[Gretl-users] Cointegration Engle-Granger ?

Hi guys, I tried the Engle-Granger cointegration test but I noticed something. Here is my procedure: 1.I estimated a equatation in cointegration dialogue with the following variables l_CPI, l_REER, MMR and l_Y using 12 lags and got these results down. 2. These ADF tests for these variables are wright and the results are the same if you do ADF tests through main Gretl window using the same number of lags. 3.In step 5 we have the cointegration results and they are the same if you estimate a OLS model through main Gretl window. 4.In step 6 we have ADF tests for the residuals. Tests are without a constant. I tried to check the same test through main Gretl window using the same specification ie. I saved residuals from OLS regression (which I done through the main Gretl window) and tested with the ADF test without a constant. Here are the results: Step 6: Dickey-Fuller test on residuals Augmented Dickey-Fuller tests, order 12, for uhat sample size 86 unit-root null hypothesis: a = 1 test without constant estimated value of (a - 1): -0,333671 test statistic: t = -2,46579 asymptotic p-value 0,4903 Augmented Dickey-Fuller tests, order 12, for uhat sample size 86 unit-root null hypothesis: a = 1 test without constant model: (1 - L)y = (a-1)*y(-1) + ... + e estimated value of (a - 1): -0,333671 test statistic: t = -2,46579 asymptotic p-value 0,01324 Test statistics and estimated values, sample size are the same, but p-values are different? cointegration and from the second (from OLS saved residuals) that there is a cointegration relationship. Why is that? Did I make same mistaque in comparing? In addition I have some one suggestion. There sould be a possibility in E-G cointegration to have a possibility to do ADF tests-test down from maximal lag order or to have a possibility to see full ADF tests output in automatic E-G cointegration procedure. With unique lag order specification one can't decide which is the wright lag order. *********************************** Step 1: testing for a unit root in l_CPI Augmented Dickey-Fuller tests, order 12, for l_CPI sample size 86 unit-root null hypothesis: a = 1 test with constant estimated value of (a - 1): -0,0107669 test statistic: t = -1,2609 asymptotic p-value 0,6499 with constant and trend estimated value of (a - 1): -0,0838183 test statistic: t = -2,51286 asymptotic p-value 0,3218 with constant and quadratic trend estimated value of (a - 1): -0,111113 test statistic: t = -1,86159 asymptotic p-value 0,8653 Step 2: testing for a unit root in l_REER Augmented Dickey-Fuller tests, order 12, for l_REER sample size 86 unit-root null hypothesis: a = 1 test with constant estimated value of (a - 1): -0,0393688 test statistic: t = -1,28046 asymptotic p-value 0,641 with constant and trend estimated value of (a - 1): -0,244764 test statistic: t = -4,98476 asymptotic p-value 0,0001 with constant and quadratic trend estimated value of (a - 1): -0,132398 test statistic: t = -1,67802 asymptotic p-value 0,9135 Step 3: testing for a unit root in MMR Augmented Dickey-Fuller tests, order 12, for MMR sample size 86 unit-root null hypothesis: a = 1 test with constant estimated value of (a - 1): -0,0881019 test statistic: t = -1,97264 asymptotic p-value 0,2992 with constant and trend estimated value of (a - 1): -0,137154 test statistic: t = -2,13296 asymptotic p-value 0,5267 with constant and quadratic trend estimated value of (a - 1): -0,263064 test statistic: t = -2,29448 asymptotic p-value 0,6817 Step 4: testing for a unit root in l_Y Augmented Dickey-Fuller tests, order 12, for l_Y sample size 86 unit-root null hypothesis: a = 1 test with constant estimated value of (a - 1): 0,0899723 test statistic: t = 1,56751 asymptotic p-value 0,9995 with constant and trend estimated value of (a - 1): -1,13912 test statistic: t = -3,52897 asymptotic p-value 0,0363 with constant and quadratic trend estimated value of (a - 1): -1,06705 test statistic: t = -2,83268 asymptotic p-value 0,3813 Step 5: cointegrating regression Cointegrating regression - OLS estimates using the 99 observations 1998:01-2006:03 Dependent variable: l_CPI VARIABLE COEFFICIENT STDERROR T STAT P-VALUE const 5,03993 0,562006 8,968 <0,00001 *** l_REER -0,305748 0,0911986 -3,353 0,00115 *** MMR -0,0121330 0,00102131 -11,880 <0,00001 *** l_Y 0,225401 0,0419834 5,369 <0,00001 *** Unadjusted R-squared = 0,845111 Adjusted R-squared = 0,840219 Durbin-Watson statistic = 0,633496 First-order autocorrelation coeff. = 0,668083 Step 6: Dickey-Fuller test on residuals Augmented Dickey-Fuller tests, order 12, for uhat sample size 86 unit-root null hypothesis: a = 1 test without constant estimated value of (a - 1): -0,333671 test statistic: t = -2,46579 asymptotic p-value 0,4903 P-values based on MacKinnon (JAE, 1996) There is evidence for a cointegrating relationship if: (a) The unit-root hypothesis is not rejected for the individual variables. (b) The unit-root hypothesis is rejected for the residuals (uhat) from the cointegrating regression. Augmented Dickey-Fuller tests, order 12, for uhat sample size 86 unit-root null hypothesis: a = 1 test without constant model: (1 - L)y = (a-1)*y(-1) + ... + e estimated value of (a - 1): -0,333671 test statistic: t = -2,46579 asymptotic p-value 0,01324 P-values based on MacKinnon (JAE, 1996) _________________________________________________________________ Windows Live™ Messenger has arrived. Click here to download it for free! http://imagine-msn.com/messenger/launch80/?locale=en-gb