Bi-variate Tobit Using CMP Command

[Yuyang Chen]

Hi everyone, I am trying to estimate a model with two equations and each of them is imbedded in Tobit model. It went well with "bitobit estimate" command whereas I cannot write a proper code to estimate using "cmp" command written by @David Roodman. There are two question I have:
One is, I tried to follow the instruction to first, execute "cmp setup"; second, execute "cmp (y1=xlist1)(y2=xlist2), ind(2 2)". However, it doesn't converge as it does using "bitobit estimate" command.
Another one is, though "2" in the code indicates "observation is left-censored for this (tobit) equation at the value stored in the dependent variable", the estimation using "cmp" still sensor the maximum value of the dependent variable.

Can anyone help with the question I have, especially the first question. As I am aiming to, eventually, estimate a TRI-variate Tobit model. And "cmp" command would be the only available one to run the estimation as far as I know.

Many Thanks,
Yuyang

[David Roodman]

Move toward your goal incrementally, and look at the examples in the help file. First try to fit single-equation models with cmp. You'll see that your try does not match the help file examples. Then build up to 2 and 3 equations.