2sls with 2 simultaneous equations

For econometric discussions not necessarily related to EViews.

Moderators: EViews Gareth, EViews Moderator

AgentRose
Posts: 14
Joined: Sat Apr 02, 2022 7:23 am

2sls with 2 simultaneous equations

Postby AgentRose » Sun Apr 03, 2022 2:16 am

Say I got 2 equations:
(1) Y1 = Alpha + Y2 + AX1 + BX2 + CX3 + E
(2) Y2 = Beta + Y1 + DX4 + BX2 + FX5 + GX6 + U

How would you do the 2sls? I keep getting singular matrix error messages or insufficient instrument variables?
I have been inputting equation 1 as model specification and X1, X2, X3, X4, X5, and X6 as the instruments, but keep getting error messages so I tried different things but still keep receiving errors or really unsure of the answers.

startz
Non-normality and collinearity are NOT problems!
Posts: 3775
Joined: Wed Sep 17, 2008 2:25 pm

Re: 2sls with 2 simultaneous equations

Postby startz » Sun Apr 03, 2022 7:07 am

Show us the commands you use and maybe attach the EViews workfile.

AgentRose
Posts: 14
Joined: Sat Apr 02, 2022 7:23 am

Re: 2sls with 2 simultaneous equations

Postby AgentRose » Sun Apr 03, 2022 7:31 am

startz wrote:Show us the commands you use and maybe attach the EViews workfile.


Quick, Estimate Equation, 2sls

Equation specification: Y1 c Y2 + AX1 + BX2 + CX3 + E

Instruments: AXI BX2 CX3 DX4 FX5 GX6

I think I got near a singular matrix for this one. How are you meant to do 2sls when you got 2 simultaneous equations with 2 endogenous variables and 6 exogenous variables? The variable X2 get repeated in both equations, and the seconds' equation have more variables.

AgentRose
Posts: 14
Joined: Sat Apr 02, 2022 7:23 am

Re: 2sls with 2 simultaneous equations

Postby AgentRose » Sun Apr 03, 2022 7:33 am

AgentRose wrote:
startz wrote:Show us the commands you use and maybe attach the EViews workfile.


Quick, Estimate Equation, 2sls

Equation specification: Y1 c Y2 X1 X2 X3 + E

Instruments: XI X2 X3 X4 X5 X6

I think I got near a singular matrix for this one. How are you meant to do 2sls when you got 2 simultaneous equations with 2 endogenous variables and 6 exogenous variables? The variable X2 get repeated in both equations, and the seconds' equation have more variables.

AgentRose
Posts: 14
Joined: Sat Apr 02, 2022 7:23 am

Re: 2sls with 2 simultaneous equations

Postby AgentRose » Sun Apr 03, 2022 7:35 am

How do you let Eviews know which variables are endogenous and which variables are exogenous in tsls, and the variables that form the 2 equations?

startz
Non-normality and collinearity are NOT problems!
Posts: 3775
Joined: Wed Sep 17, 2008 2:25 pm

Re: 2sls with 2 simultaneous equations

Postby startz » Sun Apr 03, 2022 7:37 am

AgentRose wrote:
startz wrote:Show us the commands you use and maybe attach the EViews workfile.


Quick, Estimate Equation, 2sls

Equation specification: Y1 c Y2 + AX1 + BX2 + CX3 + E

Instruments: AXI BX2 CX3 DX4 FX5 GX6

I think I got near a singular matrix for this one. How are you meant to do 2sls when you got 2 simultaneous equations with 2 endogenous variables and 6 exogenous variables? The variable X2 get repeated in both equations, and the seconds' equation have more variables.


Is the variable named X2 or BX2? When using EViews this way you don't need to include coefficient names or the error term or + signs.

AgentRose
Posts: 14
Joined: Sat Apr 02, 2022 7:23 am

Re: 2sls with 2 simultaneous equations

Postby AgentRose » Sun Apr 03, 2022 11:32 am

startz wrote:
AgentRose wrote:
startz wrote:Show us the commands you use and maybe attach the EViews workfile.


Quick, Estimate Equation, 2sls

Equation specification: Y1 c Y2 + X1 + X2 + X3 + E

Instruments: XI X2 X3 X4 X5 X6

I think I got near a singular matrix for this one. How are you meant to do 2sls when you got 2 simultaneous equations with 2 endogenous variables and 6 exogenous variables? The variable X2 get repeated in both equations, and the seconds' equation has more variables.


Is the variable named X2 or BX2? When using EViews this way you don't need to include coefficient names or the error term or + signs.
Quick, Estimate Equation, 2sls

Equation specification: Y1 c Y2 X1 X2 X3

Instruments: XI X2 X3 X4 X5 X6

I think I got near a singular matrix for this one. How are you meant to do 2sls when you got 2 simultaneous equations with 2 endogenous variables and 6 exogenous variables? The variable X2 get repeated in both equations, and the seconds' equation has more variables.

I didn't put any signs, coefficient or error terms when inputting the variables on the Eviews screen

AgentRose
Posts: 14
Joined: Sat Apr 02, 2022 7:23 am

Re: 2sls with 2 simultaneous equations

Postby AgentRose » Sun Apr 03, 2022 11:35 am

How would you do it when there is an unequal number of variables in the two equations. I am sure the equation is over-identified so 2sls should work right, or at least produce some results from half the equation?

startz
Non-normality and collinearity are NOT problems!
Posts: 3775
Joined: Wed Sep 17, 2008 2:25 pm

Re: 2sls with 2 simultaneous equations

Postby startz » Sun Apr 03, 2022 11:42 am

That looks correct. So it's probably a problem with your data.

AgentRose
Posts: 14
Joined: Sat Apr 02, 2022 7:23 am

Re: 2sls with 2 simultaneous equations

Postby AgentRose » Sun Apr 03, 2022 4:46 pm

startz wrote:That looks correct. So it's probably a problem with your data.

Could it not be something to do with the second equation having 1 additional variable compared to equation 1 or that the variable X2 exists in both equations?

startz
Non-normality and collinearity are NOT problems!
Posts: 3775
Joined: Wed Sep 17, 2008 2:25 pm

Re: 2sls with 2 simultaneous equations

Postby startz » Sun Apr 03, 2022 5:13 pm

AgentRose wrote:
startz wrote:That looks correct. So it's probably a problem with your data.

Could it not be something to do with the second equation having 1 additional variable compared to equation 1 or that the variable X2 exists in both equations?


Nope. The former means that the first equation is overidentified and the latter means that X2 is exogenous and serving as its own instrument. Both are fine for 2sls.

AgentRose
Posts: 14
Joined: Sat Apr 02, 2022 7:23 am

Re: 2sls with 2 simultaneous equations

Postby AgentRose » Mon Apr 04, 2022 3:04 am

startz wrote:
AgentRose wrote:
startz wrote:That looks correct. So it's probably a problem with your data.

Could it not be something to do with the second equation having 1 additional variable compared to equation 1 or that the variable X2 exists in both equations?


Nope. The former means that the first equation is overidentified and the latter means that X2 is exogenous and serving as its own instrument. Both are fine for 2sls.


You were right! THANK YOU so MUCH!!

AgentRose
Posts: 14
Joined: Sat Apr 02, 2022 7:23 am

Re: 2sls with 2 simultaneous equations

Postby AgentRose » Mon Apr 04, 2022 3:18 am

For the first stage of the equations would you just input LS of the structural equation for equations 1 and 2 to get the F-stat or would you input the reduced form equation to get the F-stat?

startz
Non-normality and collinearity are NOT problems!
Posts: 3775
Joined: Wed Sep 17, 2008 2:25 pm

Re: 2sls with 2 simultaneous equations

Postby startz » Mon Apr 04, 2022 6:11 am

For the first-stage F you want the reduced form.

AgentRose
Posts: 14
Joined: Sat Apr 02, 2022 7:23 am

Re: 2sls with 2 simultaneous equations

Postby AgentRose » Mon Apr 04, 2022 5:20 pm

startz wrote:For the first-stage F you want the reduced form.


Wouldn't you get the same generalised equation but like different letters instead?

How would you do it with so many variables? What difference does it make? Cause I did get an answer but I think I did it the structural formula way and get a low F-stat. Can you just straight up put equation 2 in the brackets with the rest of equation 1 and not simplify the equation further on Eviews?


Return to “Econometric Discussions”

Who is online

Users browsing this forum: No registered users and 17 guests