st: 2sls and ols using predicted values - inference

[Manuela Deidda <manuelad.deidda@gmail.com>]

Dear statalist,

I have the following equation to be estimated with IV:

(1)  ivregress 2sls y1 x1 x2 x3 (x4=z1 z2)


I need to implement several commands that do not allow the iv
structure (i.e. movestay, suest....).

I tried to compute 2sls "manually":

(2) regress x4 z1 z2 x1 x2 x3
(3) predict res, res

(4)  regress y1 x1 x2 x3 res

As expected, the coefficient associated to my explanatory variables
are the same as in equation (1), however the standard errors are
slighltly different. This may be a problem when performing test for
example.


I know it's probably a silly question, but is there a way to compute
manually 2sls so that I can obtain correct standard errors?

Thanks in advance for your help!
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/