st: bitobit estimate

[Christos Konstantinidis <chkons@agro.auth.gr>]

Dear Sir/Madam,
Iam Christos Konstantinidis and I am a Phd Student in the Aristotle  
University of Thessaloniki, Greece. Within the context of my thesis I  
use stata for panel data analysis. I have to estimate simultaneously  
both individually for eight different sectors of food and beverage  
industry and for the whole sample two equations using the bi-tobit  
syntax. The employed bi-tobit syntax is the one written by you  
"bitobit estimate, y1(depvar1) (pnp) y2(depvar2) (ms) x1(varlist) (ms  
pnplag gr opc ks) x2(varlist) (pnp mslag lev age ks) censor1(varname)  
(pnp_censor) censor2(varname) (ms_censor)"and the equations have the  
form below:
pnp=a0+a1ms+a2pnplag+a3gr+a4opc+a5ks
ms=b0+b1pnp+b2mslag+b3lev+b4age+b5ks.
In three of the eight sectors the estimation is ok. However in the  
rest five sectors as well as in the whole sample the estimation cannot  
be performed and messages such as ''numerical derivatives are  
approximate, flat or discontinuous region encountered" or ''could not  
evaluate numerical derivatives, missing values encountered" appear.
Please could you help? Could you also inform me whether the truncated  
limits have been determined correctly in my syntax? Notice that the  
dependendent variables, profitability (pnp) and market share (ms) are  
truncated and take values between -1 and 1 and 0 and 1 respectively.
Thank you in advance.
Christos Konstantinidis


*
*   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/