Residual On VEC

[fafancute]

Hi

I have a problem in obtaining the residual after I do the Vector error correction. After I do the test the resid File wont Updated, somethimes its give the NA value.
Do anyone have a clue what is wrong with this?? I really need your Help!
I need this resid file for the Error Correction Model

Cheers

[EViews Gareth]

Could you provide more details?

Which version of EViews are you using, and what's the build date (both available from Help->About EViews).

Could you post your Workfile?

Could you give exact instructions on what you're doing?

[fafancute]

Hi Its May 4th 2007. Standard edition.

Im Have to make the ECM test for the Co integration vector. I need the Residual from the VEC estimation, but apparantly the Resid file wont update.

Here is the work file

http://rapidshare.com/files/438909783/e ... ambbmf.wf1

http://rapidshare.com/files/438909797/e ... lambmf.wf1

[EViews Gareth]

Which version of EViews are you using? (EViews 5 or EViews 6?)

Could you give exact instructions on what you are doing?

[fafancute]

Im using Eviews 6

here is the instruction about what I have to do for the ECM

-ECM test with Johansen Cointegration vectors
- Open the variables you want to test as a VAR by selecting first the dependent variable and then the independent variable and use the right mouse button to open as a VAR
- Choose the right options and after the right screen is opened by clicking OK in the top the Cointegration Equation is provided
- Then generate a new series of the form residual
- After this new series is saved use Quick – Estimate Equation to make an equation of the form: d(ld) c d(ld(-1)) d(ld(-2)) d(ld(-3)) d(ld(-4)) d(ly(-1)) d(ly(-2)) d(ly(-3)) d(ly(-4)) residuldly(-1)
- Here the T-test result is provided
- For the two F-test use View – Coefficient test – redundant variables and provide the factors you want to test. In the example equation given above this would be
d(ly(-1)) d(ly(-2)) d(ly(-3)) d(ly(-4)) for the first F-test and d(ly(-1)) d(ly(-2)) d(ly(-3))
d(ly(-4)) residuldly(-1) for the second F-test.


Thanks For you help :)

[EViews Gareth]

Hard to follow what you're doing if you don't tell us which variables you are using.... etc... What's the exact specification of the VEC you're estimating?

[fafancute]

So actually I have to check the Causality betwen the Financial development and Economic growth.
For the Financial development indicator I use the 1st: ratio of bank deposit lialbilities to nominal GDP (its LM in the work file) 2nd: bank claim on private sector to nominal GDP (its LD in the work file) and I want to see the if this indicator cause economic growth which is GDP percapita (LY in the work file).

Further more: firts I have to test the co integration first. And it ends up that the variable are co integrated, Therefore I have to proced with the causality testing. And the Teacher wants us to Use the Error Correction Model to determine the causality. The step is just like I posted above.

The Type of VEC that I have to estimate is just the regular VEC testing. No exogenous variable and test with intercept and no trend.

[EViews Gareth]

Seems to work for me.

The version of EViews 6 you are using is massively out of date. You should update to the latest version:
http://eviews.com/download/download.html

[fafancute]

Yeah, All of my other friend also did not find this kind of problem. Only me. I still dont know where do I do wrong.
is importing data from Excel file could be the cause?? I already Update my eviews and its still wont work.

So when I chose the dependend and independent variable Chose open - open as VAR then Chose the VEC set lag value and click "OK" the resid file should be updated right?? or not??

Cheers