GMM Instruments with @dyn or lagged(X to Y)

[gawry]

I'm trying to understand GMM Instruments and what happens when you use the @dyn as opposed to simply lagging a variable. For example suppose I have a variable called GDP

What should be the resulting instrument matrix if you use one of those as instruments?
@dyn(GDP, -2, -3)
GDP(-2 to -3)

I can see that we loose degrees of freedom when using the non-dynamic form of the lagged variable but is not clear the form of the matrix.

In the documentation you mention "The next page is where you will specify your dynamic period-specific (predetermined) instruments. The instruments should be entered with the “@DYN” tag to indicate that they are to be expanded into sets of predetermined instruments, with optional arguments to indicate the lags to be included. If no arguments are provided, the default is to include all valid lags (from -2 to “-infinity”). Here, we instruct EViews that we wish to use the default lags for N as predetermined instruments." -- http://www.eviews.com/help/helpintro.ht ... 23ww189356

Does anyone knows what is the reference journal article is being used to implement this? How does the instrument matrix look like?

Best Regards,
Gustavo

[EViews Gareth]

http://www.eviews.com/help/content/pane ... l#ww189750

@dyn is the set of all lags up to T-2 as described above. For the t=3 observation you use t=1 lagged value. For the t=4 observation you can use both t=1 and t=2.

A standard lag operator would not capture this increasing pattern.