Skip to content


I cannot guarantee the data are error free, although I have spent a ton of time cleaning the data and checking for errors.  
If you do spot any errors, I would appreciate you letting me know.

The data span 1992-2014.

If you use this data, please cite our paper on symmetry in pay for luck:

Daniel, N.D., Li, Y., and Naveen, L., 2020. Symmetry in Pay for Luck. Review of Financial Studies 33, 3174–3204.

(i) Data Description: These are the main luck and skill measures used in our paper (described in the legend to Table 2). Please refer to Daniel, Li, and Naveen (2020) for details on construction.

(ii) Luck and Skill Data: Unique identifier is Execucomp variable CO_PER_ROL. The columns correspond to CO_PER_ROL, YEAR, LUCK, SKILL.

The data span 1992-2014.

If you use this data, please cite our incentives paper:

Coles, J.L., Daniel, N.D., and Naveen, L., 2006. Managerial Incentives and Risk-Taking.  Journal of Financial Economics 79, 431-468

Also please cite Core and Guay (2002) as our calculations are based on their paper’s methodology:

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

(i)  Note

(ii) Compensation Data (gvkey, year, co_per_rol, delta, vega, firm-related wealth)

(iii) SAS Program (saved as text file since SAS program file not permitted to be posted)

The data have been updated and span 1996-2022.

If you use any of this data, please cite our Co-opted Boards paper:

Coles, J.L., Daniel, N.D., and Naveen, L., 2014.  Co-opted Boards.  Review of Financial Studies 27, 1751-1796.

(i) Data description:  See Appendix of our paper above for  working with Riskmetrics data on boards. The appendix also provides a more detailed explanation of these variables.


(FRACDIRAFTER, and FRACDIRAFTERINDEP represent co-opted directors and co-opted independent directors as a fraction of the total board; TWFRACDIRAFTER, and TWFRACDIRAFTERINDEP represent the  corresponding tenure-weighted measures as defined in our paper)

(iii) Link Data: for merging the Riskmetrics data with Compustat.

  1. MTGDATE: the date of the annual meeting, in SAS date format (as mentioned in our 2014 paper, we assign directors on the slate at the annual meeting in a given fiscal year as the directors for that year).
  2. PERMNO: CRSP identifier
  3. GVKEY: Compustat identifier
  4. YEAR: fiscal year of data
  5. CUSIP6: six-digit Cusip identifier from ISS/Riskmetrics in legacy file or first six-digits from ISS/Riskmetrics directors file
  6. CUSIP8: first eight digits of 9-digit Cusip identifier from ISS/Riskmetrics
  7. DIDOWN: the director ID created by us as in Coles, Daniel, and Naveen.  This should be unique for a given director in the dataset.
  8. LEGACY_DIRECTOR_ID: director ID from ISS/Riskmetrics directors legacy file
  9. DIRECTOR_DETAIL_ID: director ID from ISS/Riskmetrics directors file

To merge the data with ISS Directors data, match by MEETINGDATE, CUSIP8, and DIRECTOR_DETAIL_ID.

To merge with ISS Directors Legacy data,  match by MEETINGDATE, CUSIP6,  and LEGACY_DIRECTOR_ID.

There are some filters applied to the data. For example, we start with the CRSP-Compustat merged database from WRDS, so we only have firms that have a match on both CRSP and Compustat.  We keep only share codes 10 and 11.  There are some other steps during the program where observations get dropped because of additional filters (see appendix of our RFS paper for more details).