Advances and Applications in Statistics
Volume 41, Issue 1, Pages 21 - 67
(July 2014)
|
|
A ROBUST AND PRACTICAL METHOD OF UNIT ROOT TESTS
Ka Ho Wu, Zhao-Guo Chen and Wensheng Wang
|
Abstract: Among the criteria for evaluating a unit root test method, the followings are the most important: it is capable of maintaining power and minimizing the size distortion in difficult situations; there is no major effect of the distributions of the underlying stationary series on the results; it may eliminate the impact of the initial conditions on the results. That is because a priori information on them is usually unavailable in practical applications. The graphical polyvariogram method is an excellent tool for identifying the multiplicity of unit roots, but uncertainty cannot be always avoided. This paper further develops polyvariogram methodology by deriving the asymptotic distribution of a test statistic and proposes several testing procedures. As the above criteria are concerned, simulation studies clearly demonstrate the superiority of the proposed method over the traditional methods in the literature. |
Keywords and phrases: ARIMA models, asymptotic distribution, linear series, polyvariogram, power of tests, size distortion. |
|
Number of Downloads: 380 | Number of Views: 1132 |
|