st: nonlinear combination from multiple model results

["b.tracy" <bst73@yahoo.com>]

I am trying to create a nonlinear combination of parameters from two different models (and I am not succeeding).

As an example:
Model 1
reg y1 x1 x2

Model 2
reg y2 x3 x4

I now would like to estimate something like:
elas = 1 + [y1]_b[x1]/scalar1 + [y2]_b[x3]/scalar2

I have been using -nlcom- for other, similar estimates (with parameters from only one model), but I cannot find a way to make it work to include results from different models.

My specific case is a bit more complex than my example above, as my estimates are from multiple, multiple regression models (i.e., -reg3-).
My specific case regards the estimation of a total price elasticity from nested translog cost functions (which requires parameters from both estimations), in the event someone has done this before using a different approach.
 
Any help would be greatly appreciated.

Regards,

Brandon.
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/