Generate dummy variable with inlist

[Rochelle Zhang]

Hi all,

I have a string variable splti in data set that denotes the credit rating of a firm's debt. This rating variable can take values such as "AAA" "A" etc, see below.

I want to have a dummy variable LowRate set equal to 1 if the debt rating is below investment grade , which includes ratings such as "BBB-", "CC","CCC","CCC+", "CCC-","D", "SD".

I used the script below, but it is incomplete, I do not have the value of 0, only value of 1

gen LowRate=1 if inlist(splti,"BBB-", "CC","CCC","CCC+", "CCC-","D", "SD")

replace LowRate=. if splti==""  // for firms having missing value for splti

my sample has 50000 observations, but only a fraction has rating data as shown below

/*
          A |        296        6.80        6.80
         A+ |        131        3.01        9.82
         A- |        386        8.87       18.69
         AA |         29        0.67       19.36
        AA+ |          9        0.21       19.56
        AA- |         87        2.00       21.56
        AAA |          9        0.21       21.77
          B |        266        6.11       27.89
         B+ |        291        6.69       34.57
         B- |        139        3.20       37.77
         BB |        297        6.83       44.60
        BB+ |        288        6.62       51.22
        BB- |        334        7.68       58.90
        BBB |        560       12.87       71.77
       BBB+ |        540       12.41       84.18
       BBB- |        570       13.10       97.29
         CC |          6        0.14       97.43
        CCC |         18        0.41       97.84
       CCC+ |         69        1.59       99.43
       CCC- |          8        0.18       99.61
          D |         10        0.23       99.84
         SD |          7        0.16      100.00
*/

could you help with the correct coding of LowRate?

Thank you,

Rochelle

[Nick Cox]

 
 gen LowRate=1 if inlist(splti,"BBB-", "CC","CCC","CCC+", "CCC-","D", "SD")

generates 1 if the string variable splti takes on any of those values and missing otherwise.

 
 replace LowRate=. if splti==""  // for firms having missing value for splti

does no harm, but it changes nothing as if the variable is missing then the indicator is already missing.

If you want a (0, 1) indicator that is simply produced by

 
 gen LowRate= inlist(splti,"BBB-", "CC","CCC","CCC+", "CCC-","D", "SD")

If you want (0, 1, missing) indicator that is missing if the string variable then that is

 
 gen LowRate= inlist(splti,"BBB-", "CC","CCC","CCC+", "CCC-","D", "SD") if !missing(splti)

The basic ideas were all covered in https://www.stata.com/support/faqs/d...rue-and-false/

A longer discussion by Clyde Schechter and friend was published in https://journals.sagepub.com/doi/abs...36867X19830921

A way to think about this is: Where does your code instruct Stata ever to return a value of 0?

[Rochelle Zhang]

Thank you Nick. This is very helpful !!!