Calculate delta (pay-performance sensitivity), vega (risktaking incentives), and firm-specific wealth (inside equity) for executives on Execucomp

Several papers use measures of delta (pay-performance sensitivity), vega (risktaking incentives), and firm-specific wealth (inside equity) for executives on Execucomp. For example,

1.  Core, J., Guay, W., 2002. Estimate the value of employee stock option portfolios and their sensitivities to price and volatility. Journal of Accounting Research 40, 613-630.

  • develops a method to calculate delta and vega using information provided by Execucomp in the pre-2006 period

2.  Coles, J., Daniel, N., Naveen, L., 2006. Managerial incentives and risk-taking. Journal of Financial Economics 79, 431-468.

  • estimates delta and vega

3.  Daniel, N., Li, Y., and Naveen, L. 2013. No asymmetry in pay for luck. Working Paper.

  • calculates a measure of firm-specific wealth using executives’ stock and option portfolios

Naveen makes publicly available her SAS program used to calculate delta, vega, and firm-specific wealth. See her homepage here. However,

  • Her program is not self-executable because she uses three external datasets for which she does not provide SAS codes used to create them.
  • Her program is to be executed locally so not portable (does not support PC SAS or SSH connection).
  • Her program calculates these measures only up to fiscal year 2010.

I improve Naveen’s program to make it self-contained and executable on its own. Specifically, I recreate the three datasets within the new program and update dataset references to point to the sever end. Now you can run the program via PC SAS or SSH connection, and specify the start year and end year of the period of interest. So you can easily update the data up to the most recent date.

I write a little more details in the overview section in the new program. As evidenced in the overview, I believe that I successfully replicate Naveen’s data using the new program. However, if you decide to use the new program, the accuracy of the generated data is your own responsibility.

Lastly, please cite Naveen’s work if you use the new program. I would be appreciated if you are generous enough to acknowledge my work.

This entry was posted in SAS. Bookmark the permalink.

12 Responses to Calculate delta (pay-performance sensitivity), vega (risktaking incentives), and firm-specific wealth (inside equity) for executives on Execucomp

  1. Thomas says:

    Hey Kai,

    First of all, thanks for sharing your SAS-script, it’s very useful!

    As I aim to gather the Delta’s and Vega’s of the CEOs of companies in the S&P 1500 during 2007 – 2014, I hope you can help me solving a question that is related to the output of the script (I want to match the variables Delta and Vega to my dataset of which the GVKEY is the primary identifier);

    – The output shows multiple rows of data per company per year. I assume this is because it also includes data on other executives, besides the CEO. I wish to collect only the data of CEOs. Do you think this is possible?

    Best regards,

    • Kai Chen says:

      You are right – the output includes other executives than CEO. You can modify the program to only include CEO. Another way (may be easier) is to SQL query if CO_PER_ROL is linked to a CEO flag in Execucomp.

  2. Bo says:

    Hi, Kai,

    In the following part:

    if fyr=6 then assumed_grantyear=year-1;
    else assumed_grantyear=year;

    should it be “fyr<=6" instead of "fyr=6", since options granted in the first 5 months should also be assigned the previous year as the grant year?

    • Kai Chen says:

      Bo has informed me that Naveen’s program is correct—Naveen uses the code because of the way Compustat defines data year. Naveen shows the example in details in the program. Thank you Bo for letting me know this.

  3. Paul says:

    Hi Kai,

    Excellent code. If you can I think it might be valuable to add in ‘moneyness’ calculations too:

    Campbell, T.C., Gallmeyer, M., Johnson, S.A., Rutherford, J., Stanley, B.W., 2011. CEO
    optimism and forced turnover. J. Financ. Econ. 101, 695-712.

  4. Socrates says:

    Hi Kai,
    Thank you very much for this excellent code.
    I am not able to find these two variables: fybegdt fyenddt. I am using Stata so I need to download them.
    Are they supposed to be available on CRSP/Compustat Merged Database – Security Monthly?
    Many thanks for your help.

    • Kai Chen says:

      The first several lines in the program generates fybegdt and fyenddt. They are not something that can be directly downloaded from WRDS. You need SAS to run the program. Stata will not do the work.

  5. Li says:

    Hi Kai:
    Thank you very much for the codes. However, the coperol from the download data, I assume, is modified and ranked by the order of firms in the sample, instead of the true co_per_rol? How could I solve it? I checked the codes but couldn’t figure out where the definition of co_per_rol has been changed.

    • Kai Chen says:

      I don’t quite understand your question. If I remember correctly, co_per_rol is the unique id of an executive, no matter which company he/she works for. The code won’t change co_per_rol throughout.

  6. Li says:

    Nevermind. I figured it out. It’s due to the sorting in the SAS data.

  7. Emily says:

    Hi Kai,
    Thanks for sharing the codes! I’m a bit confused about the following codes at the end of the program.

    data deltavega;
    set deltavega;
    if optiondelta=0 then optiondelta=.;
    if delta=0 then delta=.;

    You corrected Naveen’s codes by using sum function. Sum function itself returns missing value if all variables summed up are missing. If we set zero optiondelta and delta to missing, aren’t we dropping some observations with legitimate value zero?

    • Kai Chen says:

      Hi Emily, thanks for letting me know. I think you’re correct. I cannot recall exactly why I added these codes. Probably because I misunderstood the sum function. I commented them out. Glad I had that disclaimer from day 1 🙂

Leave a Reply

Your email address will not be published. Required fields are marked *