Almost Any Way You Figure It, Executive Pay Remains Irrational
Whenever I analyze chief executive pay in a single year, I get a lot of screaming from certain companies who claim that my estimates of their pay for the particular year are artificially high.
For example, companies explain that they make stock option grants only every three years. With some justice, they argue that in the years when they makes option grants, the media jumps all over them because the CEO’s pay is so high. But in the odd years when they make no option grants, the reporters are nowhere to be seen.
OK, so let’s do some rearranging.
I started with the 900 companies comprising the S&P; 500 and S&P; Mid-Cap 400 index groups. From these companies, I selected 618 firms where the CEO had served in his position for at least the three years--from 1992 through 1994. Using proxy reports, I then averaged each CEO’s pay over the three-year period, taking into account his base salary, annual bonus, restricted stock grants, estimated present value of stock option grants, payouts under other types of long-term incentive plans and other miscellaneous compensation. (I use the phrase “his” pay, because among the 618 CEOs, there was only one women--Sally Kasaks of Ann Taylor Stores.)
Three-year pay for the group ranged from an almost-unnoticeable $52,000 per year for Stuart R. Levine at Cabletron Systems in Rochester, N.H., to the eye-popping $35.9 million per year for Sanford I. Weill at Travelers Inc. in New York. The average pay for the entire group was $2.3 million per year.
What might account for why one CEO among the 618 earns $52,000 per year, while another earns almost $36 million per year?
Well, one factor is company size. Thirty-one percent of the variation in pay levels can be accounted for differences in company size. In measuring company size, I gave equal weight to annual revenue, or an income statement measure of size; to invested capital, a balance sheet measure of size, and to the number of employees, a human capital measure of size.
*
Another factor is company performance. I measured company performance by calculating total shareholder returns (stock price appreciation plus reinvested dividends) in time windows of one, two, three, four and five years, all ending at the end of 1994. Then I averaged the results from all five windows, and what I found was that company performance counts for very little compared to size. Adding performance improves the explanation to pay variation only from 31% to 35%.
Granted there is some rationality in the pay levels of these 618 CEOs, in the sense that bigger companies pay more to their CEOs, and so do better-performing companies. Still, some 65% of pay variation cannot be explained by either size or performance, and it cannot be explained by CEO tenure either.
Here are a few other findings from the study:
* For the same amount of company size and company performance, companies located in the New York City area offer 21% more salary, 30% more salary and bonus and 48% more total pay. Even though the cost of living in Los Angeles is arguably as high as it is in New York, CEOs in Southern California do not receive any pay premium after correcting for size and performance.
* Of 19 industries studied within the 618-company group, only three stand out as paying differently compared to other industries. Interestingly, two of them are health-care related. In the pharmaceutical industry and for the same amount of size and performance, CEOs receive 21% more in salary, 28% more in salary and bonus and 45% more in total pay. And in the insurance industry (which includes HMOs and other health-care insurers), the premiums are 26% for salary, 32% for salary and bonus and 49% for total pay. It seems ironic that at the very time when we are experiencing a national crisis in health care costs, health care-related companies have chosen to offer their CEOs a huge pay premium.
* Only one industry systematically pays less, and that is the power utility industry. There, after taking account of size and performance, your typical CEO earns 16% less in salary, 31% less in salary and bonus and 47% less in total pay.
I also compared the compensation of the 30 CEOs in Southern California to the other 588 chief executives in the survey. I looked at the actual three-year average annual compensation and calculated a competitive level of compensation taking into account both size and performance and the percent by which actual pay exceeds (or is under) competitive pay.
As a group, the 30 CEOs in the survey from Southern California were unremarkable. After correcting for differences in company size and performance, the average CEO in Southern California earned 8.4% more than the competitive rate for base salary, 18.2% more than the competitive rate of base salary and bonus, and only 2.5% more than the competitive rate of total pay. And none of these differences were statistically meaningful.
However, upon closer perusal the data show two terrific buys and one who’s not so great.
The terrific buys are Computer Science’s William R. Hoover and Northrop Grumman’s Kent Kresa. Hoover’s total shareholder returns in the five time windows of performance ranked him at the 83rd percentile of the 618-company group. In other words, only 17% of the companies in the group outperformed Computer Sciences.
As for Kresa, his performance ranked him at the 81st percentile of the group. Yet both CEOs earned far less than they should have earned, given their larger-than-average size and excellent performance--49% less for Hoover and 54% less for Kresa.
On the other hand, we have Kaufman & Broad Home’s Bruce Karatz. He runs a relatively small company (21st percentile of size) that is also underperforming. His shareholder returns for all time windows five through one positioned him, respectively, at the 21st, 58th, 29th, 21st and fourth percentiles of the 618-company distribution.
(Karatz was traveling in Europe and could not be reached for comment. But a spokesman for the firm said that although he had not seen the current study, the company has repeatedly disagreed with Crystal’s methodology in the past. “The bottom line is,” he said, “that the company firmly believes that Mr. Karatz’s compensation is fully commensurate with performance, which has been stellar under his 15 years of leadership.”)
As it now stands, company size is 2.7 times more important in driving executive pay than is shareholder return performance. Most shareholders would argue that it is the latter that should be 2.7 times more important than the former.
Moreover, there is far too much noise in CEO pay packages. In an ideal world, at least 60% to 70% of the variation in CEO pay, not merely 35%, ought to be explainable by such rational factors as company size and company performance.
A lot of work remains to be done if CEOs across America are to be paid in a rational fashion.
