Page 1 of 1

Single-equation cointegration test

Posted: Wed Jan 16, 2019 5:27 am
by farrel
Dear all,

I estimate cointegration between wages and prices (in logs) using Single-equation cointegration test.
I aim at finding cointegration between some cointegration there, however, the results show that according to Tau-statistic I cannot reject Null, but according to z-statistic I reject Null.
What is principle difference between Tau- and z-statistics? Could anyone suggest what conclusion I can make in this regard?

Results are here:

Date: 16/01/19 Time: 14:20
Series: LOG(HICPSERVX) LOG(WAGETOT)
Sample (adjusted): 2005Q1 2018Q3
Included observations: 55 after adjustments
Null hypothesis: Series are not cointegrated
Cointegrating equation deterministics: C
Automatic lags specification based on Schwarz criterion (maxlag=10)


Dependent tau-statistic Prob.* z-statistic Prob.*
LOG(HICPSERVX) -2.776511 0.1906 -27.89261 0.0034
LOG(WAGETOT) -2.667696 0.2278 -26.64448 0.0050

*MacKinnon (1996) p-values.

Intermediate Results:
LOG(HICPSERVX) LOG(WAGETOT)
Rho - 1 -0.144848 -0.140836
Rho S.E. 0.052169 0.052793
Residual variance 0.000133 0.000293
Long-run residual variance 0.001902 0.004031
Number of lags 3 3
Number of observations 51 51
Number of stochastic trends** 2 2

**Number of stochastic trends in asymptotic distribution


Best
Andrei