- ISS did not provide transparency into its backup information and data.
- ISS seemingly overlooked the reality that a company's performance has a more statistically significant impact on CEO pay than board leadership.
- ISS was quick to connect some dots but failed to connect them with the kind of precision that is expected of corporate America.
Institutional Shareholder Services ("ISS") published a report in March 2016 arguing that CEO compensation is impacted by companies' board structures. More specifically, ISS argues that CEOs of companies with boards chaired by an "insider" have higher compensation than CEOs of companies with boards chaired by an "outsider."
Given ISS's orientation toward one-size-fits-all corporate governance scorecards—and the fine print found on the last page of the report (warning readers that the information may not be accurate and that anyone who relies on the information does so at his or her own risk), we decided to dig into the report's findings before accepting its conclusion. We think that our caution was warranted.
Generally, the report does not provide much transparency into its underlying data, which makes it difficult to analyze. Our requests for the backup information received no response. Nonetheless, we were able to use the data included in the report to rerun some of the calculations.
The report separates S&P 500 companies into four board categories—those chaired by: (i) an individual deemed by ISS to be an "insider," (ii) the CEO of the company, (iii) an individual deemed by ISS to be an "affiliated outsider," and (iv) an individual deemed by ISS to be an "independent outsider."
The report then concludes that:
- The average total compensation of CEOs in the insider and combined role categories is higher than the average total compensation of CEOs in the affiliated outsider and independent outsider categories, and
- Therefore, board structure has a significant impact on CEO pay.
Although the ISS report does not identify the companies it assigned to each of these categories, ISS does provide the identity of four companies in the "insider" category, and describes the compensation of these CEOs as being high outliers.
Removing these so-called outliers from the "insider" category reduces the average total compensation for the "insider" category from $15.6 million to $11.8 million. In short, when the outliers are removed from the "insider" category, the resulting "insider" category's average annual compensation is reduced by almost 25 percent and is almost equal to the average annual compensation of the "affiliated outsider" category.
We reviewed the proxies of the so-called "outliers" and discovered a few facts relevant to the conclusions drawn that were not mentioned by ISS:
- One outlier's stock price increased 1600 percent over the five-year period ended December 31, 2014.
- Another outlier's revenues almost doubled since its CEO took leadership and 98 percent of the CEO's 2014 compensation was put at risk with stringent performance based measures.
- A third outlier decreased its then-CEO's compensation three years in a row, and ultimately split the CEO position into two roles in response to shareholder feedback obtained in outreach efforts.
We believe that our findings demonstrate the danger in trying to adopt a one-size-fits-all approach to corporate governance and compensation practices. Public company boards and compensation committees use many types of information when making compensation decisions (feedback received through direct shareholder outreach efforts, performance measures, long-term strategies, etc.). Any report that attempts to identify a single reason for compensation variances among public company CEOs will, in our view, inevitably fail to account for valid variances.
The first two of the so-called outliers described above, for example, adhered to ISS's top corporate governance compensation policy—creating a meaningful link between pay and performance—by providing a large portion of their executives' compensation packages in the form of performance-based/at-risk awards. The compensation packages of those CEOs have proven to create an effective incentive to maximize the value of their respective companies, as evidenced by the companies' strong performances. Yet the strong performances, which resulted in high compensation, are not taken into account in ISS's analysis.
The report also fails to acknowledge that a company's revenue has a more statistically significant impact on CEO pay than does the category applicable to the company's chair. In addition to testing the impact of the identity of the board chair on CEO pay, the report also analyzed the impact of:
- Three-year "indexed" total shareholder return of the company versus the S&P 500,
- Company revenues (averaged over the three-year period),
- The CEO's tenure, and
- Whether there was a change in CEO during the course of the three years.
According to the report, in order to be statistically significant, the t-statistic of a variable must be greater than 2 or less than -2. Of the five variables tested, only two of the variables had t-statistics that were statistically significant: the chair code at -2.53 and the company revenue code at 5.36—more than two-and-a-half times greater. Any student of statistics would not "bury the lede" by focusing on a variable of lesser significance.
The authors of the ISS report nonetheless jump to the conclusion that "insiders" are not the best monitors of shareholder interests in the boardroom. This conclusion is unsupported as well as unfair, particularly given that, under stock exchange rules, CEO pay is set solely by independent directors and may not be set by "insiders." In fact, CEO chairs cannot even serve on compensation committees. In our experience, compensation committees work hard to ensure an appropriate level of pay and linkage to performance. To blindly follow ISS's conclusion would require shareholders to oust from the board the very leaders who have made many of these companies attractive investments in the first place and who have carefully applied the very principles that ISS espouses.