Dear all,
This question is more related to research than to STATA. However, since many
STATALISTER are familar with xtabond2-related research maybe somebody can
help me.
Comparing OLS and GMM after out-of-sample prediction for a dynamic model, my
result is, that prediction based on OLS dominates the prediction based on
GMM. This result surprised me, since the OLS estimator is biased in a
dynamic setting if residuals are autocorrelated.
However, empirical work usually only compares the data fit of OLS and GMM
based on in-sample estimation (see Arellano/Bond for instance). I could not
find empirical results on out-of-sample results.
My question is: Can anybody cite me some sources that deal with
out-of-sample prediction comparison based on OLS and GMM estimator?
Thanks in advance
Niko
*
* 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/
