Wednesday, April 29, 2015

Is your CEO worth his (her) pay? The Pricing and Valuing of Top Managers!

It is the time of the year when stories about CEO compensation are the news of the day, and investors and onlookers alike get to ask whether a CEO can really be worth tens or even hundreds of millions of dollars, in annual compensation. Before I join the crowd, it behooves me to list my biases to start, because it will allow you to make a judgment on whether I am letting these biases color my conclusions. First, I believe that in some companies, CEOs not only add very little in value to the company but may actively be value destroyers, and that in many companies, what CEOs get paid is out of sync with the value that is created by them. Second, before you put me in the economic populist camp, I also believe the only group that should have a role in deciding how much a CEO gets paid in a publicly traded company is its stockholders, and that politicians, regulators and societal nannies should not get involved. Third, if you, as a stockholder, are disconcerted by the disconnect between CEO pay at your company, and CEO value added, you should be cheering on activist investors and pushing for more power to stockholders, rather than less.

CEO Compensation: The Landscape
Companies in the United States are required to break out and report what they pay their CEOs in summary compensation tables, filed with the Securities Exchange Commission (SEC). The numbers are still trickling in from 2014, but here is what we have learned so far:
  1. The early returns suggest that CEOs were paid more this year than they were last year, with collective pay increasing about 12.1% at the largest companies. 
  2. The portion of that compensation that was in cash increased to 37% from 35% in the prior year, with the bulk of the the remaining coming from stock grants (31%) and options (23%); pension gains (6%) and perks (2%) rounded out the rest. 
  3. The highest paid CEO in 2014 was Nick Woodman, CEO of GoPro, who was granted 4.5 million restricted stock units, valued at $284.5 million. 
Both the New York Times and the Wall Street Journal have visual links, where you can compare CEO pay across sectors, at least for larger companies, though the data is still from 2013.

It is true, as many others have pointed out, that CEO pay has been increasing at rates far higher than pay for those lower in the pay scale, for much of the last three decades. In the graph below, I look at the evolution of average CEO pay since 1992, broken down broadly by sector:

Source: Compustat ExecuComp
Since 1992, the annual compounded increase in CEO pay of 7.64% has been higher than the growth in revenues, earnings or other profitability measures.  Of all the drivers of CEO pay changes over time, none seems to be as powerful as stock market performance, as is clear in this graph going back further to 1965:

Source: Economic Policy Institute (EPI)
For those clamoring for legislative restrictions on CEO pay, note that it was a law designed to restrict executive compensation, passed by Congress in 1993, that set in motion an explosion of stock-based compensation that we have seen since. If you break down CEO compensation by its component parts (salary, bonus, equity-based compensation and other), you can see the shift towards equity-based compensation over time:

Source: Compustat ExecuComp

While there has been much talk about the ratio of CEO pay to that of an average employee, and that ratio has indisputably jumped over the last three decades, I found this ratio of how much a CEO gets paid, relative to the next highest paid employee in the company, for S&P 500 companies, using 2012-2014 data to be a more useful statistic.
Source: Compensation Advisory Partners
The median CEO is paid 2.30 times the next highest-paid employee at an S&P 500 company, which raises the follow up question of whether he or she adds that much more in value. While we can debate that question, the bottom line is that CEO compensation is large, rising and increasingly equity-based, that the growth in pay has accelerated in the last 20 years, and more so in the United States than in the rest of the world.

Determinants of CEO Pay 
Not only is CEO pay high, but it varies across time and across companies. There are three broad theories, not mutually exclusive, as to why you see that variance.
  1. Reward for performance: If a CEO is paid to run a company, it stands to reason that he or she should do well when the company does well, though you can measure "doing well" on three dimensions. The first is profitability, with higher profits (and growth in those profits) translating into higher pay. The second is the quality of the profitability, where the focus is on profit margins and returns on invested capital, with higher numbers on either measure resulting in bigger payoffs for CEOs. The third is to use a market measure of performance, by looking at stock price movements, with CEOs who deliver superior returns (either in absolute terms or relative to the market or the sector) getting fatter paychecks.
  2. CEO Market: If you start off with the presumption that it takes a unique skill set to become a CEO and that there is a market for CEOs, the compensation package that the company negotiates with a CEO will be determined by how the market prices his or her skills.
  3. CEO Power: If a CEO is powerful, he or she may be able to get a pay package (from a captive board) that is at odds with performance and much higher than what the market price would have been. There are multiple factors that determine CEO power, starting with his or her stock ownership. CEOs who own controlling stakes (and control does not require 50%+) in companies will be more powerful than CEOs that do not. The second is corporate governance, with all of the inputs that determine whether it is strong or weak; companies with weak or compliant boards of directors and little accountability to shareholders will be more likely to over pay their CEOs. The third is CEO tenure, since there is evidence that the longer a CEO stays in place, the more likely it is that the board will be moulded to meet his or her needs.
This excellent survey article on the topic  summarizes the evidence, and it is both supportive and inconsistent with each of these theories. Xavier Gabaix, my colleague at the Stern School of business, finds that dominant variable explaining changes in CEO pay over the last three decades has been changes in market value, with CEO pay increasing as market value increases. However, that theory in inconsistent with what happened in the 1950s and 1960s, when US stock market capitalization increased with no dramatic jump in CEO pay. There is some evidence that there are market forces at play, especially in the variation of CEO pay across sectors, but it is difficult to see how market forces can explain the rise in CEO pay in the aggregate. Finally, there is evidence that CEO pay is both more constrained and more tied to performance at companies where activist investors have stakes. At the same time, CEO pay seems to increase (rather than decrease) at firms, after they adopt good corporate governance practices or at least the appearance of such practices. In addition, none of these theories explain why CEO compensation in the US has gone up so much more than CEO compensation in the rest of the world. At the risk of muddying the waters further, I would suggest three more theories that may better explain the pricing of CEOs across time, companies and markets.
  1. The Stock Compensation Delusion: The delusion that stock-based compensation is cheap or even costless seems to be widespread among accountants, analysts and companies. Until the accounting rules changed in 2007, options were valued at exercise value in accounting statements, creating the illusion that you could give away millions of options to CEOs costlessly. Even though the accounting rules have changed, analysts and companies seem intent of reversing the effect of the rules, adding back stock-based compensation to earnings, a senseless practice that I have taken issue with before. One of the more dangerous consequences of this mistreatment of stock-based compensation is that boards of directors continue to be cavalier about granting large stock-based compensation awards to CEOs.
  2. The Momentum Game: The way companies set CEO pay is more evidence of the me-tooism that characterizes so much of what companies do in corporate finance, where they base how much to invest, how much to borrow and how much to pay in dividends/stock buybacks on what other companies in their peer group do. In deciding pay packages for CEOs, boards look at how much CEOs are paid at the peer groups (a subjective grouping to begin with). Like any other market, this relative pricing game can lead to CEO compensation packages climbing the ladder, since all it takes is one company over paying its CEO to set off overpayments across the entire sector. Put more bluntly, the answer that a board generally offers to the question of why it is paying so much to a CEO is that everyone else is doing it.
  3. The Celebrity CEO: In the last two decades, we have seen the rise of celebrity CEOs, more interested in developing superstar status than in managing the companies they are put in charge of, and boards that are willing to pay premium prices (in pay packages) to attract them. The consequences are predictable, with CEOs making big bets (even if the odds are against them), hoping that they will pay off, with the payoff measured less in profits and performance and more in social media mentions and online exposure. (Rakesh Khurana, a professor at the Harvard Business School and Dean of Harvard College, had made this point persuasively in his work on superstar CEOs.)
There is little new or original that I am adding to the discussion. There is a market for CEOs that sets compensation levels, but the prices that emerge from this market may have little correlation with performance and have more to do with mood, momentum and celebrity status. If you add to that the fact that directors seem to either have no interest or no understanding of the value of stock-based compensation grants, you have the makings of a perfect storm.

A Framework for analyzing the value added by a CEO
If the market for CEOs sets CEO pay, the big question then becomes how this market pricing measures up to the value that these CEOs can be expected to add to the company. To measure the value added (or destroyed) by a CEO at a company, you have to identify how the CEO affects the drivers of value. At the risk of over generalizing, here is how I see the effect of good and bad CEOs on value:

Can you actually value a CEO? I think it can be done, though it will require you to understand how that CEO plans to change the company and quantify the effects. Thus to measure the impact of a CEO on a company's value, you would have to value the company twice, once with an Auto CEO and the other with the CEO in question. With an Auto CEO, the company will stick with the tried and the true, doing what it has done historically in terms of market focus, marketing strategies and risk profile. With the CEO in question, the value effect will depend upon the changes you see that CEO making in the company on one or more of these dimensions. If you are willing to be specific about those changes, you can use this spreadsheet to see the value added by a new CEO and how much you would be willing to pay on an annual basis for that CEO.

I took the spreadsheet for a spin, using Microsoft as my example, partly because Satya Nadella, Microsoft's CEO, topped the list of highest paid CEOs last year and partly because I have been a stockholder in Microsoft since this post in 2013. If you assume that all that Mr. Nadella can do is slow the slide in margins, he will be able to add $7.74 billion in value to the company (about 2.25% of the company's value), translating into annual compensation of $670 million a year. While I do not agree with everything that Mr. Nadella has done over the last year at Microsoft, I think that he has done enough for me as a stockholder that I don't begrudge him his $84 million pay package.  I believe that Microsoft's board will have to monitor revenue growth and margins to see if he continues delivering, but this analysis indicates how dangerous it is to conclude that a CEO is being paid too much, just because he or she has a large pay package. Using this framework, we can make judgments on the types of companies that CEOs are most likely to make a difference, in good or bad ways.
  1. Life Cycle: A CEO should be able to make a much bigger difference in value early in the life cycle, when potential markets are still being defined and setting and having a coherent and consistent narrative can make the difference between winners and losers. As the firm matures, the CEO's capacity to affect value positively will decrease, though the capability of creating damage (by over reaching) may increase.
  2. Market and Competition: A CEO should have a much bigger impact on value in companies that operate in competitive businesses, where finding and maintaining a competitive edge separates the winners from the losers. In fact, there is research that backs up this contention, with CEO pay being higher, on average, and more tied to performance in more competitive businesses.
  3. Macro versus micro firms: I believe that the value effect that a CEO has on a company is more muted at companies whose value is driven primarily by macro forces (commodity prices, exchange rates, interest rates) than at companies where value is more determined by micro factors (markets targeted, pricing policies etc.). After all, if 80% of the variation in earnings across time is caused by oil price movements, there is not a whole lot that you can do, as a CEO, to affect earnings, an argument that Harold Hamm, CEO of Continental Resources, used in his recent divorce fight.
  4. Small versus Large firms: The value impact that a CEO can have at a large firm will be much higher in absolute terms than at a small firm, simply because the effects of small operating changes in the company can translate into large absolute changes in value. While this may seem to contradict the life cycle argument, we can reconcile the two, if we think about percentage changes in value. A CEO at small, young company will have a much higher percentage effect on value than the CEO at a larger, more mature company, but the latter will still have the larger effect on dollar value.
The skill set and qualities that make for a value-adding CEO will vary across companies. For a start-up or young growth company, the qualities that create a stand out CEO will be imagination, charisma and narrative skills. For a mature company, CEO greatness may stem from understanding capital markets and sector dynamics and less from vision and imagination. In decline, a company may be best served by a CEO who can deal with slipping revenues and shrinking margins, while managing the return of cash to stockholders and lenders in the firm. It should come as no surprise then, that what makes a founder CEO an asset early in a company's life may him a liability, as the firm matures, and that the skills that made a CEO successful in turning around one company may not be the ones that work in another. 

The End Game
If it is true that CEOs are priced and not valued, and that as stockholders in many companies, we are overpaying for our CEOs, what can we do to remedy this problem? I would not look to regulators, governments or tax laws to fix this problem, since these fixes (like the 1993 compensation law) not only operate as bludgeons, but have unintended and perverse consequences. I think that the answer has to be in good corporate governance in the true sense of the words, where shareholders are provided the information and the power to make decisions about CEO pay in the companies that they invest in, and use both effectively. I believe that information on CEO compensation should be revealed to stockholders on a comprehensive and timely basis, that shareholders should have a say on how much CEOs get paid and the power to replace directors who are casual about compensation. The SEC has moved in the right direction on all three fronts, with increased disclosure requirements on CEO pay, by requiring that shareholders be given a "say on pay" at companies and by easing challenges to incumbent directors. On each one, though, there has been push back from defenders of the status quo and the SEC has been unable or unwilling to go the distance.

Even in a world where disclosure is complete and shareholders are empowered, I recognize that giving shareholders the power to challenge and change management at public companies does not mean that they will use that power often or wisely. Just as voters in a democracy get the government that they deserve, shareholders in companies get the CEOs (and CEO pay packages) that they deserve. Paraphrasing Shakespeare, shareholders who complain loudly and often about CEO excesses should recognize that "the fault is not in their stars but in themselves", have to stop looking to others to make things right and start voting with their shares rather than their feet.

  1. The Value of a CEO 
  2. Valuing Satya Nadella's value to Microsoft
  3. Microsoft 10K and historical financials
Data Attachments
  1. Aggregate CEO Compensation breakdown by sector from 1992-2013


Unknown said...

Excellent article professor! A little elaboration on a specific determinant of CEO pay is merited here. The market for managerial talent mostly comprises of the top management directly under the CEO, since the search costs for identifying external managers fitting the CEO role are quite high, and that is why companies prefer to groom CEOs internally. Even then, CEO's pay acts as a motivator for the top management to compete for CEO's position. I believe this can actually incentivize the top management to compete through their respective team's performance for the highest prize i.e. CEO's managerial power and compensation. Too little can be demoralizing, and too much can drain resources. I see that internal labor markets also have an important role to play in deciding pay for the CEO.

Anonymous said...

How did you arrive at $670 million per year as value for MS CEO? Thanks.

Bill said...

Very interesting article Professor!

Regarding the Sales/Capital ratio,
I went into the spreadsheets and couldn't figure out why you used 3.55.

In the Global Industry Averages (tab), Software companies range from 0.88 to 1.03.
I also computed MSFT Revenues/Equity and got an average of 1.11.

Am I missing something?


Anonymous said...

How do individual investors by proxy (by way of large index funds become active? Seems like big ETF/Index funds aren't in a good position to be activists