MM Estimations in Eviews 8

[tawani]

Hello there,

I am Vineet, PhD student and I need to run MM estimation for my analysis. I have set of 800 companies. I have their daily share prices and Market data matched date wise. However, how do I run MM estimation without running Eviews for 800 times?
Is there a way to run bunch of regression equations in one go?

Thanks,
Vineet

[EViews Gareth]

Write an EViews program to do it. There are plenty of examples in the programming forum.

[tawani]

Thanks Gareth!
I have successfully written a program to generate M and MM Methods for bunch of regression equations. Following is my program. I run one equation at a time and append all my results in a Spool to export to excel for my further analysis.

My data is arranged vertically. With s_id as a first numeric column to identify pairs.Then I have 5 columns X1, X2, X3, X4, X5. I have 166 Samples with 250 paired observations each arranged vertically in Excel. All imported to Eviews.
The coding works perfectly. However, as of now, it is transferring all the values of the output. Could you please modify the program to just give me Sample-ID, Intercept and Slope coefficient(s) with std errors, t-stats and P values? Pretty much central part of the output? I tried using Coefs objects for no avail. I could not get my head around to use it. Your help is much appreciated.

-----------------------------------------------------------------------------------------------------------------------------------------------------------------------
Spool MEstCBA ' where results shall be appended
String SName

Sample ss @all if s_id = 1 ' Defining First Sample Set as per Sample - ID


!counter = 1 ' To keep track of Sample ID and Run iterations

while !counter <= 166

smpl ss ' Assigning first Sample Set


equation eq1.robustls y c x1
'equation eq1.robustls y c x2
'equation eq1.robustls y c x3 x4 x5
'equation eq1.robustls(method=mm, tuning=2.937, mtuning=3.44,cov=type2,trial = 500) y c x1
'equation eq1.robustls(method=mm, tuning=2.937, mtuning=3.44,cov=type2,trial = 500) y c x2
'equation eq1.robustls(method=mm, tuning=2.937, mtuning=3.44,cov=type2,trial = 500) y c x3 x4 x5

do eq1.output(p)


%Sname = "Sample - " + @str(!counter) ' Each output shall be Displayed with Sample ID
MEstCBA.displayname !counter %Sname ' Each output shall be Displayed with Sample ID

!counter = !counter + 1

ss.set @all if s_id = !counter ' Next Sample Id assigned to the sample set

wend
--------------------------------------------------------------------------------------------------------------------------------------------------------

Thank you,
Vineet