Unit Root Test

TITLE:

Impact of Model Specification Decisions on Unit Root Tests

ABSTRACT:

Performance of unit tests depends on several specification decisions prior to their application e.g., whether or not to include a deterministic trend. Since there is no standard procedure for making such decisions, therefore the practitioners routinely make several arbitrary specification decisions. In Monte Carlo studies, the design of DGP supports these decisions, but for real data, such specification decisions are often unjustifiable and sometimes incompatible with data. We argue that the problems posed by choice of initial specification are quite complex and the existing voluminous literature on this issue treats only certain superficial aspects of this choice. We also show how these initial specifications affect the performance of unit root tests and argue that Monte Carlo studies should include these preliminary decisions to arrive at a better yardstick for evaluating such tests.

REFERENCE:

Zaman, Asad & Atiq-ur-Rehman, "Impact of Model Specification Decisions on Unit Root Tests", MPRA Paper No. 19963, posted 12. January 2010.

TITLE: A Test of Unit Root Tests

CHANGED TITLE TO: An Empirical Evaluation of Observational Equivalence and Unit Root Tests for GDP Series.

Unit root tests have become a standard part of conventional econometric methodology ever since Nelson and Plosser (1982) showed that real data may have unit roots, and conventional regression procedures break down in presence of unit roots. Presence or absence of unit roots affects not only inference but also determines theoretical properties of models related to equilibrium. Despite the obvious importance of unit roots, which has led to widespread use of unit root tests, there remain some central problems which are unresolved to date. The main contribution of this paper is to formulate and resolve two such problems which have not received much attention in the unit root literature.

The first problem is that of observational equivalence of unit root and stationary series, as established by Faust (1996). What sense is there in unit root testing if every finite data series can be equally well fitted by a unit root process as well as a stationary process? The second problem has to do with a plethora of tests which have been devised for unit root testing. It has been shown that many of these tests have low power, and cannot reject the null hypothesis. Tests have been devised with null hypothesis of unit root, as well as that of stationarity. It can be demonstrated by choosing a suitable test, we can get any result we desire regarding any data series.

Keywords: Unit Root Tests, Stationarity, GDP

JEL Classification: C01, C15, C22