Formula to calculate covariance of coefficients from two regressions

[Bart Rokicki]

Dear Statalisters,
Does anyone know what is the formula to calculate the covariance of two coefficients from two different regressions?
It seems easy once we have the command in statistical software but it is not that straightforward elsewhere...
I found the formula for covariance of coefficients from one regression (below). Still, I cannot find the one used to calculate the same thing out of the 2 regressions.

Any help will be really appreciated.
Bart

[Michal Brzezinski]

Have you looked at the Stata manual entry for suest?
Hope this helps,
Michal

[Bart Rokicki]

Yes, I just did. Thanks! However, suest does not work with my specification. Need to see though whether can use the formula provided and calculate the covariance by hand.

[Steve Samuels]

See FAQ 12 about saying that something, your case, suest, "doesn't work".

[Bart Rokicki]

Sorry for that Steve.
I basically mean that you cannot apply suest for instrumental variables estimation (e.g. ivregress). This is an issue that was already discussed in other posts.

[Steve Samuels]

Well, I don't know anything about ivregress, but your best bet for a helpful answer is to illustrate the problem with an accessible data set, or provide one with dataex from SSC. Show the code and results. All this is requested in the FAQ I quoted.

[Joao Santos Silva]

Dear Bart,

Would you be willing to bootstrap?

All the best,

Joao