiment manytimes and design the data reconstruction program; Firstly introduces the rationality andgenerating principle of the experimental data; and then conducted the Hilbert huang transformrespectively, do many experiments to sampling parameters FS of Hilbert transform accordingto the characteristics of financial data by minutes; finally use the most close to the actualparameter values for the Hilbert transform, in this process, first of all, do analysis andsummary to the results of EMD decomposition, and then analyze the generated Hilbertspectrum and Hilbert marginal spectrum, summarized the intraday trend of stock index futuresand the main cycle. |
