Nonlinear least squares specification for exp(c(1)*t)

For technical questions regarding estimation of single equations, systems, VARs, Factor analysis and State Space Models in EViews. General econometric questions and advice should go in the Econometric Discussions forum.

Moderators: EViews Gareth, EViews Moderator

acemi
Posts: 46
Joined: Thu Jun 23, 2016 10:28 pm

Nonlinear least squares specification for exp(c(1)*t)

Postby acemi » Thu Jun 23, 2016 10:38 pm

Hello I'm trying to specify my derived macroeconomic model which includes exponential function of time t :

logy=(1-exp(-beta*t))alfa(1-theta)logs + ... +error

data is for year end 1993 - 2009.
my question is how do I specify exp(-beta*t) part since I dont include a specific time series variable "t" ?
do I omit the t in the expression exp(-beta*t), I'm confused.
Best regards

startz
Non-normality and collinearity are NOT problems!
Posts: 3797
Joined: Wed Sep 17, 2008 2:25 pm

Re: Nonlinear least squares specification for exp(c(1)*t)

Postby startz » Fri Jun 24, 2016 12:31 am

Use @trend.

acemi
Posts: 46
Joined: Thu Jun 23, 2016 10:28 pm

Re: Nonlinear least squares specification for exp(c(1)*t)

Postby acemi » Sun Jun 26, 2016 10:10 am

Thank yo for reply, but from what I found in a brief search @trend is a linear trend operator. So where would I exactly include in my model I couldnt be sure exp(-beta*t) :

exp(-beta @trend) ?

Thank you in advance.

startz
Non-normality and collinearity are NOT problems!
Posts: 3797
Joined: Wed Sep 17, 2008 2:25 pm

Re: Nonlinear least squares specification for exp(c(1)*t)

Postby startz » Sun Jun 26, 2016 10:16 am

exp(-beta*@trend)

acemi
Posts: 46
Joined: Thu Jun 23, 2016 10:28 pm

Re: Nonlinear least squares specification for exp(c(1)*t)

Postby acemi » Tue Jan 03, 2017 12:11 am

Hello,
I specified an equation including exp(-beta @trend) where @trend denotes "t" as time in Eviews7.
My data set contains years as time (e.g 1991, 1992 ,etc to 2009) for panel data consisting of 12 countries.
My questions are,

1. does eviews insert 1991,1992, ... in place of "t" or does it insert 1,2,3 .. ?
2. For panel data does it reset the year value when it moves to a new country data (i.e. start from 1991 or 1 again) ? (or does it start from 1 to continue increasing t until it reaches end of whole data set. (i.e. t = 12 countries * 19 years = 228)

The reason I am asking is that, when I manually calculate exp(-beta @trend) expression in excel, the whole expression equals zero for t = 1991, 1992,..
but if I use t = 1,2, ... expression equals a positive value. I have to simulate previously so that I ensure there are no negative exp () values.

Thank you and regards,

startz
Non-normality and collinearity are NOT problems!
Posts: 3797
Joined: Wed Sep 17, 2008 2:25 pm

Re: Nonlinear least squares specification for exp(c(1)*t)

Postby startz » Tue Jan 03, 2017 7:38 am

Either look at the Help for @trend, or do show @trend

acemi
Posts: 46
Joined: Thu Jun 23, 2016 10:28 pm

Re: Nonlinear least squares specification for exp(c(1)*t)

Postby acemi » Tue Jan 03, 2017 10:27 pm

Thank you show @trend answered my question


Return to “Estimation”

Who is online

Users browsing this forum: No registered users and 2 guests