st: Inconsistancy of Monte Carlo in STATA

[Nomoreno <nomorenook@netscape.net>]

Hello all,

I encounter a weird STATA problem and I am hoping I can get some help here.

I wrote some code for a one-shot Monte Carlo simulation, and I have another master do file to call this one-shot Monte Carlo simulation program and run it 100 times. In this master file, I also set two global variables, N and T as for the dimension of the data for one shot Monte Carlo program. N = 5, and T varies from 20, 30, ... to 80. The results for N = 5 and T=30 is very weird. To be more specific, the results are way off my exception. However, if I just run N=5 and T =30 Monte Carlo simulation alone, I get correct expected results.

I am wondering whether anyone here has any clue what's going on with STATA? I mean, it is the same program and just T varies from 20 to 80. All the simulation results for any T are close to what I expect except T = 30 produced very weird outcomes. If I ran that simulation alone, I got expected results. It does not make sense to me and this inconsistency drives me crazy...

Thank you very much and I really appreciate your help. 

Best,
Xiaobo



*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/