Hello,
New to the forum. I am a student researcher with an issue conducting bounds test in Eviews. My model contains a quadratic term with an "^" on the said term. When I try to run the bounds test it displays a syntax error in "^" message. This happens even when I just multiply the term by itself to go around the squaring using the symbol "^". Please any suggestions on how to adjust my squared term to go around this problem would be appreciated.
Bounds Test, ARDL quadratic model
Moderators: EViews Gareth, EViews Moderator

 Fe ddaethom, fe welon, fe amcangyfrifon
 Posts: 12501
 Joined: Tue Sep 16, 2008 5:38 pm
Re: Bounds Test, ARDL quadratic model
Could you provide your workfile with the equation?
Follow us on Twitter @IHSEViews
Re: Bounds Test, ARDL quadratic model
Thanks Gareth. Please find it attached here. My equations could be assessed by doubleclicking on any of the names like, Bkkn, Permian, Eagle Ford, Haynesville etc among the files. And my equation could be found within the estimation tab. Thanks. Pascal
 Attachments

 tightoil_plays.wf1
 (360.54 KiB) Downloaded 165 times

 Fe ddaethom, fe welon, fe amcangyfrifon
 Posts: 12501
 Joined: Tue Sep 16, 2008 5:38 pm
Re: Bounds Test, ARDL quadratic model
We'll investigate, but might take a little while. You can always create a new variable equal to the squared term and use that inside your estimation instead.
Follow us on Twitter @IHSEViews

 Posts: 2
 Joined: Mon Sep 11, 2017 9:40 pm
Re: Bounds Test, ARDL quadratic model
wpascals wrote:Hello,
New to the forum. I am a student researcher with an issue conducting bounds test in Eviews. My model contains a quadratic term with an "^" on the said term. When I try to run the bounds test it displays a syntax error in "^" message. This happens even when I just multiply the term by itself to go around the squaring using the symbol "^". Please any suggestions on how to adjust my squared term to go around this problem would be appreciated.

 Posts: 2
 Joined: Mon Sep 11, 2017 9:40 pm
Re: Bounds Test, ARDL quadratic model
I have undertaken my research endeavor on the Environmental Kuznets curve, depicting quadratic relationship between environmental degradation and Per Capita Income.
Some of the variables are stationary at I(0) while other are I(1). So the suggested model is ARDL model. My general econometric model is given below
lnE= ∝o +∝1 (lnY) + ∝2 (lnY)2+ ∝3 (lnEN) + ∝4 (lnTO) + ∝5 (lnUP)+ ∝6 (lnFDI) + εt ...................... (1)
Where
E= Carbon dioxide emission per capita (used as a proxy for environmental degradation/quality) measured in metric tons
Y = Real GDP per capita (used as a proxy for real income) in dollars
EN = energy consumption per capita measured in metric tons
TO = trade openness (Trade to GDP ratio is the sum of export and import as a ratio to GDP)
UP = urban population as a proportion of total population
FDI= foreign direct investment
et= error term
For analysis, I use Eviews9. My Model is loglog model. When I run ARDL Model it shows syntex error with Multicollinearity, Perhaps due to linear association of Per capita income and per capita income square. My questions are
1. How to overcome the problem of multicollinearity and to get accurate results. even when I estimate the long run relationship given in Equ (1), it shows the syntax error of perfect multicolinearity. I am not in the position to sort out this problem.
2. How to find out the turning point of the curve (threshold level of income beyond which further increase in the per capita income leads to decline in the environmental pollution, the carbon emission)
Thanks in advance
Some of the variables are stationary at I(0) while other are I(1). So the suggested model is ARDL model. My general econometric model is given below
lnE= ∝o +∝1 (lnY) + ∝2 (lnY)2+ ∝3 (lnEN) + ∝4 (lnTO) + ∝5 (lnUP)+ ∝6 (lnFDI) + εt ...................... (1)
Where
E= Carbon dioxide emission per capita (used as a proxy for environmental degradation/quality) measured in metric tons
Y = Real GDP per capita (used as a proxy for real income) in dollars
EN = energy consumption per capita measured in metric tons
TO = trade openness (Trade to GDP ratio is the sum of export and import as a ratio to GDP)
UP = urban population as a proportion of total population
FDI= foreign direct investment
et= error term
For analysis, I use Eviews9. My Model is loglog model. When I run ARDL Model it shows syntex error with Multicollinearity, Perhaps due to linear association of Per capita income and per capita income square. My questions are
1. How to overcome the problem of multicollinearity and to get accurate results. even when I estimate the long run relationship given in Equ (1), it shows the syntax error of perfect multicolinearity. I am not in the position to sort out this problem.
2. How to find out the turning point of the curve (threshold level of income beyond which further increase in the per capita income leads to decline in the environmental pollution, the carbon emission)
Thanks in advance
Re: Bounds Test, ARDL quadratic model
As far as I understand it is not a problem using the squares of variables in ARDL estimation then. But this is not NARDL I thinkI Can ve call it parabolic estimation?
"Re: Bounds Test, ARDL quadratic model
Post by EViews Gareth » Tue Aug 02, 2016 9:28 am
We'll investigate, but might take a little while. You can always create a new variable equal to the squared term and use that inside your estimation instead."
"Re: Bounds Test, ARDL quadratic model
Post by EViews Gareth » Tue Aug 02, 2016 9:28 am
We'll investigate, but might take a little while. You can always create a new variable equal to the squared term and use that inside your estimation instead."
Who is online
Users browsing this forum: No registered users and 7 guests