Panel Unit Root Tests Study With Structural Breaks Model

The panel data is the mixed data of the time and cross-section data, so that the model can be better used for internal research of similar economical problems. This unique advantage makes the theory of unit root tests have a very wide range of application. Most studies have been dependent on the short panel data of time T is fixed and cross-section N is large, which have better small sample properties. The assumption of T-fixed makes the panel unit root tests can be well applied to the empirical study, only needing a relatively short time period can observe a large number of the cross section of the unit. So the main purpose of the article is to explore and improve the T-fixed panel unit root tests research with structural breaks. Moreover, considering the structural breaks in the process of panel data unit root tests is also a major hot spot in the field of statistics and econometrics. Therefore, expanding the panel unit root tests with structural breaks and doing empirical analysis have important theoretical significance and application value.At the time of deeply discussing the panel unit root tests with structural breaks, the article mainly deduces the limit distribution of test statistics under the known break point and unknown break point in the use of central limit theorem on the assumption of panel disturbance terms have heterogeneity and serial correlation.In addition, the article selects all of the 12-month period monthly closing price in the year of 2008 of 100 listed companies shares of real estate development industry as a specific research data in order to make the research have more application value instead of theoretical significance. Then the paper uses R software analyzing the data, clearly investigating the effects of structural breaks on the panel unit root test results, and obtaining that it is very necessary to consider the structural breaks in the data generation process.The innovation of the article is mainly manifested in the following two aspects:first, discussing the fixed T panel unit root tests with structural breaks on the assumption of panel disturbance terms have heterogeneity and serial correlation and drawing the limit distribution of test statistics. Second, The previous literatures which studying structural breaks of fixed-T panel unit root tests have used the simulation to drawn relevant conclusions, the results of this simulation have certain reference value, but there is deviation in the process of practical application. As a result, the article selects the specific economic data to analysis and explore the affect of structural breaks on the unit root test results, which makes the theory of the paper have more guidance meaning in practice.