Estimating beta's for each id and store in a variable - panel data

[Rafael Maschke]

Dear Stata users,

I have an unbalanced panel dataset with 4329 observations and about 400 variables. Time periods range from 2003 to 2010 with gaps.

I would to run a time-series regression to estimate beta coefficients and store them in a variable for each mfiid and its time period, taking into account the panel structure (gaps). _roe is the dependent, _sp500 the independent variable. The dependent variable varies by mfiid each year and the independent variable is the same for each mfiid each year.

Sorted by year and mfiid:

* Example generated by -dataex-. To install: ssc install dataex
clear
input long mfiid int year float(_roe _sp500)
100000 2007    .2913083 -.010829791
100000 2008   .17713334   -.4216229
100000 2009   .14123334     .201956
100001 2004   .04955833    .0472585
100001 2005 -.008699998  -.01277183
100001 2006   .02328333   .08855005
100001 2007   .21060835 -.010829791
100001 2009  .015933335     .201956
100001 2010  .017058332   .09580792
100004 2008   .20683333   -.4216229
100008 2006   -.4828166   .08855005
100008 2007   -.4197917 -.010829791
100008 2008  -2.9300666   -.4216229
100008 2009    9.721633     .201956
100012 2005       .3659  -.01277183
100012 2006   -.4431167   .08855005
100012 2007   1.6969082 -.010829791
100012 2008   1.2119334   -.4216229
100012 2009    .9820334     .201956
100012 2010   .16185834   .09580792
100013 2009  -.10906667     .201956
100013 2010  -.13484167   .09580792
100016 2005      -.0101  -.01277183
100016 2006 .0035833344   .08855005
100016 2007  .016408335 -.010829791
100016 2008   .05173333   -.4216229
100016 2009   .07153333     .201956
100016 2010   .08925833   .09580792
100017 2003      .17895    .2237602
100017 2004   .16935833    .0472585
100017 2006   .09988333   .08855005
100020 2007  .003608331 -.010829791
100020 2008  .010133334   -.4216229
100020 2009   .09823333     .201956
100020 2010  .016058333   .09580792
100021 2005      -.0768  -.01277183
100021 2006  -.05961667   .08855005
100021 2007 -.017191667 -.010829791
100021 2008  -.04696666   -.4216229
100021 2009 -.023466665     .201956
100021 2010 -.014741667   .09580792
100024 2006    .9799833   .08855005
100024 2009   .51533335     .201956
100024 2010  .014058333   .09580792
100026 2004   17.870058    .0472585
100026 2005       -5.09  -.01277183
100026 2006    9.523084   .08855005
100026 2007  -2.4865916 -.010829791
100026 2008   .54033333   -.4216229
100026 2009   1.4377333     .201956
100026 2010    .4943583   .09580792
100027 2003   1.0200499    .2237602
100027 2004    .8740584    .0472585
100027 2005       .7555  -.01277183
100027 2006   .16448334   .08855005
100029 2006  -.16821668   .08855005
100029 2007   .27910832 -.010829791
100029 2008   .07793334   -.4216229
100029 2009  .005633332     .201956
100030 2005      -.0131  -.01277183
100030 2006  .024183333   .08855005
100030 2007   .02750833 -.010829791
100030 2008   .05113334   -.4216229
100031 2005        -.02  -.01277183
100031 2006  -.03771666   .08855005
100031 2007 -.033691667 -.010829791
100031 2009   .14953333     .201956
100031 2010   .19615833   .09580792
100032 2004   .06425834    .0472585
100032 2005      -.0216  -.01277183
100032 2008   .17953333   -.4216229
100032 2010   .17735833   .09580792
100033 2005    .9403999  -.01277183
100033 2006    .7986833   .08855005
100033 2007   .07110833 -.010829791
100033 2008   .05233334   -.4216229
100033 2009   .03813333     .201956
100033 2010   .23085834   .09580792
100036 2003     1.58285    .2237602
100036 2004    16.50846    .0472585
100036 2005       .0053  -.01277183
100036 2006    .6539834   .08855005
100036 2007   .11130834 -.010829791
end

Sorted by year the data looks as follows:

* Example generated by -dataex-. To install: ssc install dataex
clear
input long mfiid int year float(_roe _sp500)
100324 2003       .13935 .2237602
100118 2003      -.05345 .2237602
100296 2003       .18615 .2237602
100092 2003      -.05935 .2237602
100720 2003       .14455 .2237602
100104 2003      -.02805 .2237602
101107 2003   .015449997 .2237602
100648 2003       .08945 .2237602
100302 2003       .16415 .2237602
100467 2003       .40555 .2237602
100303 2003       .14535 .2237602
100281 2003       .23025 .2237602
100523 2003       .08065 .2237602
100326 2003       .10955 .2237602
100673 2003       .71915 .2237602
100211 2003  -.066250004 .2237602
100096 2003      -.11025 .2237602
100492 2003      -.22445 .2237602
100312 2003       .06345 .2237602
100431 2003       .02765 .2237602
100565 2003       .55315 .2237602
100477 2003       .08275 .2237602
100713 2003      -.04965 .2237602
100547 2003  -.036650002 .2237602
100684 2003       .07855 .2237602
100556 2003       .05095 .2237602
100404 2003       .07495 .2237602
100280 2003       .25795 .2237602
100080 2003       .19575 .2237602
100288 2003       .04715 .2237602
100309 2003       .01885 .2237602
100253 2003     -2.49875 .2237602
100127 2003      -.46735 .2237602
100749 2003  -.012650002 .2237602
100357 2003       .09915 .2237602
100617 2003      -.18485 .2237602
100027 2003    1.0200499 .2237602
100255 2003   .011749998 .2237602
100346 2003       .22665 .2237602
100233 2003   .007949997 .2237602
100348 2003       .08205 .2237602
100350 2003       .05175 .2237602
100813 2003       .14445 .2237602
100741 2003       .14795 .2237602
100155 2003      -.02585 .2237602
100243 2003       .09685 .2237602
100173 2003      -.05955 .2237602
100733 2003  -.011250002 .2237602
100664 2003       .31375 .2237602
100705 2003       .16505 .2237602
100678 2003      -.02375 .2237602
100590 2003      -.22435 .2237602
100378 2003      -.01945 .2237602
100188 2003 -.0014500022 .2237602
100700 2003       .36955 .2237602
100420 2003       .05695 .2237602
100500 2003       .04465 .2237602
100402 2003       .23905 .2237602
100298 2003       .19165 .2237602
100716 2003       .19895 .2237602
100542 2003    .26064998 .2237602
100339 2003       .39035 .2237602
100641 2003      -.02645 .2237602
100168 2003  -.072850004 .2237602
100158 2003      -.27875 .2237602
100097 2003       .39195 .2237602
100611 2003   -.28904998 .2237602
100512 2003       .46595 .2237602
100299 2003       .50835 .2237602
100225 2003       .50175 .2237602
100036 2003      1.58285 .2237602
100386 2003       .15465 .2237602
100329 2003       .01035 .2237602
100017 2003       .17895 .2237602
100576 2003   -.29035002 .2237602
100442 2003       .12485 .2237602
100831 2003       .06045 .2237602
100089 2003      -.37675 .2237602
100601 2003    .27354997 .2237602
100769 2003      -.16215 .2237602
100747 2003       .11985 .2237602
100515 2003       .08925 .2237602
100222 2003      -.06445 .2237602
100765 2003      -.02665 .2237602
100681 2003      -.03485 .2237602
100217 2003   .005849998 .2237602
100098 2003      -.08055 .2237602
100540 2003   .018349998 .2237602
100464 2003       .01055 .2237602
100660 2003  -.005450003 .2237602
100574 2003      -.02425 .2237602
100236 2003       .17965 .2237602
100205 2003  -.018150002 .2237602
100369 2003      -.02675 .2237602
100714 2003       .33685 .2237602
100497 2003      -.05555 .2237602
100376 2003       .11915 .2237602
100325 2003       .00925 .2237602
100666 2003   .009949997 .2237602
100242 2003       .11325 .2237602
end

Help is highly appreciated. Thanks a lot in advance!

Kind regards,
Rafael

[Clyde Schechter]

From your description, I can't tell if you want to do a separate regression for each year, with the various mfiid's being the different observations in the regression sample, or the other way around: a separate regression for each mfiid, with the years giving the different observations in the regression sample. For illustrative purposes, I'll assume the latter:

rangestat (reg) _roe _sp500, by(mfiid) interval(year . .)

-rangestat- is written by Robert Picard, Nick Cox, and Roberto Ferrer; it is available from SSC.

[Rafael Maschke]

Dear Clyde Schechter,

thanks a lot for your response and this very useful code. The assumption you made correctly applies to my project. Sorry for being unclear.

I have also used this code in order to determine the beta of two groups within the panel data (profit_status):

 rangestat (reg) _roe _sp500, by(profit_status) interval(year . .)

I have a question regarding the beta estimates I would like to ask you. Does the resulting beta for each Non-profit and Profit subgroup over the years take into account the panel data gaps in such a way that mfiid's that are subject to a shorter or longer time period add their corresponding weight to the beta estimate(s)? I am not sure if this makes sense, but I reckoned that beta's with shorter time intervals, say 3 years, should give a lower weight to the average (e.g.) "Profit" beta estimate than beta's with 8 years of data.

* Example generated by -dataex-. To install: ssc install dataex
clear
input int year long mfiid str10 profit_status float(_roe _sp500) double(mfiid_beta profitstatus_beta)
2007 100000 "Non-profit"    .2913083 -.010829791 -.01230429498659591  .0676808639326907
2008 100000 "Non-profit"   .17713334   -.4216229 -.01230429498659591  .0676808639326907
2009 100000 "Non-profit"   .14123334     .201956 -.01230429498659591  .0676808639326907
2004 100001 "Non-profit"   .04955833    .0472585 -.43151791583880483  .0676808639326907
2005 100001 "Non-profit" -.008699998  -.01277183 -.43151791583880483  .0676808639326907
2006 100001 "Non-profit"   .02328333   .08855005 -.43151791583880483  .0676808639326907
2007 100001 "Non-profit"   .21060835 -.010829791 -.43151791583880483  .0676808639326907
2009 100001 "Non-profit"  .015933335     .201956 -.43151791583880483  .0676808639326907
2010 100001 "Non-profit"  .017058332   .09580792 -.43151791583880483  .0676808639326907
2008 100004 "Profit"       .20683333   -.4216229                   . .22769555983445572
2006 100008 "Non-profit"   -.4828166   .08855005   15.21401296646321  .0676808639326907
2007 100008 "Non-profit"   -.4197917 -.010829791   15.21401296646321  .0676808639326907
2008 100008 "Non-profit"  -2.9300666   -.4216229   15.21401296646321  .0676808639326907
2009 100008 "Non-profit"    9.721633     .201956   15.21401296646321  .0676808639326907
2005 100012 "Non-profit"       .3659  -.01277183 -1.3618103679868323  .0676808639326907
2006 100012 "Non-profit"   -.4431167   .08855005 -1.3618103679868323  .0676808639326907
2007 100012 "Non-profit"   1.6969082 -.010829791 -1.3618103679868323  .0676808639326907
2008 100012 "Non-profit"   1.2119334   -.4216229 -1.3618103679868323  .0676808639326907
2009 100012 "Non-profit"    .9820334     .201956 -1.3618103679868323  .0676808639326907
2010 100012 "Non-profit"   .16185834   .09580792 -1.3618103679868323  .0676808639326907
2009 100013 "Non-profit"  -.10906667     .201956                   .  .0676808639326907
2010 100013 "Non-profit"  -.13484167   .09580792                   .  .0676808639326907
2005 100016 "Profit"          -.0101  -.01277183 .015465619825001544 .22769555983445572
2006 100016 "Profit"     .0035833344   .08855005 .015465619825001544 .22769555983445572
2007 100016 "Profit"      .016408335 -.010829791 .015465619825001544 .22769555983445572
2008 100016 "Profit"       .05173333   -.4216229 .015465619825001544 .22769555983445572
2009 100016 "Profit"       .07153333     .201956 .015465619825001544 .22769555983445572
2010 100016 "Profit"       .08925833   .09580792 .015465619825001544 .22769555983445572
2003 100017 "Profit"          .17895    .2237602  .18605594287548533 .22769555983445572
2004 100017 "Profit"       .16935833    .0472585  .18605594287548533 .22769555983445572
2006 100017 "Profit"       .09988333   .08855005  .18605594287548533 .22769555983445572
2007 100020 "Profit"      .003608331 -.010829791  .09572634580600582 .22769555983445572
2008 100020 "Profit"      .010133334   -.4216229  .09572634580600582 .22769555983445572
2009 100020 "Profit"       .09823333     .201956  .09572634580600582 .22769555983445572
2010 100020 "Profit"      .016058333   .09580792  .09572634580600582 .22769555983445572
2005 100021 "Profit"          -.0768  -.01277183 .030577672464920214 .22769555983445572
2006 100021 "Profit"      -.05961667   .08855005 .030577672464920214 .22769555983445572
2007 100021 "Profit"     -.017191667 -.010829791 .030577672464920214 .22769555983445572
2008 100021 "Profit"      -.04696666   -.4216229 .030577672464920214 .22769555983445572
2009 100021 "Profit"     -.023466665     .201956 .030577672464920214 .22769555983445572
2010 100021 "Profit"     -.014741667   .09580792 .030577672464920214 .22769555983445572
2006 100024 "Non-profit"    .9799833   .08855005  -.2686100941367364  .0676808639326907
2009 100024 "Non-profit"   .51533335     .201956  -.2686100941367364  .0676808639326907
2010 100024 "Non-profit"  .014058333   .09580792  -.2686100941367364  .0676808639326907
2004 100026 "Non-profit"   17.870058    .0472585   8.093220141687869  .0676808639326907
2005 100026 "Non-profit"       -5.09  -.01277183   8.093220141687869  .0676808639326907
2006 100026 "Non-profit"    9.523084   .08855005   8.093220141687869  .0676808639326907
2007 100026 "Non-profit"  -2.4865916 -.010829791   8.093220141687869  .0676808639326907
2008 100026 "Non-profit"   .54033333   -.4216229   8.093220141687869  .0676808639326907
2009 100026 "Non-profit"   1.4377333     .201956   8.093220141687869  .0676808639326907
2010 100026 "Non-profit"    .4943583   .09580792   8.093220141687869  .0676808639326907
2003 100027 "Profit"       1.0200499    .2237602  1.0082801249665767 .22769555983445572
2004 100027 "Profit"        .8740584    .0472585  1.0082801249665767 .22769555983445572
2005 100027 "Profit"           .7555  -.01277183  1.0082801249665767 .22769555983445572
2006 100027 "Profit"       .16448334   .08855005  1.0082801249665767 .22769555983445572
2006 100029 "Profit"      -.16821668   .08855005  -.1929731451801049 .22769555983445572
2007 100029 "Profit"       .27910832 -.010829791  -.1929731451801049 .22769555983445572
2008 100029 "Profit"       .07793334   -.4216229  -.1929731451801049 .22769555983445572
2009 100029 "Profit"      .005633332     .201956  -.1929731451801049 .22769555983445572
2005 100030 "Profit"          -.0131  -.01277183 -.07494323541065033 .22769555983445572
2006 100030 "Profit"      .024183333   .08855005 -.07494323541065033 .22769555983445572
2007 100030 "Profit"       .02750833 -.010829791 -.07494323541065033 .22769555983445572
2008 100030 "Profit"       .05113334   -.4216229 -.07494323541065033 .22769555983445572
2005 100031 "Non-profit"        -.02  -.01277183   .8757502637805205  .0676808639326907
2006 100031 "Non-profit"  -.03771666   .08855005   .8757502637805205  .0676808639326907
2007 100031 "Non-profit" -.033691667 -.010829791   .8757502637805205  .0676808639326907
2009 100031 "Non-profit"   .14953333     .201956   .8757502637805205  .0676808639326907
2010 100031 "Non-profit"   .19615833   .09580792   .8757502637805205  .0676808639326907
2004 100032 "Profit"       .06425834    .0472585  -.1563752523359082 .22769555983445572
2005 100032 "Profit"          -.0216  -.01277183  -.1563752523359082 .22769555983445572
2008 100032 "Profit"       .17953333   -.4216229  -.1563752523359082 .22769555983445572
2010 100032 "Profit"       .17735833   .09580792  -.1563752523359082 .22769555983445572
2005 100033 "Profit"        .9403999  -.01277183  .36818820953588766 .22769555983445572
2006 100033 "Profit"        .7986833   .08855005  .36818820953588766 .22769555983445572
2007 100033 "Profit"       .07110833 -.010829791  .36818820953588766 .22769555983445572
2008 100033 "Profit"       .05233334   -.4216229  .36818820953588766 .22769555983445572
2009 100033 "Profit"       .03813333     .201956  .36818820953588766 .22769555983445572
2010 100033 "Profit"       .23085834   .09580792  .36818820953588766 .22769555983445572
2003 100036 "Profit"         1.58285    .2237602  2.4887138330042635 .22769555983445572
2004 100036 "Profit"        16.50846    .0472585  2.4887138330042635 .22769555983445572
2005 100036 "Profit"           .0053  -.01277183  2.4887138330042635 .22769555983445572
2006 100036 "Profit"        .6539834   .08855005  2.4887138330042635 .22769555983445572
2007 100036 "Profit"       .11130834 -.010829791  2.4887138330042635 .22769555983445572
2008 100036 "Profit"     -.025066666   -.4216229  2.4887138330042635 .22769555983445572
2009 100036 "Profit"     -.006966667     .201956  2.4887138330042635 .22769555983445572
2010 100036 "Profit"       .03235833   .09580792  2.4887138330042635 .22769555983445572
2003 100039 "Profit"         -.15105    .2237602 -1.3043965932947752 .22769555983445572
2004 100039 "Profit"      -.10604167    .0472585 -1.3043965932947752 .22769555983445572
2005 100039 "Profit"           .2303  -.01277183 -1.3043965932947752 .22769555983445572
2006 100039 "Profit"     -.008116666   .08855005 -1.3043965932947752 .22769555983445572
2005 100041 "Non-profit"       .6233  -.01277183   .7659891154256351  .0676808639326907
2006 100041 "Non-profit"    .9760834   .08855005   .7659891154256351  .0676808639326907
2007 100041 "Non-profit"   .25010833 -.010829791   .7659891154256351  .0676808639326907
2008 100041 "Non-profit"   .36893335   -.4216229   .7659891154256351  .0676808639326907
2008 100042 "Profit"      -.22946666   -.4216229 -.19658897029645822 .22769555983445572
2009 100042 "Profit"      -.52516663     .201956 -.19658897029645822 .22769555983445572
2010 100042 "Profit"      -.02404167   .09580792 -.19658897029645822 .22769555983445572
2004 100043 "Profit"       -6.986542    .0472585 -13.055812227775231 .22769555983445572
2005 100043 "Profit"           .2743  -.01277183 -13.055812227775231 .22769555983445572
2006 100043 "Profit"       1.2188834   .08855005 -13.055812227775231 .22769555983445572
end
format %ty year

Kind regards,
Rafael

[Rafael Maschke]

Dear Clyde Schechter,

thanks a lot for your response and this very useful code. The assumption you made correctly applies to my project. Sorry for being unclear.

I have also used this code in order to determine the beta of two groups within the panel data (profit_status):

 rangestat (reg) _roe _sp500, by(profit_status) interval(year . .)

I have a question regarding the beta estimates I would like to ask you. Does the resulting beta for each Non-profit and Profit subgroup over the years take into account the panel data gaps in such a way that mfiid's that are subject to a shorter or longer time period add their corresponding weight to the beta estimate(s)? I am not sure if this makes sense, but I reckoned that beta's with shorter time intervals, say 3 years, should give a lower weight to the average (e.g.) "Profit" beta estimate than beta's with 8 years of data.

* Example generated by -dataex-. To install: ssc install dataex
clear
input int year long mfiid str10 profit_status float(_roe _sp500) double(mfiid_beta profitstatus_beta)
2007 100000 "Non-profit" .2913083 -.010829791 -.01230429498659591 .0676808639326907
2008 100000 "Non-profit" .17713334 -.4216229 -.01230429498659591 .0676808639326907
2009 100000 "Non-profit" .14123334 .201956 -.01230429498659591 .0676808639326907
2004 100001 "Non-profit" .04955833 .0472585 -.43151791583880483 .0676808639326907
2005 100001 "Non-profit" -.008699998 -.01277183 -.43151791583880483 .0676808639326907
2006 100001 "Non-profit" .02328333 .08855005 -.43151791583880483 .0676808639326907
2007 100001 "Non-profit" .21060835 -.010829791 -.43151791583880483 .0676808639326907
2009 100001 "Non-profit" .015933335 .201956 -.43151791583880483 .0676808639326907
2010 100001 "Non-profit" .017058332 .09580792 -.43151791583880483 .0676808639326907
2008 100004 "Profit" .20683333 -.4216229 . .22769555983445572
2006 100008 "Non-profit" -.4828166 .08855005 15.21401296646321 .0676808639326907
2007 100008 "Non-profit" -.4197917 -.010829791 15.21401296646321 .0676808639326907
2008 100008 "Non-profit" -2.9300666 -.4216229 15.21401296646321 .0676808639326907
2009 100008 "Non-profit" 9.721633 .201956 15.21401296646321 .0676808639326907
2005 100012 "Non-profit" .3659 -.01277183 -1.3618103679868323 .0676808639326907
2006 100012 "Non-profit" -.4431167 .08855005 -1.3618103679868323 .0676808639326907
2007 100012 "Non-profit" 1.6969082 -.010829791 -1.3618103679868323 .0676808639326907
2008 100012 "Non-profit" 1.2119334 -.4216229 -1.3618103679868323 .0676808639326907
2009 100012 "Non-profit" .9820334 .201956 -1.3618103679868323 .0676808639326907
2010 100012 "Non-profit" .16185834 .09580792 -1.3618103679868323 .0676808639326907
2009 100013 "Non-profit" -.10906667 .201956 . .0676808639326907
2010 100013 "Non-profit" -.13484167 .09580792 . .0676808639326907
2005 100016 "Profit" -.0101 -.01277183 .015465619825001544 .22769555983445572
2006 100016 "Profit" .0035833344 .08855005 .015465619825001544 .22769555983445572
2007 100016 "Profit" .016408335 -.010829791 .015465619825001544 .22769555983445572
2008 100016 "Profit" .05173333 -.4216229 .015465619825001544 .22769555983445572
2009 100016 "Profit" .07153333 .201956 .015465619825001544 .22769555983445572
2010 100016 "Profit" .08925833 .09580792 .015465619825001544 .22769555983445572
2003 100017 "Profit" .17895 .2237602 .18605594287548533 .22769555983445572
2004 100017 "Profit" .16935833 .0472585 .18605594287548533 .22769555983445572
2006 100017 "Profit" .09988333 .08855005 .18605594287548533 .22769555983445572
2007 100020 "Profit" .003608331 -.010829791 .09572634580600582 .22769555983445572
2008 100020 "Profit" .010133334 -.4216229 .09572634580600582 .22769555983445572
2009 100020 "Profit" .09823333 .201956 .09572634580600582 .22769555983445572
2010 100020 "Profit" .016058333 .09580792 .09572634580600582 .22769555983445572
2005 100021 "Profit" -.0768 -.01277183 .030577672464920214 .22769555983445572
2006 100021 "Profit" -.05961667 .08855005 .030577672464920214 .22769555983445572
2007 100021 "Profit" -.017191667 -.010829791 .030577672464920214 .22769555983445572
2008 100021 "Profit" -.04696666 -.4216229 .030577672464920214 .22769555983445572
2009 100021 "Profit" -.023466665 .201956 .030577672464920214 .22769555983445572
2010 100021 "Profit" -.014741667 .09580792 .030577672464920214 .22769555983445572
2006 100024 "Non-profit" .9799833 .08855005 -.2686100941367364 .0676808639326907
2009 100024 "Non-profit" .51533335 .201956 -.2686100941367364 .0676808639326907
2010 100024 "Non-profit" .014058333 .09580792 -.2686100941367364 .0676808639326907
2004 100026 "Non-profit" 17.870058 .0472585 8.093220141687869 .0676808639326907
2005 100026 "Non-profit" -5.09 -.01277183 8.093220141687869 .0676808639326907
2006 100026 "Non-profit" 9.523084 .08855005 8.093220141687869 .0676808639326907
2007 100026 "Non-profit" -2.4865916 -.010829791 8.093220141687869 .0676808639326907
2008 100026 "Non-profit" .54033333 -.4216229 8.093220141687869 .0676808639326907
2009 100026 "Non-profit" 1.4377333 .201956 8.093220141687869 .0676808639326907
2010 100026 "Non-profit" .4943583 .09580792 8.093220141687869 .0676808639326907
2003 100027 "Profit" 1.0200499 .2237602 1.0082801249665767 .22769555983445572
2004 100027 "Profit" .8740584 .0472585 1.0082801249665767 .22769555983445572
2005 100027 "Profit" .7555 -.01277183 1.0082801249665767 .22769555983445572
2006 100027 "Profit" .16448334 .08855005 1.0082801249665767 .22769555983445572
2006 100029 "Profit" -.16821668 .08855005 -.1929731451801049 .22769555983445572
2007 100029 "Profit" .27910832 -.010829791 -.1929731451801049 .22769555983445572
2008 100029 "Profit" .07793334 -.4216229 -.1929731451801049 .22769555983445572
2009 100029 "Profit" .005633332 .201956 -.1929731451801049 .22769555983445572
2005 100030 "Profit" -.0131 -.01277183 -.07494323541065033 .22769555983445572
2006 100030 "Profit" .024183333 .08855005 -.07494323541065033 .22769555983445572
2007 100030 "Profit" .02750833 -.010829791 -.07494323541065033 .22769555983445572
2008 100030 "Profit" .05113334 -.4216229 -.07494323541065033 .22769555983445572
2005 100031 "Non-profit" -.02 -.01277183 .8757502637805205 .0676808639326907
2006 100031 "Non-profit" -.03771666 .08855005 .8757502637805205 .0676808639326907
2007 100031 "Non-profit" -.033691667 -.010829791 .8757502637805205 .0676808639326907
2009 100031 "Non-profit" .14953333 .201956 .8757502637805205 .0676808639326907
2010 100031 "Non-profit" .19615833 .09580792 .8757502637805205 .0676808639326907
2004 100032 "Profit" .06425834 .0472585 -.1563752523359082 .22769555983445572
2005 100032 "Profit" -.0216 -.01277183 -.1563752523359082 .22769555983445572
2008 100032 "Profit" .17953333 -.4216229 -.1563752523359082 .22769555983445572
2010 100032 "Profit" .17735833 .09580792 -.1563752523359082 .22769555983445572
2005 100033 "Profit" .9403999 -.01277183 .36818820953588766 .22769555983445572
2006 100033 "Profit" .7986833 .08855005 .36818820953588766 .22769555983445572
2007 100033 "Profit" .07110833 -.010829791 .36818820953588766 .22769555983445572
2008 100033 "Profit" .05233334 -.4216229 .36818820953588766 .22769555983445572
2009 100033 "Profit" .03813333 .201956 .36818820953588766 .22769555983445572
2010 100033 "Profit" .23085834 .09580792 .36818820953588766 .22769555983445572
2003 100036 "Profit" 1.58285 .2237602 2.4887138330042635 .22769555983445572
2004 100036 "Profit" 16.50846 .0472585 2.4887138330042635 .22769555983445572
2005 100036 "Profit" .0053 -.01277183 2.4887138330042635 .22769555983445572
2006 100036 "Profit" .6539834 .08855005 2.4887138330042635 .22769555983445572
2007 100036 "Profit" .11130834 -.010829791 2.4887138330042635 .22769555983445572
2008 100036 "Profit" -.025066666 -.4216229 2.4887138330042635 .22769555983445572
2009 100036 "Profit" -.006966667 .201956 2.4887138330042635 .22769555983445572
2010 100036 "Profit" .03235833 .09580792 2.4887138330042635 .22769555983445572
2003 100039 "Profit" -.15105 .2237602 -1.3043965932947752 .22769555983445572
2004 100039 "Profit" -.10604167 .0472585 -1.3043965932947752 .22769555983445572
2005 100039 "Profit" .2303 -.01277183 -1.3043965932947752 .22769555983445572
2006 100039 "Profit" -.008116666 .08855005 -1.3043965932947752 .22769555983445572
2005 100041 "Non-profit" .6233 -.01277183 .7659891154256351 .0676808639326907
2006 100041 "Non-profit" .9760834 .08855005 .7659891154256351 .0676808639326907
2007 100041 "Non-profit" .25010833 -.010829791 .7659891154256351 .0676808639326907
2008 100041 "Non-profit" .36893335 -.4216229 .7659891154256351 .0676808639326907
2008 100042 "Profit" -.22946666 -.4216229 -.19658897029645822 .22769555983445572
2009 100042 "Profit" -.52516663 .201956 -.19658897029645822 .22769555983445572
2010 100042 "Profit" -.02404167 .09580792 -.19658897029645822 .22769555983445572
2004 100043 "Profit" -6.986542 .0472585 -13.055812227775231 .22769555983445572
2005 100043 "Profit" .2743 -.01277183 -13.055812227775231 .22769555983445572
2006 100043 "Profit" 1.2188834 .08855005 -13.055812227775231 .22769555983445572
end
format %ty year

Kind regards,
Rafael

1. If I use the rangestat code above by profit_status I get the same results as with a simple linear regression:

reg _roe _sp500 if profit_status == "Profit"

=

rangestat (reg) _roe _sp500, by(profit_status) interval(year . .)

2. I also tried to follow -help rangestat- and added a by option at the end of the code [options], but I get the error message "by not allowed":

rangestat (reg) _roe _sp500, by(mfiid) interval(year . .) by profit_status

3. I have tried this code, which gives me the right estimates for Profit-mfiids:

 rangestat (reg) _roe _sp500 if profit_status == "Profit", by(mfiid) interval(year . .)

Is there a way to execute step three for both profit_status groups at once?

[Clyde Schechter]

Yes. Your error was in trying to specify -by()- twice (and, on top of that, omitting the parentheses the second time)..

rangestat (reg) _roe _sp500 by(mfiid profit_status) interval(year . .)

As for your question in #3, I'm not sure what kind of "weighting" you are talking about here. Each regression is carried out with exactly the number of observations that meet the conditions specified by -by()- and -interval()- and have non-missing values for _roe and _sp500. No more and no less. The results you get are the same as you would get if you did each of those regressions separately with corresponding -if- conditions. It's just normal regression. And that is what is usually wanted here. But it is not, in the formal statistical sense of the term, "weighting." If you do have some kind of weighting, in the statistical sense of the term, in mind then please post back with a full explanation.

[Rafael Maschke]

Thanks a lot for helping me out Clyde!