CEO Pay for Performance: Reality or Illusion
Fifty years ago CEO pay was primarily based on ‘pay for time’ (salary, benefits and perquisites — pay that is insensitive to individual or company performance. Over the last 50 years the design of executive pay in general and CEO pay in particular has shifted dramatically from ‘pay for time’ to ‘pay for performance’– cash incentive pay dependent on individual and company performance and equity pay (restricted stock and stock options) dependent on company stock price performance.
July 21 2009 by John W. Hamm
“I am concerned that executive compensation has, for a number of years, become
disconnected from actual long-term company performance.” Mary Schapiro, head
of the Securities & Exchange Commission, Los Angeles Times, May 24, 2009.
Fifty years ago CEO pay was primarily based on ‘pay for time’ (salary, benefits and perquisites — pay that is insensitive to individual or company performance. Over the last 50 years the design of executive pay in general and CEO pay in particular has shifted dramatically from ‘pay for time’ to ‘pay for performance’– cash incentive pay dependent on individual and company performance and equity pay (restricted stock and stock options) dependent on company stock price performance. The average CEO ‘pay for time’ in 2007 constituted only 23% of total pay, while equity pay was twice that amount at 56% and cash incentives made up the 21% balance.
Yet Ms. Schapiro of the SEC apparently feels that this ‘pay for performance’ is an illusion and that executive pay is disconnected from company performance. She is not alone. Corporate stock performance in 2008 was one of the worst in decades with the Dow Jones average down over 34%. So how did CEO pay in 2008 reflect this dismal performance?
- The Wall Street Journal annual survey found that the median total pay for CEOs at 200 large companies declined by only 3.4%.
- The AP analysis for CEO pay in 2008 for 309 of the S&P 500 companies fell 7% while stock prices for the group fell 36%, reflecting “…a mismatch between pay and stock performance throughout corporate America.”
- According to Corporate Secretary, RiskMetrics Group recently identified 960 companies with ‘CEO-Pay Disconnect.’
- “Executives are… rewarded richly when the company does poorly and the shareholders have been hammered.” Senator Chuck Schumer (D-NY)
Based on these studies and the opinions of government leaders, it would seem that ‘pay for performance’ is a mere illusion. Yet things are not always what they appear to be. Based on the analysis below, it seems more likely that:
- Pay for Performance in CEO pay is a reality, not an illusion;
- Executive pay does track the risks and rewards of shareholders; and
- The reason Ms. Schapiro, Senator Schumer and the media conclude there is a disconnect between pay and company performance is that Ms. Schapiro’s own SEC calculates equity pay, i.e., stock option and stock award pay (56% of the CEO pay package) as fixed ‘pay for time’ rather than variable ‘pay for performance.’
Let me clarify one point before I get too far down the road. This analysis is in no way intended to justify the overall size of CEO pay. Those who argue that CEO pay is obscenely high will find some of my conclusions very supportive, while those who defend the overall size of CEO pay as market-based and appropriate for the level of responsibility will also find some of my conclusions very supportive. In short, this paper focuses only on the effectiveness of the pay for performance design in CEO compensation, rather than the overall size of the pay package itself.
Why the Growth in Equity Pay?
Why did equity pay as a percentage of total pay grow from virtually nothing to twice the size of a CEO’s pay for time? Five words: “Pay for Stock Price Performance.” Stock options and restricted stock awards are designed to align the risks and rewards of executives with the risk and rewards of shareholders. The key reward driver of equity pay is future change in company stock price after the award is granted.
In a stock option or unvested stock award, if the post-grant stock price goes up, both the executive and shareholders are rewarded (think 2006 and 2007). If the post-grant stock price peaks and then falls, both the executive and the shareholders lose (think 2008). However, a stock option has a ‘stop-loss’ feature, limiting the executive’s loss when the stock price falls below the date of grant stock value. As we shall see, compensatory stock option are by no means free from significant risk of loss even with this ‘stop loss’ feature. Unlike the stock option, a stock award has no ‘stop loss’ feature (think restricted stock in General Motors).
Is it ‘Pay for Performance’ or ‘Pay Before Performance’
The SEC requires each public company to disclose in its annual proxy statement the Total Pay of its CEO. Amazingly, under the SEC rules, Equity Pay included in Total Pay for the year does not take into account post-grant changes in the company stock price!
Nothing is more important in determining the pay from an option or stock award than stock price performance during the option or award performance period. Aligning pay with future changes in stock price is the very reason for granting compensatory options and stock awards in the first place.
In other words, compensatory options and stock awards are tools designed to deliver future pay conditioned on future performance but the SEC requires companies to treat them as up-front pay regardless of future performance. Of course Ms. Schapiro finds a disconnect between pay and company performance: her agency treats options and awards the same as ‘time based pay’ like salary — a fixed amount insensitive to future performance. But this does not fit reality. Just as changes in stock price during the year impact the risks and rewards of shareholders, such changes also impact the risks and rewards of executives holding unexercised options and unvested awards.
It’s Kind of Like a Horse Race
Last May as the horses were at the gate at the Kentucky Derby, Friesan Fire was the favorite to win with post-time odds of 7-2. In other words, if Friesan Fire should win as expected, a bettor would receive $7 for each $2 wager made: $7 is the potential reward, the payment of which is dependent upon future performance.
A little over two minutes later, Friesan Fire was the 18th of 19 horses to cross the finish line and a $2 wager on the horse paid $0 reward for such dismal performance. So which is the more accurate description of the bettor’s reward for the race (pay for performance)?
- $7 for $2, determined before the race (potential reward), or
- $0 for $2, determined after the race (actual reward).
The former is the racetrack consensus immediately before the race and the latter is the actual pay received based on the performance.
Bad analogy you say because the performance period in the Kentucky Derby is only 2 minutes long while an option performance period can be as long as 10 years. But the 10-year option period can easily and accurately be divided up into 10 annual performance periods using change in stock price from the beginning and ending of each year to determine annual performance and reward. And more importantly, the sum of the rewards for each annual performance period will exactly equal the option profit received by the executive at exercise. You sure can’t say that about the SEC date of grant option value.
Bad analogy you say because there is a lot of “intellectual” authority that backs up the pre-performance stock option valuation formula (e.g., the SEC, the Financial Accounting Standards Board, and a Nobel prize for economics). Do you really think that a formulaic estimated value, no matter how widely recognized, has more weight and authority than the actual reward paid after the performance period at option exercise? Fortunately, this is the system authorized by no less an authority than the Internal Revenue Service.
If the proper time to value the reward for betting on a horse race is before the race, then why run the race at all? Obviously, they run the race to determine if the performance will, in fact, result in the payment of the potential reward. Similarly, if you are going to reward a CEO based solely on future stock price growth (a stock option), why calculate his reward before the performance period begins? Future performance (change in stock price) may result in a much bigger reward or it may result in no reward at all.
But the Economist Begs to Differ
The economist asserts that the SEC position of determining the value of compensatory stock options (Pay Before Performance) is appropriate since the compensatory event occurs at the date the option is granted and what happens to the option thereafter is not relevant for purposes of determining pay. Not relevant? Not relevant to whom? By contract design, the CEO can’t sell the option on the date of grant and unless the stock price increases thereafter, the option will continue to be worthless to him. To the CEO, the only reward from an option is future stock appreciation (Pay for Performance).
The economist compares the employer granting an executive a stock option with an employer purchasing on behalf of the CEO a $2 ticket on Friesan Fire before the Kentucky Derby. The CEO would have $2 of pay, since before the race the ticket is worth $2 regardless of what it might be worth after the race. Therefore, the issue is not what an option may be worth in the future but what it is worth at grant. But unlike the $2 race ticket that could be resold by the CEO for $2 before the race, a stock option cannot be sold — it can only be exercised and the only value that can ever be liquidated by the executive is future stock appreciation, if any. The IRS would tax the receipt of the $2 race ticket because it is negotiable, but does not tax the receipt of the compensatory stock option because it is not negotiable. In fact, if the employer purchased a publicly traded stock option and transferred it to the CEO, the IRS would tax the option at grant precisely because publicly traded options are, by definition, negotiable. But the whole purpose of a compensatory stock option is not to allow the executive to benefit from the option upfront (at grant), but only benefit from future stock growth (Pay for Performance).
CEO Pay Study Methodology
I tracked the CEO pay for the Dow Jones Industrial companies for the years 2006 (Dow up 16%), 2007 (Dow up 6%) and 2008 (Dow down 34%). I limited the study to 25 of the 30 Dow companies reporting pay data on a consistent performance period and those companies with the same CEO for most of the 3-year period (the “Study Companies”).
I converted CEO “Reported Pay” from the SEC proxy statements into “Revised Pay” by:
- Subtracting the reported value of “Option Awards” (before-the-race potential)
- Adding/ subtracting change in unexercised option values and unvested stock awards due solely to changes in company stock price for the year subject only to the ‘stop-loss’ feature of options.
Average CEO Pay for the 25 Study Companies
Listed below is the 3-year average stock price performance for the 25 Study Companies and the average SEC Reported Pay for the 25 CEOs:
Notice the lack of Reported Pay variance during the 3-year period in spite of the wide variance in stock performance. The average CEO received $9.76 million in Equity Pay in 2008 according to the SEC proxy statements even though the shareholders suffered a loss of 32% on their holdings in that year. Not a good picture; certainly not ‘Pay for Performance.’ But, I contend it is also not reality.
Now let’s look at the pay results taking into account changes in post-grant stock prices on the average CEO’s unexercised options and unvested stock awards during those years:
This is what Pay for Performance should look like and reflects reality. Revised Pay is highest in 2006 due to the 16% increase in stock price and next highest in 2007 with an 8% stock gain. The negative $18,444,262 equity pay in 2008 reflects the impact of the 32% drop in stock prices on the unexercised stock options and unvested stock awards held by the average CEO during 2008. This $18 million paper loss by the average CEO in 2008 is just as real as the paper loss of shareholders in 2008 with similar equity positions.
The average CEO (and shareholders) may well recoup this $18 million loss after 2008 if stock prices recover before the options expire or the awards vest, but that should be treated as future rewards in future years for future performance. On the other hand, they may never recoup such losses (think GM).
Example: General Motors CEO
At the beginning of 2008, the GM stock price was $24.89 and the GM CEO held over 3 million option shares, all under water except the 2006 grant of 400,000 option shares, which, if exercised at the beginning of 2008, would have yielded a $1,596,000 profit.
Early in 2008 the CEO was granted an additional million-share option, but by year-end with the stock price tumbling to $3.20, all the option shares, including the 2006 grant, were under water. In other words, during 2008, the CEO’s option position went from a positive $1,596,000 to $0 despite the new million-share option grant.
The SEC, however, reported that the GM CEO received over $8 million in option pay during 2008! Does this make any sense? In fact, the SEC reported that the CEO had received over $15 million in pay from options during the 3-year period 2006 to 2008, even though all his options expired under water by 2009 due to either his forced resignation or the demise of GM’s equity.
Yet, the press dutifully reported, based on the SEC’s pay charts, that while shareholders were losing nearly everything, the CEO had pocketed $15 million in pay from options even though he never got a dime from such options. This is not transparency; nor is this reality. You can say a lot of things about the CEO of GM, but enriched by options to the tune of $8 million in 2008, is not one of them.
Example: Exxon CEO
The CEO of Exxon held 464,545 option shares at the end of 2005 and during the next 3 years he received no new option grants and exercised 137,238 otherwise expiring options. I want to focus on the 327,307 option shares that he held throughout the 3-year period and compare this option-holding CEO with an Exxon shareholder who actually owned 327,307 shares for the same period:
The rewards for the Exxon shareholder and the CEO for the 36% and 22% stock price performance in 2006 and 2007 were exactly the same, namely, $6,696,701 and $5,583,857. Moreover, the ‘rewards’ for the Exxon shareholder and CEO for the negative 15% stock price performance in 2008 were also exactly the same — a loss of $4,536,475. The stop-loss feature of options did not come into play since the drop in stock price in 2008 did not put any of his options under water. Thus the CEO’s in-the-money options went down by $4.5 million during 2008, while the shareholder’s stock holdings also went down by the same $4.5 million.
So what did the SEC and media say his stock option pay was for 2006, 2007 and 2008? $0 and $0 and $0, because he did not receive any new options during those years.
Wait, you say, there is a big difference between shareholders who have to invest their own capital upfront, and executives who do not. That is a huge executive advantage over shareholders, but once the equity grant is made, this investment difference ceases to impact the alignment the executives and shareholders possess for upside opportunity and downside risk of loss (subject only to the stop-loss feature of options).
If you include the impact of changes in stock price on all of his equity pay (all of his options and his unvested stock awards), the difference in pay is quite significant. Notice the SEC understates pay in the good performance years (2006 and 2007) and overstates pay in the bad performance year (2008):
More to the point, viewed from a pay for performance perspective, the SEC’s Reported Pay in 2008 is up 34% over 2007 (from $16 million to $22 million) even though shareholders experience a 15% decline in value, while in reality, the CEO experienced an 81% reduction in Revised Pay in 2008 (from $35 million to $6 million) due to the 15% decline is stock price for that year. That is Negative Pay for Negative Performance.
Example: American Express CEO
The CEO of American Express had over 6 million unexercised stock options at the beginning of 2006 and, net of exercises and grants, had over 9 million at the end of 2008. American Express stock price appreciated 18% in 2006 but declined 14% in 2007 and 64% in 2008—dismally poor performance for which American Express shareholders certainly suffered. Yet CEO pay, as reported by the SEC, varied very little:
The American Express shareholders must have been outraged that their CEO was rewarded with a 4.8% increase in pay in 2008, a year their own holdings declined by 64% and that the difference between pay in a year of 18% stock growth (2006) and a year with 64% stock decline (2008) is a mere 6%. Clearly not “Pay for Performance”.
Now let’s look at the results if changes in post-grant stock prices are taken into account on unexercised options and unvested stock awards:
As of the end of 2008, none of the CEO’s unexercised option shares were above water, yet over the 3-year period, the SEC reported over $28 million in CEO option pay and $22 million in stock award pay. How does this make any sense?
Due to 2008 stock price performance, the CEO suffered a $79 million negative reward on his options and unvested stock awards aligned with American Express shareholders. This amount is after giving effect to his stop-loss feature on his options. So much for no risk of loss on stock options.
Example: McDonald’s CEO
During 2007, the stock price of McDonalds appreciated by 33%, decidedly good performance for the shareholders of McDonalds. The SEC Reported Pay for its CEO in 2007 was down 26% from 2006. This is certainly not consistent with a “Pay for Performance” strategy: reduced pay for stellar performance.
Yet, as a direct result of the 33% increase in share price in 2007, the CEO’s unexercised options appreciated by over $23 million and his unvested stock awards appreciated by nearly $2 million resulting in a 2007 Revised Pay increase of 29% over 2006 Revised Pay. Now that is stellar pay for stellar performance.
But wait, you say, isn’t that like counting increases or decreases of an employee’s 401k plan as part of his or her pay? No, I am not including as part of the CEO’s pay any increases or decreases in personally held assets; nor the changes in the value of his company stock holdings obtained from previously exercised stock options or vested stock awards. I only include changes in the value of his compensatory unvested stock awards and unexercised stock options.
Example: CEO of JP Morgan
The SEC reported the following pay information for the CEO of JP Morgan:
Now let’s look at the results if changes in post-grant stock prices are taken into account on unexercised options and unvested stock awards:
Let me conclude by illustrating my findings in the two side-by-side charts below.
In the chart on the left, the three bars measure the average CEO Reported Pay under the SEC Pay Before Performance rules for 2006 – 2008. Superimposed on that chart is a line representing the average stock performance for each year (up 16%, up 8% and down 32%). Notice the small pay variance even though stock price performance is volatile.
In the chart on the right, the three bars measure the average CEO Revised Pay (reflecting change in stock price) for the same 3-year period with the same superimposed line representing average stock performance. Notice how pay is way up in 2006 when stock performance is up 16%; not as high in 2007 when stock performance is only 8% and way down (negative, in fact) when stock price falls 32%.
One could easily conclude that Pay for Performance is an illusion under the SEC reporting rules illustrated in the left chart, but would conclude otherwise by observing the more realistic Revised Pay chart on the right.
A retired Ernst & Young partner and an attorney John W. Hamm, CPA, firstname.lastname@example.org served as an executive compensation consultant to public and private companies and their boards of directors designing pay for performance programs.