Hi All,
I have performed the modified ADF on my variable using the
At level I get this
DF-GLS for lgdp2 Number of obs = 16
Maxlag = 8 chosen by Schwert criterion
DF-GLS tau 1% Critical 5% Critical 10% Critical
[lags] Test Statistic Value Value Value
------------------------------------------------------------------------------
8 -1.972 -3.770 -3.432 -2.725
7 -1.464 -3.770 -3.114 -2.532
6 -1.128 -3.770 -2.965 -2.471
5 -0.718 -3.770 -2.950 -2.512
4 -0.565 -3.770 -3.033 -2.625
3 -0.087 -3.770 -3.178 -2.781
2 -0.259 -3.770 -3.349 -2.949
1 -0.763 -3.770 -3.509 -3.100
Opt Lag (Ng-Perron seq t) = 4 with RMSE .4052668
Min SIC = -1.068416 at lag 1 with RMSE .4928773
Min MAIC = -1.224771 at lag 2 with RMSE .4729084
At first difference
DF-GLS for D.lgdp2 Number of obs = 15
Maxlag = 8 chosen by Schwert criterion
DF-GLS tau 1% Critical 5% Critical 10% Critical
[lags] Test Statistic Value Value Value
------------------------------------------------------------------------------
8 -0.936 -3.770 -3.702 -2.892
7 -0.917 -3.770 -3.257 -2.604
6 -1.075 -3.770 -3.024 -2.482
5 -1.423 -3.770 -2.960 -2.489
4 -1.293 -3.770 -3.021 -2.590
3 -1.905 -3.770 -3.163 -2.748
2 -3.540 -3.770 -3.343 -2.927
1 -3.790 -3.770 -3.517 -3.091
Opt Lag (Ng-Perron seq t) = 0 [use maxlag(0)]
Min SIC = -2.090892 at lag 1 with RMSE .2934692
Min MAIC = 27.11756 at lag 1 with RMSE .2934692
1. Am I correct in saying that at level there is presence of unit root given that the absolute value of the t stat is less than absolute critical values at 1, 5 and 10% while at the first difference the variable is made stationary and the null is rejected at 1, 5 and 10%?
2. since the dfgls already includes the trend in the above, is there a way to run it can so as to show results that are based on inclusion of trend and intercept
3. when I run the ADF, the var is only stationary at 10%. Is it ok if I present my data showing the for the ADF test the var is only stationary at 10 while for the modified ADF at 1, 5 and 10%, there I assume this variable is stationary and proceed to run regressions?
Augmented Dickey-Fuller test for unit root Number of obs = 20
---------- Interpolated Dickey-Fuller ---------
Test 1% Critical 5% Critical 10% Critical
Statistic Value Value Value
------------------------------------------------------------------------------
Z(t) -3.258 -4.380 -3.600 -3.240
------------------------------------------------------------------------------
MacKinnon approximate p-value for Z(t) = 0.0735
I have performed the modified ADF on my variable using the
Code:
dfgls
DF-GLS for lgdp2 Number of obs = 16
Maxlag = 8 chosen by Schwert criterion
DF-GLS tau 1% Critical 5% Critical 10% Critical
[lags] Test Statistic Value Value Value
------------------------------------------------------------------------------
8 -1.972 -3.770 -3.432 -2.725
7 -1.464 -3.770 -3.114 -2.532
6 -1.128 -3.770 -2.965 -2.471
5 -0.718 -3.770 -2.950 -2.512
4 -0.565 -3.770 -3.033 -2.625
3 -0.087 -3.770 -3.178 -2.781
2 -0.259 -3.770 -3.349 -2.949
1 -0.763 -3.770 -3.509 -3.100
Opt Lag (Ng-Perron seq t) = 4 with RMSE .4052668
Min SIC = -1.068416 at lag 1 with RMSE .4928773
Min MAIC = -1.224771 at lag 2 with RMSE .4729084
At first difference
DF-GLS for D.lgdp2 Number of obs = 15
Maxlag = 8 chosen by Schwert criterion
DF-GLS tau 1% Critical 5% Critical 10% Critical
[lags] Test Statistic Value Value Value
------------------------------------------------------------------------------
8 -0.936 -3.770 -3.702 -2.892
7 -0.917 -3.770 -3.257 -2.604
6 -1.075 -3.770 -3.024 -2.482
5 -1.423 -3.770 -2.960 -2.489
4 -1.293 -3.770 -3.021 -2.590
3 -1.905 -3.770 -3.163 -2.748
2 -3.540 -3.770 -3.343 -2.927
1 -3.790 -3.770 -3.517 -3.091
Opt Lag (Ng-Perron seq t) = 0 [use maxlag(0)]
Min SIC = -2.090892 at lag 1 with RMSE .2934692
Min MAIC = 27.11756 at lag 1 with RMSE .2934692
1. Am I correct in saying that at level there is presence of unit root given that the absolute value of the t stat is less than absolute critical values at 1, 5 and 10% while at the first difference the variable is made stationary and the null is rejected at 1, 5 and 10%?
2. since the dfgls already includes the trend in the above, is there a way to run it can so as to show results that are based on inclusion of trend and intercept
3. when I run the ADF, the var is only stationary at 10%. Is it ok if I present my data showing the for the ADF test the var is only stationary at 10 while for the modified ADF at 1, 5 and 10%, there I assume this variable is stationary and proceed to run regressions?
Augmented Dickey-Fuller test for unit root Number of obs = 20
---------- Interpolated Dickey-Fuller ---------
Test 1% Critical 5% Critical 10% Critical
Statistic Value Value Value
------------------------------------------------------------------------------
Z(t) -3.258 -4.380 -3.600 -3.240
------------------------------------------------------------------------------
MacKinnon approximate p-value for Z(t) = 0.0735
