Research On The Extension And Application Of Cointegration Theory

Modern economic theory regards macroeconomics as a complex system with dynamic random characteristics.The system changes with the individual’s action,and the individual reacts further with the changes of the system.Under the influence of this feedback-negative-feedback economic mechanism,the variables themselves are prone to exhibit non-stationary characteristics,and the variables tend to form long-term complex synergistic associations Cointegration theory provides a basic method for modeling non-stationary data,which has been fully studied and widely applied.However,due to the complexity of the economic system itself,some important cointegration relationships supported by economic theory and practical experience may be rejected by the cointegration test,and thus cannot be analyzed by means of cointegration theory.As the economic data itself constantly presents new features and the understanding of economic data is continuously deepened,the parametric model shows certain limitations in the analysis of complex synergistic integral relations.Nonparametric kernel regression method,with strong data resilience,has unique advantages in fitting complex cointegration relationship.This paper expands the research scope of the cointegration theory with the help from nonparametric kernel regression method,and gives a simple and robust tool for some complex cointegration relations,which were difficult to analyze before and originally need using many different types of parameter models together.Considering that the current Chinese economy is in the complex situation of profound changes in the external environment and internal structure optimization and adjustment,it is of great practical significance to provide effective empirical research tools for complex synergy integration systems at this stage.This paper firstly organizes the existing tools of the theory of cointegration research,laying the foundation for the study of more complex cointegration problems.To modeling and analyzing cointegration relationships with nonlinear characteristics,kernel regression method was introduced into cointegration theory and cointegration test method was improved.Furthermore,to modeling and analyzing cointegration relationships with structure change characteristics,time-varying non-parametric model was introduced while its estimating process was simplified and corresponding cointegration test method was given.The study found that cointegration theory could have a good perform when analyzing cointegration relationship with complex characteristics with the help of nonparametric kernel regression methods.The joint-test which combining the ADF test and the local DW test was powerful and useful when testing the cointegration relationship with nonlinear features.And the generalized local DW test extended from local DW test can be used to test whether there is a cointegration relationship among variables when variables having structure-change features.Results of Monte Carlo experiment are consistent with theory.The paper also applies the method to the empirical study of the Chinese economy,capturing the complex cointegration that cannot be found before,including the cointegration relationship with cyclical structure change between the M2 and economic growth and the cointegration relationship with S-type nonlinear features between import and economic growthThe main innovations of the research are as follows.Firstly,this paper provides a new tool based on the kernel regression method for the identification and analysis of complex cointegration relations,thus expanding the application range of the cointegration theory.Secondly,the diagnostic method of spurious-regression in non-parametric kernel regression model is improved,and the cointegration test method which is compatible with the kernel regression estimation method is developed.Third,by applying the cointegration method to the empirical research of several hot issues in China economy,some deeper understanding of the monetary policy impact mechanism and the development law of import trade in the China economy were achieved with the help of the expended cointegration theory.