|
Abstract:
|
Using methodology in panel unit root tests we propose a new approach to univariate unit root tests. Our method leads to an asymptotically normal distribution of the OLS estimator and is robust to structural changes while the power of the test does not drastically worsen. The main idea is that under the assumption that the process has a unit root we transform a univariate time series process to a double-index process in such a way that the segments are independent, i.e., we divide the T observations into m sets of size n where m sets of data are independent of each other. For this transformed data, we apply the sequential limits to obtain the normal limiting distribution, which is the same as in Levin and Lin (1992). An advantage of this technique is that an undetected break has a relatively minor effect on the asymptotic results which, in fact, disappears as the segments increase.
|
