st: non stationary time series with SUR

["Teresa Linz" <tlinz@uni-bonn.de>]

Dear stata subscribers,

I have a problem with some time series analysis. I am 
estimating 45 monthly time series with seemingly unrelated 
regression. I have five independent variables which are 
estimated in a cost function as a translog functional 
form. However two of the variables are non stationary 
(also the transformed ln variables are non stationary). 
what can I do about that. Do I have to include them in a 
differenced form in the SUR model? How do I do that?
The command I am using right now looks like that:
sureg (Sw = lnPw lnPe lnPl lnPc lny) (Se = lnPw lnPe lnPl 
lnPc lny) (Sl = lnPw lnPe lnPl lnPc lny) (Sc = lnPw lnPe 
lnPl lnPc lny)(lnTC = lnw lne lnl lnc lnd lny lnww lnee 
lnll lncc lndd lnyy lnwe lnwl lnwc lnwd lnel lnec lned 
lnlc lnld lncd lnwy lney lnly lncy lndy), constraints(1 2 
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 
25 26 27 28 29 30 31) isure corr nolog.
Here lnl and lnc are non stationary as well as all 
interrelated variables.

It would be nice to get an answer about this. Thank you 
very much.

Greetings,
Teresa




______________________________

Teresa Linz, M.Sc.

Junior Researcher

Center for Development Research

Walter-Flex-Str. 3

53113 Bonn

Germany

Tel.: +49 228/73 4962
Fax:  +49 228/73 1869
Internet: www.zef.de
*
*   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/