Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Asreg rolling window returns missing values

    Hi,

    I am dealing with missing values when using asreg. I have read Attaullah Shah's explanation of the package but I still don't know where I'm messing up.

    Goal:
    My aim is to calculate abnormal cash flows from operation (CFO). In order to do so I must first run a regression on CFO of each industry-year. Then I subtract those coefficients from the actual CFO for every firm.

    Stata:
    Run a cross-sectional regression for each industry-year and save the coefficients.

    I run the following commands
    bys sic : asreg CFO scassets scsales scdsales, robust window(year 5)
    list year _b_cons _b_scassets _b_scsales _b_scdsales in 1/25

    Problem:
    My issue is the following: whenever I run the command I get missing values in certain years.

    Code:
    * Example generated by -dataex-. To install: ssc install dataex
    clear
    input long gvkey double year int sic str10 rating float(POST treated treated2 treated3 PROD DEXP CFO) double at float(scsales scassets)
      3362 2007 2000 "BBB+" 0 0 0 1  .6884373  .1846732   .00862659  13682.5  .9805838 .00008449157
      3362 2013 2000 "BBB-" 1 0 0 1  .6505319  .1247372   .07601947  19366.4  .8675491 .00004900687
      3362 2012 2000 "BBB"  1 1 0 1 1.0465395   .222297   .12342355  20405.3 1.3539184 .00008739808
      3362 2010 2000 "BBB"  1 1 0 1  .7840263  .1755921   .11520702  11408.7 1.0481428 .00008519339
      3362 2006 2000 "BBB+" 0 0 0 1  .8010228 .19678298   .07781902  11835.5  .9937474 .00008248783
      3362 2011 2000 "BBB"  1 1 0 1  .8861746  .2145906   .09221033  11441.9 1.1624988  .0000876524
      3362 2015 2000 "BBB-" 1 0 0 1  .4280934  .1331361  .068828315  13390.6  .6637081 .00005700539
      3362 2008 2000 "BBB+" 0 0 0 1  .7063987 .15496437  .009062671  11073.3  .9304732 .00007308606
      3362 2009 2000 "BBB"  0 1 0 1  .7453424  .2042571   .13299559    11738 1.0908582 .00009030732
      3362 2016 2000 "BBB"  1 1 0 1  .3398802 .11188445   .08778546  10096.3 .58450705 .00007467925
      3362 2014 2000 "BBB-" 1 0 0 1  .6097158 .11453342     .076452  17542.2   .817519 .00005163582
     10793 2012 2011 "BBB-" 1 0 0 1 2.7893596 .12970825   .10721705    11896  3.005871 .00009032608
      5709 2014 2011 "A"    1 0 0 0 1.5670245  .1607948    .1519319 5455.619  1.895135 .00020342237
      5709 2002 2011 "A"    0 0 0 0  .6962887 .17685835   .07717835 2220.196  .9173973  .0002346096
     10793 2014 2011 "BBB"  1 1 0 1  2.853248 .15414305   .09673975    23956  3.086146 .00008212203
      5709 2013 2011 "A"    1 0 0 0  1.584465   .162017   .13974798  4915.88 1.9175546  .0002191077
      5709 2009 2011 "A"    0 0 0 0 1.4504447 .18971063    .1522946 3692.055 1.8066428 .00027651267
      5709 2004 2011 "A"    0 0 0 0  1.501653  .3585828   .12315717 2533.968 1.9973395  .0004178644
      5709 2016 2011 "A"    1 0 0 0 1.1768237  .1809128   .16170608 6370.067 1.5510564 .00016287093
     10793 2010 2011 "BB"   1 0 1 0  4.707277 1.0666528    .9854583    10752 20.489824  .0007207114
      5709 2005 2011 "A"    0 0 0 0  1.612697  .3823722   .16954516 2822.406 2.1365688   .000394638
      5709 2011 2011 "A"    1 0 0 0  1.610507 .18828353   .12098888 4244.391 1.9475206 .00024667496
      5709 2015 2011 "A"    1 0 0 0 1.3289115  .1671508    .1818294 6139.831 1.6980407 .00018329726
      5709 2006 2011 "A"    0 0 0 0 1.5314457  .3686181     .115995 3060.306  2.035668 .00035430765
     10793 2016 2011 "BBB"  1 1 0 1 1.3624152 .09394018   .11806642    22373  1.603243  .0000434707
     10793 2011 2011 "BB+"  1 0 0 0  2.783575 .14025298   .09728423    11071   3.00093 .00009300595
      5709 2010 2011 "A"    1 0 0 0 1.6029158  .2018369   .13150725 4053.918 1.9557452 .00027085186
     10793 2015 2011 "BBB"  1 1 0 1 1.5148606 .11362498   .10728002    23004 1.7270412  .0000417432
      5709 2012 2011 "A"    1 0 0 0 1.6124865 .17412816   .12199182 4563.966 1.9391875 .00023560507
      5709 2003 2011 "A"    0 0 0 0 1.4182874   .362713    .1140665 2393.121 1.8918726  .0004504107
     10793 2013 2011 "BBB"  1 1 0 1  2.650723 .13349025    .1104573    12177  2.889543 .00008406187
      5709 2007 2011 "A"    0 0 0 0 1.5448792  .3423619   .11033504  3393.65 2.0236642  .0003267647
      5709 2008 2011 "A"    0 0 0 0 1.5453984  .3340091   .08032414 3616.471 1.9904536   .000294668
     12785 2014 2015 "BB"   1 0 1 0 2.2155902 .06203312    .3362411 3119.063  2.705636 .00031521855
    170527 2016 2015 "B"    1 0 0 0  .3650492 .11268491   .05448787   9360.6  .5451824 .00010845516
    170527 2012 2015 "B+"   1 0 0 0 .26878026  .3094643    .1124473   2732.3  .7487896   .000780884
     12785 2015 2015 "BB"   1 0 1 0   2.17523 .06818747    .3131799 3318.443  2.622616  .0003206091
    170527 2013 2015 "B"    1 0 0 0 .21201918 .14998353   .04362625   3473.8  .3784724   .000365992
     12785 2012 2015 "B"    1 0 0 0  2.475329 .05908879   .06296294 2913.869  2.561546  .0003154077
     12785 2016 2015 "BB"   1 0 1 0  2.064814 .06309616    .2276619 3008.218  2.390013 .00030134615
    170527 2014 2015 "B"    1 0 0 0   .557689 .15703264   .05270885   7731.1  .6940814  .0002878692
    170527 2015 2015 "B"    1 0 0 0 .43944585 .11094152   .05841342   9220.4   .601234  .0001293477
     12785 2013 2015 "B"    1 0 0 0 2.4994414 .06510759    .3015005 3172.402  2.886591  .0003431863
     62655 2011 2020 "B+"   1 0 0 0 1.2303208  .3467688   .05646711 5754.363 1.6408243 .00012568076
     62655 2009 2020 "BB-"  0 0 0 0 1.1139048  .3725135   .09355895 7843.941  1.584955 .00014204158
     62655 2016 2020 "BB-"  1 0 0 0  2.215955  .6891779   .10182416 2606.227  3.049913  .0003955673
     62655 2015 2020 "BB-"  1 0 0 0 2.1660056   .641507   .14736702 2528.015  2.932393  .0003610583
    180833 2014 2020 "BBB"  1 1 0 1  .4690999  .7117527    .2283757   3776.5  1.269192 .00028784433
    124554 2003 2020 "B"    0 0 0 0  .3330927  .2638978   .05275728  158.494  .5735661  .0044558514
    124554 2004 2020 "B"    0 0 0 0  .4485659  .3648719 -.003337666  148.374  .7304882   .004607934
    180833 2010 2020 "BBB"  1 1 0 1 .54663575  .6991258    .2483698   2293.1 1.5174612  .0004830218
     62655 2010 2020 "B+"   1 0 0 0 1.1269369  .3498856    .0681373 7956.667 1.5455097 .00012748693
    180833 2013 2020 "BBB"  1 1 0 1  .4752931  .6115033      .24756   3474.1   1.28927  .0003069179
    124554 2002 2020 "B"    0 0 0 0 .05437573 .12980792   .00302532  224.424 .24946304  .0014551805
    180833 2009 2020 "BBB"  0 1 0 1  .1045552 .13254008   .05889082   2070.3 .28868347  .0001021346
    124554 2006 2020 "B-"   0 0 0 0   .783308  .3833778 .0005940799  180.477 1.1230181   .004607934
    180833 2012 2020 "BBB"  1 1 0 1  .4732543  .6439931   .25036144   3258.2 1.4100405  .0003614284
    124554 2005 2020 "B-"   0 0 0 0  .7544179  .4105639 -.029331284  183.477 1.0091795   .004607934
     62655 2014 2020 "B+"   1 0 0 0  2.733973  .5958056   .05458371 2769.636 3.3915215  .0003568822
     62655 2013 2020 "B+"   1 0 0 0 1.2050616 .29052427  -.05567715 2802.045  1.585401  .0001758368
    180833 2011 2020 "BBB"  1 1 0 1  .6389604  .9181021    .2759583   2766.8 1.6035062  .0004360909
     62655 2012 2020 "B+"   1 0 0 0 1.4385157  .4618167   .07655791 5687.091  1.991928  .0001737812
    180833 2016 2020 "BBB"  1 1 0 1  .3076461  .5531877   .17298217   4087.7  .9361197  .0002501188
     62655 2008 2020 "BB-"  0 0 0 0  4.707277 2.2161098    .9287363 7040.192 16.111132  .0012935874
    180833 2015 2020 "BBB"  1 1 0 1  .3376936  .6296836    .2409374   3998.1 1.0780617 .00026479544
      2663 2013 2030 "BBB+" 1 0 0 1  .7470138  .3520674     .156049     8323  1.233078 .00015313935
      2663 2016 2030 "BBB+" 1 0 0 1  .5983434  .2635678    .1808629     7837   .984176 .00012362468
      2663 2011 2030 "A"    1 0 0 0  .7001275  .3774697   .18196303     6862 1.2299235 .00015933716
    185334 2016 2030 "BB-"  1 0 0 0  .4022305  .0783853   .09129146 6739.645 .58574706   .000187263
      2663 2014 2030 "BBB+" 1 0 0 1  .6167247 .25964195   .10801394     8113  .9933918 .00012014899
      2663 2012 2030 "A-"   1 0 0 0  .6417954  .3478578   .16321772     6530 1.1231419  .0001457301
    185334 2014 2030 "B+"   1 0 0 0  .3582445 .06357781   .10838207 5200.945  .5099558 .00019680425
      2663 2015 2030 "BBB+" 1 0 0 1  .6123505 .25391346    .1456921     8089   .996179 .00012325897
    185334 2013 2030 "B"    1 0 0 0 .39567265  .0770777   .05960062 5081.191  .5594338 .00022727334
    185334 2012 2030 "B"    1 0 0 0  .4053441 .07759278   .04962522 4399.988  .6063275 .00024463635
      2663 2010 2030 "A"    1 0 0 0 1.2898895   .737967     .321795     6276 2.3368952  .0003044418
    185334 2015 2030 "BB-"  1 0 0 0  .3608185 .06460672  .071700625 5340.083 .51063645 .00019227275
      9777 2016 2033 "BBB"  1 1 0 1 .27188894 .10135072   .06625334  15639.7  .4622406 .00006256217
      9777 2013 2033 "BBB+" 1 0 0 1  .3770234 .12599924   .09477624   9072.1  .6212051  .0001107199
      9777 2015 2033 "BBB"  1 1 0 1 .25738332 .10324831   .08637887  15984.1  .4626775 .00005923258
      9777 2014 2033 "BBB+" 1 0 0 1  .4129033 .12912115   .08081921  16882.6  .6274953 .00011022806
      9777 2011 2033 "BBB+" 1 0 0 1  .5050382  .1532099   .10824535 9115.226  .8183287 .00014809283
      9777 2012 2033 "BBB+" 1 0 0 1  .4057852 .12401228   .09388687   9031.8  .6470163 .00010970655
      5071 2007 2040 "BBB+" 0 0 0 1    .46685 .18964474   .09513308  19041.6  .7507754  .0000549934
      6375 2012 2040 "BBB+" 1 0 0 1  .7143937  .4295437   .14771868    15184  1.192925 .00008402655
      5071 2015 2040 "BBB+" 1 0 0 1  .4490155  .1864554   .11972956  21712.3   .754085 .00004552801
      6375 2009 2040 "BBB+" 0 0 0 1  .6134661  .4222547    .1501005    11200 1.1488215 .00009135757
      5071 2008 2040 "BBB+" 0 0 0 1  .4718196 .20484097   .09601084  17874.8  .7715371  .0000525166
      5071 2006 2040 "BBB+" 0 0 0 1 .42033285 .17158236   .09694073    18184  .6833636 .00005492393
      5071 2013 2040 "BBB+" 1 0 0 1 .48308325 .20071498   .11214582  23145.7  .7904316 .00004413452
      5071 2005 2040 "BBB+" 0 0 0 1  .3621167  .1857633   .09802945    18207  .6443042  .0000553526
      5071 2014 2040 "BBB+" 1 0 0 1  .4751768 .18835896   .10986058  21964.5  .7617095 .00004320457
      6375 2005 2040 "BBB+" 0 0 0 1  .4787403  .3571693    .1059553  10574.5  .9431717 .00009267497
      6375 2011 2040 "BBB+" 1 0 0 1  .6281759  .4032244   .13463324    11901 1.1140373 .00008440956
      5071 2004 2040 "BBB+" 0 0 0 1  .3455659 .16603425   .09274718    18066   .609497 .00005420642
      5071 2011 2040 "BBB+" 1 0 0 1  .5323409  .2425018   .12862459  21096.8  .8920132 .00005354896
      6375 2014 2040 "BBB+" 1 0 0 1  .5732842  .3335918   .11587179    15153  .9424195 .00006462453
      5071 2012 2040 "BBB+" 1 0 0 1  .5124901 .22031303   .13869402    22658  .8425022 .00004740055
      6375 2015 2040 "BBB"  1 1 0 1  .5225368  .3028443   .11159506    15265  .8890648 .00006599353
      6375 2007 2040 "BBB+" 0 0 0 1  .5882117  .4171178   .14028373    11397 1.0991226 .00009333582
    end
    These are the results I get from the asreg command (I'm sorry I don't know how to show this with dataex)

    . list year _b_cons _b_scassets _b_scsales _b_scdsales in 1/25

    +----------------------------------------------------------+
    | year _b_cons _b_scass~s _b_scsales _b_scdsa~s |
    |----------------------------------------------------------|
    1. | 2007 . . . . |
    2. | 2013 -.04137871 -1454.5413 .26266831 -.35338505 |
    3. | 2012 -.38693713 3034.2662 .21646971 -.27079479 |
    4. | 2010 -.85324587 -8938.4823 1.6439263 -.12898214 |
    5. | 2006 . . . . |
    |----------------------------------------------------------|
    6. | 2011 -.16358315 -12266.271 1.2316633 -.93329911 |
    7. | 2015 -.02253046 -724.27971 .16337456 -.08283015 |
    8. | 2008 . . . . |
    9. | 2009 . . . . |
    10. | 2016 .01235643 950.81255 .01738386 .02459045 |
    |----------------------------------------------------------|
    11. | 2014 .00971925 449.31965 .05272683 -.0136679 |
    12. | 2012 .08884997 113.33022 -.0015957 .0451358 |
    13. | 2014 .24156329 -115.83738 -.04908656 .09766839 |
    14. | 2002 . . . . |
    15. | 2014 .24156329 -115.83738 -.04908656 .09766839 |
    |----------------------------------------------------------|
    16. | 2013 -.05566587 438.12642 .04170158 -.0068965 |
    17. | 2009 2.3117199 2416.5953 -1.5396779 .76984911 |
    18. | 2004 . . . . |
    19. | 2016 .1414304 201.14575 -.01674374 -.04160379 |
    20. | 2010 -.69739426 -1098.4617 .6166464 -.54146248 |
    |----------------------------------------------------------|
    21. | 2005 . . . . |
    22. | 2011 .25514846 -104.39973 -.06177096 .11034773 |
    23. | 2015 .16020267 84.320294 -.01398477 -.11785821 |
    24. | 2006 -.04683009 -52.635429 .0792364 .21267886 |
    25. | 2016 .1414304 201.14575 -.01674374 -.04160379 |
    +----------------------------------------------------------+


    Thanks a lot.


  • #2
    The usual story here is that you don’t have enough data points to calculate the regression.

    Note that asreg (from SSC, as you are asked to explain: FAQ Advice #12) reports the number of observations. With 3 predictors I wouldn't feel comfortable with fewer than about 30, but asreg certainly needs enough data points to estimate 4 parameters.
    Last edited by Nick Cox; 04 Jun 2021, 07:15.

    Comment


    • #3
      Originally posted by Nick Cox View Post
      The usual story here is that you don’t have enough data points to calculate the regression.

      Note that asreg (from SSC, as you are asked to explain: FAQ Advice #12) reports the number of observations. With 3 predictors I wouldn't feel comfortable with fewer than about 30, but asreg certainly needs enough data points to estimate 4 parameters.
      Thanks and sorry I thought it wasn't necessary to mention the SSC part. Will do next time.

      Could it be due to something else? Because when I leave out the option robust window(year 5) it doesn't produce any missing values

      Comment


      • #4
        All we can see are your example data. In that there seems a clear pattern: regression results are missing for early years in the panel. I haven't used asreg ever but I find it difficult to imagine that whatever it means by robust can be done with fewer data points.

        Comment

        Working...
        X