Estimating two linear equations (one after the other)

[stephanie22]

Hi guys,

I am trying to estimate a model containing two linear equations.

(1) R = a0 + a1Vt-1 + a2ABt-1

I want to use the estimated R from equation 1 and use it as an independent variable in the second equation:

(2) Q = a0 + a1(R) +.....

I can obviously estimate equation 1 with standard OLS, however, I have no idea how to save these estimated R values in eviews
in order to use it in the second equation.
Also, I was wondering if there is an eviews function where you can view a graph of the estimated equation, I can only find a graph
of the residuals of the equation.

Thank you very much in advance!

[EViews Gareth]

You want the fitted values from Equation 1, which can be obtained by hitting the forecast button (fitted values are just a static forecast over the original estimation sample, after all).

What does a "graph of the estimated equation" mean?

[stephanie22]

Thank you very much!

Forget what I said about the graph, cause I get it now :D

I have one more question.

I also estimate a 'rental adjustment equation'.
First I estimate R* = a0 + a1Vt-1 + a2ABt-1

Following, I want to estimate the adjustment process, which is:

Rt-Rt-1 = a3 [R* - Rt-1]

If you fill R* in the above equation, I get:

(1) Rt = a3[a0 + a1Vt-1 + a2ABt-1] + (1-a3)Rt-1 (where a3 gives the rate of adjustment)

How can I estimate equation 1 in eviews? What should I type in the estimate equation box?

Thanks in advance!

[EViews Gareth]

Something like:

Code: Select all

R = c(4)*(c(1) + c(2)*v(-1) + c(3)*ab(-1)) + (1-c(4))*r(-1)

[stephanie22]

thank you so much!

I have one last question:
If you assume a time trend in eviews, you can estimate '@trend'
If I include this variable in the regression, the signs of my coefficients change.
This change is good for my model, however are these signs reliable if I include a trend, because
without the trend the variables have the wrong sign.