Dear all,
I am comparing the forecasting performance of two models estimated on panel
data with small T (<20) and large N (> 2000). I compare the forecasting
performance estimating the two models on T-1 years and forecasting on the
last year of the panel. Is the Diebold-Mariano test appropriate in this
case, or it is suitable only for time series? If not appropriate, anybody
can suggest another approach?
Thank you very much,
Carlo
*
* 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/
