Matching based on SIZE and Industry

[Jackson Njoroge]

. dataex ID TIC SIZE PROT Industry_code LN_AUDFEE
----------------------- copy starting from the next line -----------------------

* Example generated by -dataex-. To install: ssc install dataex
clear
input long ID str5 TIC double SIZE byte(PROT Industry_code) double LN_AUDFEE
1675149 "AA"      9.676461473957694 1 3 15.872450834297597
1158114 "AAOI"   6.1459865865441525 0 3 14.049499089318646
 320193 "AAPL"   12.732326637573383 1 3  16.66092562785641
 930245 "AASP"   -4.268697949366879 0 7 10.491274217438248
1069183 "AAXN"    6.578612118944737 1 3  14.00131769970961
1144980 "ABG"     7.899301895335059 0 6 14.580978400103845
 907654 "ABIO"   1.9205923410710235 0 3 11.947949375319936
1605331 "ABQQ"   1.5208252540774967 0 7 10.809727948556782
1070494 "ACAD"    6.291943143684131 0 3  13.71596421829101
1622996 "ACBM"    .3940670631557951 1 3 10.348974334993779
1001463 "ACCA"  -.36096986822161325 0 3  10.75042820627549
 712034 "ACCO"    7.932505719051791 0 3 15.183386333560147
1698991 "ACEL"    6.125704750966459 0 5 11.350406535472453
1069308 "ACER"   3.9251958611383935 1 3 11.635143097694478
1520697 "ACHC"    8.727859869254766 1 7 15.259819077781128
1070336 "ACHN"   5.6271101918253805 1 3 13.129432333358215
1651235 "ACIA"    6.400023198606101 1 3 14.222973668735111
 868857 "ACM"     9.594318342797532 0 7 16.056220247240233
1467373 "ACN"    10.301924017501833 1 7 16.822232226317865
1557746 "ACRS"    5.618827164330305 1 3  13.17115354157376
   2098 "ACU"     4.695714481531775 1 3  12.70076888736628
 786947 "ACUR"     .766397642299538 1 3 11.891286555386333
 796343 "ADBE"    9.839944908681455 1 7 15.736012052418026
1515156 "ADES"    5.073071607141957 1 3 12.936033799212835
1513525 "ADIL"   1.5473497201206787 0 3 11.894780707439258

  //SOME OBSERVATIONS DELETED FOR SPACE PURPOSES

   6176 "AP"      6.347597594247162 0 3 14.320024303672572
end

------------------ copy up to and including the previous line ------------------

Listed 100 out of 1730 observations
Use the count() option to list more

I am new to Stata and I would really appreciate some help here. I have some observations for companies that rotated their auditor partner and their corresponding variables (auditfees, size, etc) I have included a sample which I have omitted some control variables such as ROA, Leverage etc. What I want to do is match the companies that rotated (dummy=1) and those that did not (dummy=0) by industry and size. For size I want closest match (probably the difference shouldn't be more than 0.5)

After matching I would like to produce a report showing the matched sample. Please let me know if I'm missing to include anything that might be helpful in answering the question.

Thanks a lot

[Mike Lacy]

I like the user-written program -calipmatch- for purposes like this (-ssc describe calipmatch-). You don't say which variable in your example represents your "rotation" dummy variable, so in the example code that follows I picked PROT for illustration. Also, in the example you provided, not many companies have a match on size within 0.5, so I used 50 as the caliper width so as to create a reasonable number of matches. I don't know what sort of report you want, so I did not offer any suggestions for that.

ssc install calipmatch
calipmatch, casevar(PROT) generate(pairid) maxmatches(1) ///
   exactmatch(Industry_code) calipermatch(SIZE) caliperwidth(50)

Questions about matching are very common on StataList, and many other solutions are possible besides what I suggested, so searching the archives of StataList for postings on matching might be interesting to you.

[Phil Bromiley]

It is also possible to do the matching manually. If you sort by industry and size, you can then readily identify the next largest or smallest firm in the sample. While this seems like an incredible amount of work, it turns out you can do it relatively quickly.

[Jackson Njoroge]

Thank you Mike and Phil, I will get back to you if I encounter any problems or if I need more assistance. Stay safe.