Do compensation committees do what they say they do? Evidence from narrative disclosure in proxy statements

By prof. Sewon Kwon
Division of Business Administration
PURE Research Profile
k4js1@ewha.ac.kr

Prof. Sewon Kwon’s recent research published in Finance Research Letters, entitled “Do compensation committees do what they say they do? Evidence from narrative disclosure in proxy statements,” is about whether Compensation Committees’ stated “Compensation Philosophy” in the CD&A section of proxy statements-especially (i) pay-for-performance and (ii) attraction & retention of talent - is reflected in actual CEO pay outcomes.

The authors conduct a textual analysis of CD&A disclosures for ExecuComp firms (S&P 1500) from 2007–2016 (7,781 firm-year observations). They use Python to count the frequency of pay-for-performance and attraction/retention phrases throughout the CD&A and link these measures to excess CEO compensation (EXCESSCOMP) - the residual component of pay beyond economically predicted compensation - estimated following Core et al. (2008).

Conceptually, the paper contrasts three motives for repeating compensation-philosophy language in disclosures: salience (repetition reflects real pay features), obfuscation (repetition masks pay that deviates from the stated philosophy), and symbolism (repetition is a legitimacy signal without substantive implementation). The central hypothesis is that CEO compensation reflects firms’ emphasis on these philosophy themes.

Empirically, the results show that greater emphasis on pay-for-performance wording is positively associated with excess CEO compensation (e.g., PFP_WORDS coefficient ≈ 0.360), while attraction & retention wording is not significantly related to excess pay. The effect is especially pronounced in firms without a general counsel among top officers and in firms not hiring compensation consultants, consistent with an obfuscation interpretation. (Descriptively, sample firms mention pay-for-performance phrases about 5.7 times and attraction/retention phrases about 5.1 times on average; CD&A length is about 5,954 words.)

In terms of implications, the paper suggests that investors should be cautious when interpreting narrative statements about pay-for-performance, and it notes relevance to regulatory discussion (e.g., Dodd-Frank pay-versus-performance disclosures), because stronger pay-for-performance rhetoric can coincide with higher excess pay.

* Related Article
Natalie Kyung Won Kim, Sewon Kwon, Jae Yong Shin, Do compensation committees do what they say they do? Evidence from narrative disclosure in proxy statements, Finance Research Letters, 73, 2025

Core, J., Guay, W., Lacker, D.F., The power of the pen and executive compensation, Journal of Financial Economics, 88(1), 1–25, 2008