Time Series Cointegration Theory In The Stock Index In The Empirical Analysis

The time series which researched in quantity economic usually are non-stationary, using the classical linear model describes this kind of economic phenomena with the time series analysis often to appear the question of "the false return". In the financial system research, analyze the correlation relations between the multi-dimensional time series frequently, like short-term information, long-term balanced relations. In order to avoid the question of "the false return", compositing the short-term information and the long-term balanced relations between time series, usually, we must analyze between twoor more non-stationary time series the long-term balanced relations------co-integrationrelations.Co-integration theory is to research the co-integration relations and the existence of inspection procedure mathematics theory between the multi-dimensional time series. Using co-integration theory to analyze the long-term balanced relations between the multi-dimensional time series is the important application domain. This article in order to analyze the relations between the index data time series of Shanghai, Shenzhen, HongKong stock marketThis article works as follows:(1) This article integrated the correlation literature to introduce quite systematically the linear co-integration theory, and the non-linear co-integration theory and so on, and the present research situation of this theory in domestic and foreign.(2) This article analyzed contrast quite in detail the validity and the reliability of the inspection procedure of co-integration relations existence between the time series in different situation at the same time this article introduced the correlated co-integration theory.(3) This article carried out diagnosis research such as the basic characteristic statistics, integration examination , co-integration examination, Granger causality relation analysis and CCF examination of the stock time series which using the daily closing synthetic index data of Shanghai, Shenzhen, HongKong stock market, discussed the co-integration relations existence in these time series, examined the long-term stable balanced relations of the three stock markets with the index data time series in some certain time section which selected using the daily closing synthetic index data of these stock markets.This article research indicated that:(1) The finance time series are obvious non-stationary, which formed to use the daily closing synthetic index data of Shanghai, Shenzhen, HongKong stock market.(2) The finance time series which formed the daily closing synthetic index data of Shanghai, Shenzhen, HongKong stock market exists co-integration relations in some certain time section, which indicates there are long-term stationary balanced relations between these stock market.(3) Using co-integration theory to get the result that there are long-term stationary balanced relations among Shanghai, Shenzhen, HongKong stock market are tallied with the actual comparison.(4) This article conclusion is not consistent with certain correlation research conclusion, the reason is: The data this article selects is composite index data and the time section is longer, the frequency is higher.