Dummy varaible in regression -problem with results ouput

Fadi Ansar

Join Date: May 2021
Posts: 120

Dummy varaible in regression -problem with results ouput

28 Dec 2021, 00:21

Hi All,

I have ran a regression model which explians how the effect of bank competiton on bank risk .I am also trying to understand the effect of monetary policy regime (MPR) on bank risk.for the MPR variable I have two category (EMP and TMP).I have treated it as a dummy varaible and also I have used intercations with the same.my problem is that The result which I am getting is showing only either EMP or TMP.I am not able to show both the interactions in single table.Can any one help me on this.I have seen some of the papers as they have put both categories in single table while i was getting 0 in that case.
I am attaching both my reression results on stata as well as excel

Code:

 reg zscore  NIM  lasset CapitalRatio  i.mpr##c.C5_Aggregate i.Year,robust
note: 2019.Year omitted because of collinearity

Linear regression                               Number of obs     =        378
                                                F(19, 358)        =       3.48
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1665
                                                Root MSE          =     .46035

------------------------------------------------------------------------------------
                   |               Robust
            zscore |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------------+----------------------------------------------------------------
               NIM |    .166563   .0505932     3.29   0.001     .0670658    .2660601
            lasset |   .0370991    .027149     1.37   0.173    -.0162923    .0904906
      CapitalRatio |  -.6045754   .5546065    -1.09   0.276    -1.695272    .4861208
                   |
               mpr |
              TMP  |  -2.849336   1.144863    -2.49   0.013    -5.100838   -.5978336
      C5_Aggregate |   -22.2205   5.575988    -3.99   0.000    -33.18631    -11.2547
                   |
mpr#c.C5_Aggregate |
              TMP  |    7.18703   2.908346     2.47   0.014     1.467441    12.90662
                   |
              Year |
             2006  |  -.0931186   .0976341    -0.95   0.341     -.285127    .0988897
             2007  |  -.1506941   .1196353    -1.26   0.209    -.3859703    .0845821
             2008  |  -.3771988    .160835    -2.35   0.020    -.6934989   -.0608987
             2009  |  -.4125662   .1663068    -2.48   0.014    -.7396271   -.0855052
             2010  |  -.6597916   .2470061    -2.67   0.008    -1.145557   -.1740264
             2011  |  -.5763591   .1997539    -2.89   0.004    -.9691976   -.1835206
             2012  |  -.6730537   .2154415    -3.12   0.002    -1.096744   -.2493638
             2013  |  -.7146015   .2247925    -3.18   0.002    -1.156681   -.2725218
             2014  |  -.5656688    .196636    -2.88   0.004    -.9523757    -.178962
             2015  |  -.5367778   .1768668    -3.03   0.003    -.8846063   -.1889493
             2016  |  -.4683608   .1762386    -2.66   0.008    -.8149539   -.1217677
             2017  |  -.3803226   .1953754    -1.95   0.052    -.7645503     .003905
             2018  |   .4640251   .1954189     2.37   0.018     .0797118    .8483384
             2019  |          0  (omitted)
                   |
             _cons |   8.616593   2.136981     4.03   0.000     4.413979    12.81921
------------------------------------------------------------------------------------

similarly
tabulate mpr,gen(dMPR)

monetary |
poicy |
regime | Freq. Percent Cum.
------------+-----------------------------------
EMP | 206 53.79 53.79
TMP | 177 46.21 100.00
------------+-----------------------------------
Total | 383 100.00

. reg zscore NIM lasset CapitalRatio dMPR1#c.C5_Aggregate i.Year,robust
note: 1.dMPR1#c.C5_Aggregate omitted because of collinearity

Linear regression Number of obs = 378
F(18, 359) = 3.40
Prob > F = 0.0000
R-squared = 0.1523
Root MSE = .4636

--------------------------------------------------------------------------------------
| Robust
zscore | Coef. Std. Err. t P>|t| [95% Conf. Interval]
---------------------+----------------------------------------------------------------
NIM | .1617704 .0483224 3.35 0.001 .0667399 .2568008
lasset | .030806 .027096 1.14 0.256 -.0224808 .0840928
CapitalRatio | -.8086158 .4762358 -1.70 0.090 -1.745178 .1279466
|
dMPR1#c.C5_Aggregate |
0 | .0775196 .1512528 0.51 0.609 -.2199332 .3749723
1 | 0 (omitted)
|
Year |
2006 | -.0665558 .0967374 -0.69 0.492 -.256799 .1236873
2007 | .0244397 .1052275 0.23 0.816 -.1825 .2313795
2008 | .0312063 .1090563 0.29 0.775 -.1832633 .2456758
2009 | -.0203048 .1251241 -0.16 0.871 -.2663732 .2257636
2010 | -.0975821 .1778664 -0.55 0.584 -.447373 .2522089
2011 | .0764374 .081305 0.94 0.348 -.0834565 .2363313
2012 | .0167227 .0846394 0.20 0.843 -.1497287 .183174
2013 | .0303189 .0867024 0.35 0.727 -.1401896 .2008273
2014 | -.0938555 .1097305 -0.86 0.393 -.3096509 .1219399
2015 | -.0585723 .1009598 -0.58 0.562 -.2571193 .1399746
2016 | -.2019525 .1374338 -1.47 0.143 -.472229 .068324
2017 | -.332495 .1920758 -1.73 0.084 -.7102302 .0452401
2018 | -.2988177 .1752532 -1.71 0.089 -.6434695 .0458341
2019 | -.6637235 .1854361 -3.58 0.000 -1.028401 -.299046
|
_cons | -.354233 .3581022 -0.99 0.323 -1.058475 .3500087
--------------------------------------------------------------------------------------

Code:

 





(1)




zscore








NIM
0.167**




(3.29)








lasset
0.0371




(1.37)








CapitalRatio
-0.605




(-1.09)








0.dMPR2
0




(.)








1.dMPR2
-2.849*




(-2.49)








C5_Aggregate
-22.22***



(-3.99)








0.dMPR2#c.C5_Aggregate
0




(.)








1.dMPR2#c.C5_Aggregate
7.187*




(2.47)








2005.Year
0




(.)








2006.Year
-0.0931




(-0.95)








2007.Year
-0.151




(-1.26)








2008.Year
-0.377*




(-2.35)








2009.Year
-0.413*




(-2.48)








2010.Year
-0.660**




(-2.67)








2011.Year
-0.576**




(-2.89)








2012.Year
-0.673**




(-3.12)








2013.Year
-0.715**




(-3.18)








2014.Year
-0.566**




(-2.88)








2015.Year
-0.537**




(-3.03)








2016.Year
-0.468**




(-2.66)








2017.Year
-0.380




(-1.95)








2018.Year
0.464*




(2.37)








2019.Year
0




(.)








_cons
8.617***




(4.03)








N
378








t statistics in parentheses


="* p<0.05
 ** p<0.01
 *** p<0.001"

can any one help me on this.Your commnets are highly appreciated

Tags: None

Carlo Lazzaro

Join Date: Apr 2014

Posts: 17609
#2

28 Dec 2021, 02:40

Fadi:
as far as I can get it, you have an interaction only.
That said, to a have clearer idea of what's going on with your coefficients, try:

Code:

reg zscore NIM lasset CapitalRatio i.mpr##c.C5_Aggregate i.Year,robust allbaselevels

Kind regards,
Carlo
(StataNow 18.5)
Comment

Fadi Ansar

Join Date: May 2021
Posts: 120

28 Dec 2021, 03:31

Thankyou so much for your help.This is what I was getting after doing the commands .is there any way I can get the results like EMP*C5_aggregate and TMP*C5_aggregate as MPR inclusive of both TMP and EMP

Code:

eststo:reg zscore  NIM  lasset CapitalRatio  i.mpr##c.C5_Aggregate i.Year,robust allbaselevels
note: 2019.Year omitted because of collinearity

Linear regression                               Number of obs     =        378
                                                F(19, 358)        =       3.48
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1665
                                                Root MSE          =     .46035

------------------------------------------------------------------------------------
                   |               Robust
            zscore |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------------+----------------------------------------------------------------
               NIM |    .166563   .0505932     3.29   0.001     .0670658    .2660601
            lasset |   .0370991    .027149     1.37   0.173    -.0162923    .0904906
      CapitalRatio |  -.6045754   .5546065    -1.09   0.276    -1.695272    .4861208
                   |
               mpr |
              EMP  |          0  (base)
              TMP  |  -2.849336   1.144863    -2.49   0.013    -5.100838   -.5978336
                   |
      C5_Aggregate |   -22.2205   5.575988    -3.99   0.000    -33.18631    -11.2547
                   |
mpr#c.C5_Aggregate |
              EMP  |          0  (base)
              TMP  |    7.18703   2.908346     2.47   0.014     1.467441    12.90662
                   |
              Year |
             2005  |          0  (base)
             2006  |  -.0931186   .0976341    -0.95   0.341     -.285127    .0988897
             2007  |  -.1506941   .1196353    -1.26   0.209    -.3859703    .0845821
             2008  |  -.3771988    .160835    -2.35   0.020    -.6934989   -.0608987
             2009  |  -.4125662   .1663068    -2.48   0.014    -.7396271   -.0855052
             2010  |  -.6597916   .2470061    -2.67   0.008    -1.145557   -.1740264
             2011  |  -.5763591   .1997539    -2.89   0.004    -.9691976   -.1835206
             2012  |  -.6730537   .2154415    -3.12   0.002    -1.096744   -.2493638
             2013  |  -.7146015   .2247925    -3.18   0.002    -1.156681   -.2725218
             2014  |  -.5656688    .196636    -2.88   0.004    -.9523757    -.178962
             2015  |  -.5367778   .1768668    -3.03   0.003    -.8846063   -.1889493
             2016  |  -.4683608   .1762386    -2.66   0.008    -.8149539   -.1217677
             2017  |  -.3803226   .1953754    -1.95   0.052    -.7645503     .003905
             2018  |   .4640251   .1954189     2.37   0.018     .0797118    .8483384
             2019  |          0  (omitted)
                   |
             _cons |   8.616593   2.136981     4.03   0.000     4.413979    12.81921

Thankyou so much.

Comment

Carlo Lazzaro

Join Date: Apr 2014

Posts: 17609
#4

28 Dec 2021, 03:40

Fadi:
no, you can's, as -EMP- is your reference category.

Kind regards,
Carlo
(StataNow 18.5)
Comment
Fadi Ansar

Join Date: May 2021

Posts: 120
#5

28 Dec 2021, 04:58

Thanks so much @Carlo .In one of the research paper they have done something similar to this.Can I get something like this if I can run regression seperately for both EMP and TMP.
Attached Files
Comment

Carlo Lazzaro

Join Date: Apr 2014
Posts: 17609

28 Dec 2021, 05:57

Fadi:
you have a two-level vategorical variable; the reference category does not enter the interaction (and there's nothing you can do about that, because multiplying a given number by zero gives back zero).
If you change your reference category, you simply reverse the isssue (that, as per what above, is here to stay).
IMHO, one of the most didactic exercise to carry out during OLS postestimate tests, is to ask Stata for calculating the fitted values, calulate them by hand based on code and coefficients and then compare the results obtained with the two approaches:

Code:

use "C:\Program Files\Stata17\ado\base\a\auto.dta"
. regress price i.foreign##c.trunk if trunk<=11, allbase

      Source |       SS           df       MS      Number of obs   =        27
-------------+----------------------------------   F(3, 23)        =      1.36
       Model |  4645330.66         3  1548443.55   Prob > F        =    0.2806
    Residual |  26238937.9        23  1140823.39   R-squared       =    0.1504
-------------+----------------------------------   Adj R-squared   =    0.0396
       Total |  30884268.5        26  1187856.48   Root MSE        =    1068.1

---------------------------------------------------------------------------------
          price | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
        foreign |
      Domestic  |          0  (base)
       Foreign  |   1666.006   2271.605     0.73   0.471    -3033.168    6365.179
                |
          trunk |   -13.7037   187.6449    -0.07   0.942    -401.8767    374.4693
                |
foreign#c.trunk |
      Domestic  |          0  (base)
       Foreign  |  -98.73108   244.9716    -0.40   0.691    -605.4935    408.0314
                |
          _cons |   4493.741   1748.222     2.57   0.017     877.2681    8110.213
---------------------------------------------------------------------------------

predict fitted, xb

*Let's calculate fitted values by hand for a domestic car with trunk=10*

. di 4493.741 + (-13.7037*10)
4356.704

. list fitted if foreign==0 & trunk==10

     +----------+
     |   fitted |
     |----------|
  7. | 4356.704 |
 25. | 4356.704 |
 
 40. | 4356.704 |
     +----------+
*Let's calculate fitted values  by hand for a foreign car with trunk=10 (decimal digits rounding)*
. di 4493.741 + 1666.006 + (  -13.7037 -98.73108)*10
5035.3992

. list fitted if foreign==1 & trunk==10

     +----------+
     |   fitted |
     |----------|
 61. | 5035.398 |
 65. | 5035.398 |
     +----------+


.

Kind regards,
Carlo
(StataNow 18.5)

Announcement

Dummy varaible in regression -problem with results ouput

Comment

Comment

Comment

Comment

Comment