Dear All,

My study compares the exports performance of motor vehicles firms in India by using Tobit model and Cragg’s Double Specification model. In my sample, the dependent variable (Export Intensity) has many zero values. Thus, I consider limited dependent model specifications to be appropriate.

I got the results of Tobit and Probit model but I am not getting results of Truncated regression model.

I have Panel data of total 3414 variables (299 firms for the period of 17 years).

I am using this command truncreg Ylist Xlist, ll(0)

But I am not getting the results, it is showing

cannot compute an improvement -- discontinuous region encountered

r(430);

Please see the attached file.