I want to truncate (not winsorize) the most extreme 1 percent of earnings. However, I do not know how to do this. My dataset looks like this (simplified):
Note: earnings is already scaled by average total assets. Therefore, the values are low.
Code:
* Example generated by -dataex-. To install: ssc install dataex clear input str6 gvkey double fyear float Earn "001004" 2002 .069433704 "001004" 2003 .09366571 "001004" 2004 .06395068 "001004" 2005 .00225294 "001004" 2006 -.04376862 "001004" 2007 .05516222 "001004" 2008 .014650915 "001004" 2009 .07596077 "001004" 2010 .04182295 "001004" 2011 .10636716 "001004" 2012 .028460857 "001004" 2013 .07873188 "001004" 2014 -.09327173 "001004" 2015 .01596238 "001004" 2016 .0748177 "001004" 2017 .06239358 "001004" 2018 .01184788 "001013" 2003 .11190998 "001013" 2004 .18013138 "001013" 2005 .04881214 "001013" 2006 -.008894078 "001013" 2007 .10534018 "001013" 2008 .09784136 "001013" 2009 -.007185459 "001013" 2010 .16246234 "001034" 2003 .06254857 "001034" 2004 .04231646 "001034" 2005 .033480093 "001034" 2006 .205559 "001034" 2007 .03662908 "001038" 2002 .10530272 "001038" 2003 .12154207 "001045" 2003 .07058154 "001045" 2004 .017279265 "001045" 2005 .04276995 "001045" 2006 .05527683 "001045" 2007 .08618657 "001045" 2008 -.05465978 "001045" 2009 -.003224015 "001045" 2010 .077378 "001045" 2011 .03835765 "001045" 2012 .0459423 "001045" 2013 .06547737 "001045" 2014 .02320717 "001045" 2015 .1101205 "001045" 2016 .1618523 "001045" 2017 .09375992 "001045" 2018 .05428436 "001050" 2003 .10410062 "001050" 2004 .14352474 "001050" 2005 .03800196 "001050" 2006 .2147853 "001050" 2007 -.14851879 "001050" 2008 -.01105276 "001050" 2009 -.00946848 "001050" 2010 .12418208 "001050" 2011 .14547464 "001050" 2012 .14874977 "001050" 2013 .3618757 "001050" 2014 -.13913119 "001050" 2015 .05535443 "001050" 2016 .02353288 "001050" 2017 .035376854 "001050" 2018 -.017243665 "001056" 2003 -.04523037 "001056" 2004 .1733007 "001056" 2005 -.06380673 "001056" 2006 .09907055 "001072" 2002 .17861393 "001072" 2003 .015089057 "001072" 2004 .09200164 "001072" 2005 .03811658 "001072" 2006 .17410243 "001072" 2007 .13513482 "001072" 2008 -.04685184 "001072" 2009 .12183116 "001072" 2010 .14520466 "001072" 2011 .033670727 "001072" 2012 .026322335 "001072" 2013 -.03926912 "001072" 2014 .1094193 "001072" 2015 -.005357519 "001072" 2016 .14301506 "001072" 2017 .12882896 "001072" 2018 .06967193 "001075" 2005 .17439023 "001075" 2006 -.07582778 "001075" 2007 .04151408 "001075" 2008 .12286757 "001075" 2009 .08130658 "001075" 2010 .062811024 "001075" 2011 .0845019 "001075" 2012 .10071879 "001075" 2013 .072788104 "001075" 2014 .073774114 "001075" 2015 .07311531 "001075" 2016 .0734632 "001075" 2017 .0793887 "001075" 2018 .05596517 "001076" 2012 .032616034 end
Thank you in advance.
Roy
0 Response to Truncation of the most extreme 1 percent of earnings
Post a Comment