univariate regression

[Afsaneh]

Hi there,

I regress the dependent variable (Y) on a lagged of independent variable (x) in Eviews 9 by applying the command(Y c x(-1)). I surprise, because if I lagged X manually in an Excel file and then applying that in regression(Y c x),I get a different result from the previous estimation. Also, if this time I lagged independant variable manually forwrd (not backward like my previous regressions, by inserting a cell at the beginning of period and shift the independent variable down), I get the similar result as I apply backward lag function on independant variable in Eviews (Y c x(-1)).

I confused, as I belive I should get simillar results if I either apply lag function on independant variable in Eviews (Y c x(-1)) or lagged independant variable manually in Excel before I apply it in the regression command in Eviews (Y c x). Then, why I get different results?

Also, how it is possible to get the same results of estimation by applying a manually lag forward of independant variable and a lag backward (by Eviews) of independant variable ? Aren't these regressions theoretically and mathematically different?
Thanks

[EViews Gareth]

Perhaps you could provide the files you're using, with full steps of what you're done. Otherwise it is all a little imaginary.

[startz]

I suspect you are doing the lags backwards in Excel. Open your Excel spreadsheet. Then in EViews

Code: Select all

show y x x(-1)

See how things line up.