## Linear Detrend

For econometric discussions not necessarily related to EViews.

Moderators: EViews Gareth, EViews Moderator

Amaziah
Posts: 18
Joined: Wed Apr 22, 2015 3:22 pm

### Linear Detrend

I am have some questions about how to detrend a series with a deterministic trend and how to apply it in Eviews. Can someone check if my process is correct?

Let's say I want to regress a variable Yt on a variable Xt, but Xt exhibits a deterministic trend.

To detrend Xt, I run a regression on a time index variable, TIME. Would my Eviews command would be: ls x TIME

I then obtain the estimated residuals from the above regression, ut. Now do I use this new series, ut as my regressor for my original equation? So I would then estimate the following: ls y c u

startz
Non-normality and collinearity are NOT problems!
Posts: 3551
Joined: Wed Sep 17, 2008 2:25 pm

### Re: Linear Detrend

Amaziah wrote:I am have some questions about how to detrend a series with a deterministic trend and how to apply it in Eviews. Can someone check if my process is correct?

Let's say I want to regress a variable Yt on a variable Xt, but Xt exhibits a deterministic trend.

To detrend Xt, I run a regression on a time index variable, TIME. Would my Eviews command would be: ls x TIME

I then obtain the estimated residuals from the above regression, ut. Now do I use this new series, ut as my regressor for my original equation? So I would then estimate the following: ls y c u

Almost, the first step needs a constant

Code: Select all

`ls x c TIME`

Amaziah
Posts: 18
Joined: Wed Apr 22, 2015 3:22 pm

### Re: Linear Detrend

Okay. Now the residuals that are obtained are considered the detrended series?

startz
Non-normality and collinearity are NOT problems!
Posts: 3551
Joined: Wed Sep 17, 2008 2:25 pm

Yep.