"No valid observations" problem in SUR estimation

[andrew]

I tried to estimate a three-equation system using SUR, but an error message box pops out saying "no valid observations" in the last equation. I don't know what this is all about or how to do with it. I uploaded the workfile in the attachment. Can anybody take a look and help me? Thanks in advance!

[EViews Gareth]

At your current starting values (0 for all coefficients), the third expression in your system evaluates to all NAs, which gives no valid observations. Use different starting values (I set C equal to 1, and it seemed to work).

[andrew]

Problem solved. Thanks so much!

[Bulldozer]

I get the same error message, but when estimating the system for VECM. I have four series, I have run the JJ cointegration and found there are more than one cointegration vectors, I then run VECM, made the system (Proc>>Make System), then I tried to get the significance values through system estimation (Estimate), but the Error Message came up (No valid observation in equation....). Any suggestions?

[massoud982000]

Dear friends,
I faced the same problem as Andrew in Eviews, I would kindly request you to help me
((I tried to estimate a three-equation system using SUR, but an error message box pops out saying "no valid observations" in the last equation. I don't know what this is all about or how to do with it. I uploaded the workfile in the attachment. Can anybody take a look and help me?))
Please find the file I have attached ( I run VECM with 2 endogenous variables )
I am so thankful of your favor in advance.
Massoud

My Email: massoud982000@yahoo.com

[EViews Gareth]

Hard to help without knowing exactly what you're doing.