Hi Statalisters,
I have another question that is bothering me for quite some time and i would love to finish that puzzle in my head. I would like to calculate the variable LRETR (long-run cash effective tax rate) and im not getting the results that im hoping for.
I have searched the forum and got similar posts, but they do not seem that different to my approach:
https://www.statalist.org/forums/forum/general-stata-discussion/general/1600054-long-term-effective-tax-rate-over-5-years
https://www.statalist.org/forums/for...ctive-tax-rate
The variable is from the Paper of An, Zhang 2013 and is defined as follows:
"LRETR is the ratio of the sum of the cash tax paid during the last 3 years to the sum of the difference between pretax income and special items during the last 3 years"
What i do is the following:
The variables i used should be the right ones, because the authors copied the variable from another paper, where the original authors exactly say which compustat items they used (and my dataset is from compustat)
->The problem i have is, that the results (see stats and regression) are not consistent with the literature and I know that this is normal to a certain degree, but =0 is a bit extreme and also the high SD makes me wondering.
I would be very happy if you could enlighten me. I have probably overlooked or misinterpreted something!
I also know that i have the same problem with the DTURN variable, but thats a problem for another time and maybe a seperate post
if there is anything else I can provide please let me know.
Here is a sample from my compustat annual dataset:
My Statistics:

Stats from the original Paper:

My REG (you see LRETR is 0)

The results from the Paper that i try to replicate:

I have another question that is bothering me for quite some time and i would love to finish that puzzle in my head. I would like to calculate the variable LRETR (long-run cash effective tax rate) and im not getting the results that im hoping for.
I have searched the forum and got similar posts, but they do not seem that different to my approach:
https://www.statalist.org/forums/forum/general-stata-discussion/general/1600054-long-term-effective-tax-rate-over-5-years
https://www.statalist.org/forums/for...ctive-tax-rate
The variable is from the Paper of An, Zhang 2013 and is defined as follows:
"LRETR is the ratio of the sum of the cash tax paid during the last 3 years to the sum of the difference between pretax income and special items during the last 3 years"
What i do is the following:
Code:
rangestat (sum) txpd, interval(fyear -2 0) by(gvkey) label var txpd_sum "Sum of the cash tax paid during the last 3 years" bys gvkey fyear: gen diff_pi_spi = pi-spi label var diff_pi_spi "Difference between pretax income and special items" rangestat (sum) diff_pi_spi, interval(fyear -2 0) by(gvkey) label var diff_pi_spi_sum "Sum of the difference between pretax income and special items during the last 3 years" bys gvkey fyear: gen LRETR = txpd_sum / diff_pi_spi_sum label var LRETR "long-run cash effective tax rate LRETR"
->The problem i have is, that the results (see stats and regression) are not consistent with the literature and I know that this is normal to a certain degree, but =0 is a bit extreme and also the high SD makes me wondering.
I would be very happy if you could enlighten me. I have probably overlooked or misinterpreted something!
I also know that i have the same problem with the DTURN variable, but thats a problem for another time and maybe a seperate post

if there is anything else I can provide please let me know.
Here is a sample from my compustat annual dataset:
Code:
* Example generated by -dataex-. For more info, type help dataex clear input long gvkey int(fyear year) double(txpd pi spi) 185319 2011 2012 3.597 .97 -.997 18582 2018 2018 -1.045 3.277 0 110801 2001 2002 .046 -4.668 -3.925 8768 1999 1999 14.742 59.714 59.734 163987 2010 2010 .891 -1.34 -1.34 28236 2013 2013 .73 -10.332 -.564 62285 1999 1999 15.812 38.132 38.17 148255 2001 2001 15 -161.7 -46.6 160642 2006 2006 .092 -24.877 -4.179 133304 2008 2008 .245 -5.459 -2.914 100699 2011 2012 375.743 -2893.204 -1964.008 177446 2012 2012 .132 3.016 -.212 30231 2003 2004 .312 20.457 -.163 6529 2010 2010 29.3 45.392 -58.885 105089 2015 2015 432.509 200.174 -49.142 2829 2017 2017 .349 .4 .021 147307 2014 2014 0 -9.084 -5.114 108804 2017 2017 27.914 -125.914 -20.28 3607 2010 2010 6 -550 -520 65911 2009 2009 .23 -16.482 -14.354 142909 2002 2002 1.764 -6.537 -1.282 175055 2018 2018 12.397 -26.273 -24.458 133432 2009 2009 9.899 -3.641 -1.298 10622 2001 2002 416.09 -2813.887 -1571.083 7601 2002 2002 2.173 -1.47 0 65570 2008 2008 .352 16.79 -3.654 12852 2007 2007 .054 -12.207 -12 170527 2016 2016 73.4 -30.1 -102.7 3946 2017 2017 78.2 -175.7 -142.6 23777 2000 2000 1.756 -11.762 -3.649 13354 2005 2005 31.639 -18.096 4.971 160661 2007 2007 13.762 -15.444 -36.344 9812 2009 2009 .236 38.445 41.843 12877 1999 1999 .263 -2.445 -2.44 119034 2005 2005 55.881 9.811 0 65459 2008 2008 16.999 -941.942 -906.271 164755 2010 2010 11.048 80.792 -3.839 160949 2007 2007 30.321 -64.261 -118.217 62694 2004 2004 .001 -1.077 -.61 63938 2008 2009 1.267 10.549 0 157755 2011 2011 -2.994 5.572 -1.89 60931 2000 2000 5.027 -65.254 0 117781 2012 2012 7.103 -50.621 -14.121 175342 2012 2012 .493 -89.81 -38.641 27816 2002 2002 12.474 -34.187 -45.817 63724 1999 1999 27.6 -5.795 -5.094 10658 2005 2005 .354 -2.756 0 126718 2017 2017 2.131 -2.14 -4.187 118401 2010 2010 5.872 -24.014 -10.913 106687 2009 2009 .039 -.61 -.163 13163 2007 2008 .243 8.093 .113 15050 2004 2005 .951 -6.334 0 15274 2000 2000 . -156.727 -190.882 27845 2016 2016 3692.7 -2832 -2541.8 218399 2016 2017 93.073 334.848 -63.118 64853 2009 2010 20.219 84.834 -3.547 64630 2006 2006 9.875 -184.482 -36.35 14145 2003 2003 .546 -.636 0 112413 2008 2008 5.057 -60.039 -58.062 12904 2001 2002 -.445 -22.334 -15.067 10443 2002 2002 27 28 7 125074 2018 2018 6.415 14.798 .442 27490 2018 2018 86.605 30.57 225.982 64103 2010 2010 3.988 11.443 -4.735 26560 2016 2016 440.7 549.4 52.7 63800 2013 2013 8 -48 -38 3366 2000 2000 .819 1.045 0 12658 2007 2007 .909 .555 0 19113 1998 1999 115.207 -4.38 0 9548 2000 2000 133 299 157 27925 2008 2008 5.5 -195.514 -157.26 13645 2003 2003 -.861 .972 0 31005 2018 2018 22.356 -208.283 -224.671 12708 2006 2006 2.416 -.942 -3.155 65275 2012 2012 12.823 11.752 -.9 27490 2015 2015 90.163 -4.408 -.616 25972 2002 2002 5.914 9.885 1.913 31099 2011 2011 -.247 1.053 0 174390 2007 2007 .92 -2.702 -1.415 66059 2006 2006 .248 -8.95 -1.435 178796 2009 2009 1.721 -43.06 0 24171 2000 2001 4.7 -167.421 -110 101973 2017 2017 .586 3.184 0 18672 2013 2013 3.036 54.287 57.366 6809 1999 1999 .669 .131 .163 100590 2018 2018 718.996 3759.844 . 185194 2016 2016 .548 -16.882 -.554 27794 2017 2017 8.494 -4.664 .553 2061 2005 2005 0 8.07 -1.619 176238 2011 2011 -.852 -86.768 .3 11162 2006 2006 .876 -11.328 -2.021 145083 2000 2000 .612 2.321 0 26892 2016 2016 0 -6.668 0 2581 2005 2005 .005 .455 0 65899 2007 2007 36.615 55.987 73.409 17934 2012 2012 1.626 34.767 -4.221 105442 2005 2005 1.014 4.855 0 13003 2003 2003 19.067 27.866 -4.765 62396 2011 2011 4.235 14.333 11.171 10053 2002 2002 16 9 14 end
Stats from the original Paper:
My REG (you see LRETR is 0)
The results from the Paper that i try to replicate:
Comment