Non-linear Fixed Effects

For econometric discussions not necessarily related to EViews.

Moderators: EViews Gareth, EViews Moderator

TML
Posts: 10
Joined: Mon Jun 25, 2018 4:20 pm

Non-linear Fixed Effects

Postby TML » Thu Jun 28, 2018 9:20 am

Can anyone help me how to run non-linear fixed effects on Eview? If i have equation like Y=a+b+b^2+c+c^2, what should i write in the command on Eviews?

startz
Non-normality and collinearity are NOT problems!
Posts: 3327
Joined: Wed Sep 17, 2008 2:25 pm

Re: Non-linear Fixed Effects

Postby startz » Thu Jun 28, 2018 9:51 am

Are a,b, and c series with coefficients to be estimated?

TML
Posts: 10
Joined: Mon Jun 25, 2018 4:20 pm

Re: Non-linear Fixed Effects

Postby TML » Sat Jul 07, 2018 4:51 pm

a is the intercept while B and C is beta. Am i answering your question? sorry...i am new on Eviews so I do not really understand about your question.

startz
Non-normality and collinearity are NOT problems!
Posts: 3327
Joined: Wed Sep 17, 2008 2:25 pm

Re: Non-linear Fixed Effects

Postby startz » Sat Jul 07, 2018 5:15 pm

As written, your equation doesn't seem to have any data in it.

TML
Posts: 10
Joined: Mon Jun 25, 2018 4:20 pm

Re: Non-linear Fixed Effects

Postby TML » Sat Jul 07, 2018 5:23 pm

Thank Startz for your prompt response. Please find the actual equation on the attached file. All of equation is data. I am using panel data in Eviews.
Attachments
Equation.PNG
Equation.PNG (9.15 KiB) Viewed 191 times

startz
Non-normality and collinearity are NOT problems!
Posts: 3327
Joined: Wed Sep 17, 2008 2:25 pm

Re: Non-linear Fixed Effects

Postby startz » Sat Jul 07, 2018 5:30 pm

This equation is linear in the coefficients, so it's estimated by linear least squares. Nothing special needs to be done about the fixed effects.

TML
Posts: 10
Joined: Mon Jun 25, 2018 4:20 pm

Re: Non-linear Fixed Effects

Postby TML » Sat Jul 07, 2018 5:36 pm

Based on your explanation, does it mean that I can run it the same way as I run linear fixed effects, right?
I did run it using the same way as linear fixed effect with just different equation that I showed you. However, the result came out that p-value is higher than 60% for all variables.

startz
Non-normality and collinearity are NOT problems!
Posts: 3327
Joined: Wed Sep 17, 2008 2:25 pm

Re: Non-linear Fixed Effects

Postby startz » Sat Jul 07, 2018 6:09 pm

Well, then the individual coefficients aren't significant.

You might want to test the joint significance of beta1 and beta2 or beta3 and beta4.

TML
Posts: 10
Joined: Mon Jun 25, 2018 4:20 pm

Re: Non-linear Fixed Effects

Postby TML » Sun Jul 08, 2018 2:02 am

What do you mean by running join significant? May I know how to do it? Do you have any link to share for this?

startz
Non-normality and collinearity are NOT problems!
Posts: 3327
Joined: Wed Sep 17, 2008 2:25 pm

Re: Non-linear Fixed Effects

Postby startz » Sun Jul 08, 2018 6:01 am


TML
Posts: 10
Joined: Mon Jun 25, 2018 4:20 pm

Re: Non-linear Fixed Effects

Postby TML » Sun Jul 08, 2018 6:38 am

By doing so, it will no longer be a fixed effect estimator, right?

startz
Non-normality and collinearity are NOT problems!
Posts: 3327
Joined: Wed Sep 17, 2008 2:25 pm

Re: Non-linear Fixed Effects

Postby startz » Sun Jul 08, 2018 7:28 am

No, this is just a test of coefficients. You can do it either with fixed effects or without fixed effects.

TML
Posts: 10
Joined: Mon Jun 25, 2018 4:20 pm

Re: Non-linear Fixed Effects

Postby TML » Sun Jul 08, 2018 7:42 am

i cannot find those option based on the guideline. The guideline is in version 10 while mine is version 9.

startz
Non-normality and collinearity are NOT problems!
Posts: 3327
Joined: Wed Sep 17, 2008 2:25 pm

Re: Non-linear Fixed Effects

Postby startz » Sun Jul 08, 2018 8:12 am

Look in the built-in help system under Wald test/coefficient restriction

TML
Posts: 10
Joined: Mon Jun 25, 2018 4:20 pm

Re: Non-linear Fixed Effects

Postby TML » Mon Jul 16, 2018 9:36 am

Hi, after doing further research, non linear time fixed-effects might fit my study.
May i know how to write equation in Eviews if i want to use time fixed-effects.
I use this equation as non-linear fixed effects (Y c A _A^2 B B^2 lngdp crisis_year)
Then i tried the same equation on time fixed-effect and tick Fixed on both Cross-section and Period in Panel Option, but it shows "Near singular matrix".
Your prompt response is really appreciated.


Return to “Econometric Discussions”

Who is online

Users browsing this forum: No registered users and 4 guests