F-statistic in Augmented Dickey Fuller test
Posted: Fri Mar 30, 2018 2:15 am
Hello everyone,
I am having trouble calculating Fischer statistic in Eviews to test the hypotheses of the three models (None- Intercept- Trend) of Dickey Fuller test. According to a document I found, where they explicitly explain all the steps to do so, you can get the sum squared residuals directly from the ADF test, or by writing a small program. Normally you should obtain the same result. But when I applied it to my sample, the ADF test and the code lines gave me completely different results. Can somebody explain to me what's the difference, or whether my program is wrong somehow?
SMPL 7/25/2016 2/21/2018
GENR Y = ERREUR - ERREUR(-1)
LS Y C @trend(7/25/2016) ERREUR(-1)
SCALAR SCR3 = @SSR
SCALAR NDL = @REGOBS - @NCOEF
LS Y C
SCALAR SCR3C = @SSR
SCALAR F3 = ((SCR3C - SCR3)/2)/(SCR3/NDL)
Thank you so much.
I am having trouble calculating Fischer statistic in Eviews to test the hypotheses of the three models (None- Intercept- Trend) of Dickey Fuller test. According to a document I found, where they explicitly explain all the steps to do so, you can get the sum squared residuals directly from the ADF test, or by writing a small program. Normally you should obtain the same result. But when I applied it to my sample, the ADF test and the code lines gave me completely different results. Can somebody explain to me what's the difference, or whether my program is wrong somehow?
SMPL 7/25/2016 2/21/2018
GENR Y = ERREUR - ERREUR(-1)
LS Y C @trend(7/25/2016) ERREUR(-1)
SCALAR SCR3 = @SSR
SCALAR NDL = @REGOBS - @NCOEF
LS Y C
SCALAR SCR3C = @SSR
SCALAR F3 = ((SCR3C - SCR3)/2)/(SCR3/NDL)
Thank you so much.