Dear Statalisters,
I am using xtabond to estimate a model in which firm profits (return on
assets, to be specific) is the dependent variable. Given the persistence of
profits literature, I expected the time lag (t-1) of the DV to be positive
with a value between 0 and 1. And, indeed, ROA and its time lag is
positively correlated. The first-differenced ROA and time lag, however, are
negatively correlated in the xtabond estimates. There are some outliers and
missing observations in the ROA variable to contend with, but these issues
notwithstanding I am wondering if the negative coefficient of the time lag
is indicative of a model misspecification or other estimation problem.
Thanks in advance for your help!
Larry Plummer
University of Colorado
*
* 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/
