The intense spotlight on executive pay centers on the relationship between pay and performance for CEOs. Legacy pay-setting practices and proxy advisory firm models have largely relied upon comparisons of pay derived from disclosure in the Summary Compensation Table to total shareholder return (“TSR”). Critics of CEO pay often focus exclusively on the absolute amount of annual “pay” disclosed in the Summary Compensation Table without considering firm performance or size, or distinguishing between pay opportunity (subject to future performance) and actual pay realized.
In prior Pay Governance research, we have proposed that Realizable Pay1 is the most appropriate definition of compensation in the pay for performance assessment. We have demonstrated that high performing companies have higher levels of Realizable Pay than low performing companies2. Recently, the Conference Board Working Group on Supplemental Pay Disclosure concluded that Realizable Pay is appropriate for demonstrating the alignment of pay with performance and in comparison to peers3.
Pay Governance researched CEO pay opportunities from 2003-2012 for a constant sample of long-service chief executive officers (CEOs) at 45 Fortune 500 companies. This long-service, constant CEO sample was selected to prevent bias due to changes in incumbents, company mergers or significant acquisitions. We assessed pay and performance for the 10-year performance period ending in the first quarter of 2013.
This Viewpoint focuses on the relationships of CEO pay with performance for this study group. A future Viewpoint will evaluate the pay practices and trends in this long-tenure and high performing group of companies.
Realizable Pay has a much stronger relationship to TSR performance than compensation reflected in the Summary Compensation Table (“SCT-based Pay”). Realizable Pay is composed of cash compensation paid, cash long-term incentive payouts and the value of equity awards using the stock price at the end of the assessment period. Realized Pay is also different than SCT-based Pay, which for equity includes only equity vested or stock options exercised. Both absolute Realizable Pay and Realized Pay have much higher correlations with TSR performance than pay disclosed in proxies (SCT-based Pay includes actual cash salary and cash incentive payouts, but replaces current or realized value of equity with accounting fair value of awards and excludes pension accruals and other compensation). In our analysis, Realized Pay, which is only presented on a 5-year basis due to the limitations of pre-2007 proxy disclosures, had a comparable correlation with Realizable Pay.
We also found similar results for the comparison of these pay definitions to growth in market capitalization, which includes both TSR and any change in shares outstanding. In addition, pay disclosed in proxies has very weak (and in some cases negative) correlations to firm growth (revenues and assets) and profitability (cash flow, EPS, return on equity and return on assets), while Realizable Pay and Realized Pay have much stronger correlations with these measures. This is because committees set SCT-based stock grants mostly based on survey/peer data and not on recent company performance.
Companies with high performance have meaningfully higher realizable pay than lower performers. To test our conclusions on Realizable Pay and alignment with TSR performance, we segmented our study group into high, medium and low performers based upon 5- and 10-year TSR. The results illustrate the linkage between Realizable Pay and TSR performance. High TSR performing companies have much higher levels of Realizable Pay than companies that have weaker TSR performance.
The strongest relationship between pay and performance is demonstrated by the ratio of Realizable Pay and SCT-based Pay. We used a new definition of pay ratio—total Realizable Pay divided by SCT-based Pay—to determine how much incremental value from long-term incentives was earned or “forfeited” (this is a comparison of long-term incentive values, since salary and earned cash annual incentives are the same value in both pay definitions). In SCT-based pay, we excluded pension and other compensation. If this ratio equals 100%, then the realizable value of all long-term incentives equals the grant date fair value and all of the intended target pay opportunity was realizable.
Our analysis found that this ratio has the strongest positive correlation with TSR performance of any of the pay definitions studied.
This ratio also varies by TSR performance — high performers have a higher ratio while low performers have a lower ratio.
It is important for compensation committees to evaluate pay from multiple perspectives—using proxy-disclosed pay, survey data or custom peer group data alone is insufficient. Compensation committees should also consider the use of backward and forward looking Realizable Pay and the ratio of Realizable Pay/SCT-based Pay to evaluate the pay for performance relationship for their CEOs in comparison to peers as these measures have the strongest correlation to TSR performance. We will explore these issues further in a subsequent Viewpoint.