Compustat execucomp data processing

[wendy chan]

Hi everyone,

I am currently using Stata to process Compustat ExecuComp data to work on my thesis. I have encounter a problem with number of observations (number of CEO-year) and number of firms. After processing, I have found about 28,138 observations but I would like to know how many firms are there is the 28,138 observations since there might be one CEO for a firm for different years. How should I actually code to know how many firms are there?

The data downloaded from WRDS had different ID number for each executive/company combination, but how do I code to categorise them to number of firms?

Thanks a lot.

[Wouter Wakker]

Hi Wendy,

you can count the number of firms with the distinct command:

distinct(coname)

However, this will ignore missing values, and I see your first 5 observations have missing data for the company name. You might want to look first why that data is missing.

You can also create a firm id with encode:

encode coname, gen(firm_id)

This will give the firms a number from 1 to the total amount of firms, while retaining the company name as a value label.