The challenge to Compensation Committees is clear: ensure that compensation plans pass the test of common sense and reward top management for building the intrinsic value of the business. Compensation is the sharpest tool for ensuring that the CEO acts in the best interest of the company and its investors, and boards have to use it effectively. In addition, they need to align the CEO’s compensation with that of direct reports, so that the same principles drive the actions of the whole senior management team.

Boards must get a handle on CEO compensation once and for all. Pay for performance has long been the goal, but even wellintentioned boards have had trouble with it in practice. Something goes wrong in defining performance, measuring it, and matching rewards to it, whether it’s overrelying on a single measure of performance or creating complex systems that obscure the total package.

A whole new approach to CEO compensation is in order, one in which tax efficiencies don’t dominate and performance is measured by more than nominal stock price or any one other variable. Instead, compensation plans should be clear, straightforward, and built around a combination of objectives that reflect the board’s careful judgments about what is truly important for the company. Some of those objectives will be qualitative and therefore harder to measure, but this is where boards can shine by consistently exercising keen judgment and business savvy. As Jim Reda, managing director of James F. Reda & Associates and expert on executive compensation, says, “Boards have to get comfortable exercising discretion.” Mathematical formulas are no substitute.

Boards must exercise judgment, but they cannot be arbitrary. A compensation framework can provide the structure and rigor to get compensation right and make it fully transparent. Consistent use of the framework will build the board’s credibility with various constituencies.

Compensation consultants, HR departments, and Compensation Committees have important roles to play, but the whole board needs to get engaged in the following tasks:

Working on the compensation philosophy first before identifying a CEO’s performance objectives is a way to prevent that problem. The board should discuss what, in general terms, it wants the CEO to achieve. The philosophy has to capture the essence of what the board has in mind for the business. Most philosophies imply a balance between factors that are attractive to short-term investors and factors that build the corporation for the future. And most indicate what level of risk the board is willing to accept—or not accept.

The nature of the business will influence the time frame. Some businesses are inherently “long-tail” in that the real profitability of contracts signed today might not be evident until years down the road. The board of an insurance company, for example, will probably want to ensure a long-term view of the business, whereas a retailer or trucking company may be more short-term oriented.

Risk, too, depends in part on the nature of the business. An oil exploration or a mining company operating in countries where facilities could be appropriated by foreign governments may have to tolerate a certain amount of political risk. But to avoid compounding the risk inherent in the business, the boards of such companies might insist on superior financial strength.

The company’s situation is also an important consideration in the board’s philosophy. Is the scenario one of fast growth, turnaround, opportunistic acquisitions, or dressing up to be acquired? A board wouldn’t likely focus on cash generation, for example, if the company were in a rapidly growing industry. So industry dynamics and competition are considerations.

Johnson & Johnson’s record over the years suggests a philosophy that could be phrased as “steady performance improvement over the long haul, while making very selective transforming moves.” On the other hand, a company like WorldCom in the 1990s might have had a philosophy along the lines of “become the largest in the industry as fast as possible.” This philosophy may have contributed to some behaviors that were not in the firm’s long-term interest. Another company might look to fatten up its income or sales growth or both to prepare to be an acquisition target in a year or two.

The compensation philosophy is the starting point, but thinking the issues through in more detail provides greater assurance that the right behaviors will be rewarded, particularly in four crucial areas of the business:

Compensation Committees can be of great service to their boards by rolling up their sleeves, sorting out with management what a philosophy might be and how the CEO should behave in these four areas, and then discussing their thinking with fellow directors. Some Compensation Committees have convened off-sites dedicated to developing a compensation philosophy to bring back to the full board.

David Fuente, Chair of the Compensation Committee at Ryder Systems and at Dick’s Sporting Goods, and former CEO and Chair of Office Depot, describes Ryder’s process in these terms: “Before you bring the philosophy to the full board, you have to take the time to get your Compensation Committee off on its own for a day or two, so they form their own opinion of compensation philosophy and get a clear idea of the compensation programs that already exist. That way, the committee is fully informed and can lead that discussion at a larger board meeting.”

At Ryder, Fuente’s Compensation Committee did just that. “We basically got the executive vice president of human resources and the CEO to sit down with us and philosophically go through compensation: What role was it going to play? What various compensation programs were in place? Then we could critique them, in essence philosophically discussing where the compensation programs lined up with the strategic direction of the company.” Only then was the Compensation Committee fully prepared to go before the full board to develop the framework that links pay with performance.

Many boards make the mistake of putting their trust in a single objective, notably increasing total shareholder return or EPS, as a proxy for a CEO’s performance. But using a single objective rarely if ever captures the range of behaviors a board wants to encourage, and it creates room for people to game the system. It is the root of many a reckless acquisition spree that left the CEO richly rewarded and the company strapped with debt because the CEO threw caution to the wind in the single-minded pursuit of accomplishing sequential EPS growth or stock price appreciation.

Using total shareholder return (stock appreciation plus dividends) as a single objective is a problem in itself, especially when it is measured in absolute terms rather than in comparison with a peer group or the S&P 500. Contrary to the belief of some Nobelists in the dismal science of economics, the stock market is an indirect and often inaccurate measure of a company’s intrinsic value—the long-term franchise value of the company—at a given point in time. Stock prices are subject to the psychological whims of investors as well as to cyclical swings as valuation methodologies go out of fashion and are reinvented. As one successful hedge fund manager puts it, “The stock market has become a casino without a house.”

Indeed, the New Economy bull market demonstrated how large the gap can become between a company’s intrinsic value and its market capitalization, much to the chagrin of investors who came late to the dot-com party. In those cases, rewards that linked with stock performance alone had little connection to real corporate performance. In other extreme cases, they predisposed top executives to “make the numbers” or otherwise prop up stock prices by taking actions that, in fact, destroyed intrinsic value. Dennis Donovan, head of HR at Home Depot, notes that relying on stock market values for incentives when the market underrecognizes the company’s intrinsic value can be very demotivating to key employees—not what boards want for their managements.

Getting pay for performance right depends on choosing objectives that have a more direct connection with intrinsic value. These could include the number and quality of prospects in the drug pipeline or time-to-market for a pharmaceuticals company, for instance, brand strength for a consumer goods company, or customer satisfaction for an auto company.

To keep behaviors in balance, the CEO needs multiple objectives. They should reflect the board’s desired mix of short-term and long-term orientation, and they should not encourage more risk than the board is comfortable with. While there should be a mix of objectives, however, a CEO can’t be expected to pull two dozen levers. There are usually fewer than a dozen that capture the essentials.

A quick example shows how this might look in practice. Imagine a hypothetical discount retailer that is trying to regain its footing for the long term. If that’s what the board has in mind for the company, the philosophy would say that the company is willing to cede spectacular short-term performance to make sure the company can compete against dominant players, particularly Wal-Mart, and that it will be financially prudent in that pursuit. If it takes too hard a hit in the short term, it risks becoming a takeover target, something the board wants to avoid. Strengthening the company’s long-term competitiveness against the giants might mean finding a way to transform the company’s stores, but the company must avoid extraordinary increases in debt that would limit management’s flexibility in the future. Those are the behaviors that the board would like to see the CEO execute.

From this base, the CEO and the board must agree on the right set of objectives. Short-term objectives might include the following:

Not everything can be completed in one year, but progress must be made. A set of longer-term objectives establishes actions that will be partially completed during the year. In this case, they might be:

The board shouldn’t articulate specific initiatives. For instance, the board doesn’t have to define exactly how the CEO should differentiate the brand. It could be through developing new store formats. It could be through incorporating high-end design elements in merchandising. It could be through celebrity endorsements. It’s up to the CEO to develop that paradigm, which the board will later approve—just as long as it doesn’t involve taking on excessive debt, as stated in the fourth objective.

This set of objectives addresses the balance between short term and long term. There are many ways to improve operating cash flow—shuttering a number of stores and drastically reducing the SKUs could generate cash, for example—but not all of them will make the company more competitive in the long run. The long-term objectives are needed as a balance to ensure that the CEO protects the company’s ability to compete going forward.

Some objectives are easy to quantify and measure, but others require some translation. Qualitative factors can be assessed on a scale. For long-term objectives, the board could agree on milestones at the beginning of the year and measure progress in terms of percentage of completion. Ease of measurement should not dictate the choice of objectives.

The optimum balance between the two depends on the industry and the external conditions. The baseline could be 50:50. But in a commodity business such as copper mining, 80 percent cash and 20 percent equity might be more appropriate because there’s not much room to grow in that industry. In a growth industry such as high-tech, a good pay package might have more equity than cash to allow a higher reward for the higher risk. But letting any one element of compensation grow too large relative to the others could allow the wrong kinds of behavior to creep in. A huge short-term bonus, for example, could sway a CEO to miss some objectives in favor of those with a more immediate or bigger payoff.

The choice of peer companies is critical. It’s not enough to blindly accept the group of peers from the industry. Industry players often vary considerably in size and complexity. In some cases, a better set of peers are companies outside the industry that share characteristics such as size, opportunity, or maturity. Ten years ago, the board of a Baby Bell such as the predecessors to Verizon or SBC would never have considered Comcast or Time Warner in its comparison group. But times change. Ten years from now, the same Baby Bell might no longer consider AT&T in its comparison group. The Compensation Committee should carefully debate the list and discuss it with the full board.

There are decisions to be made about the cash bonus, too. When the tax deductibility of salaries was capped at $1 million in 1993, some boards began to award “guaranteed” bonuses to pay the CEO higher cash compensation. But Progressive boards do not treat bonuses as an entitlement. If they need to pay a salary higher than $1 million, they pay it. The bonus is only awarded based on honest judgments by the board on the CEO’s performance against specific objectives. How the bonus is to be awarded must reflect the board’s philosophy. At some companies, it is an all-or-nothing proposition, paid only if all the targets are fully achieved. At others, it is awarded on a scale. Some companies use an objective such as EPS growth, as a “toll gate” to be exceeded before bonuses for accomplishing other objectives can come into play.

Recently, there have been too many instances in which the financial performance of a company had to be restated, in some cases (such as Nortel’s) more than twice. Bonuses were not recovered from the CEOs. A positive trend is to make bonuses contingent on the accurate portrayal of financial performance. Reda has seen some boards putting their foot down by contractually seeking a repayment when this circumstance occurs. “It’s easy to do, and I’ve seen it done,” he says. One board puts the CEO’s bonus in escrow for three years.

The basic premise of equity awards is to instill a sense of ownership in CEOs and align their interest with that of long-term investors. But this concept can go too far. Equity’s value is in the long-term potential of appreciation, but there is also risk of a downturn in the industry or broader market. Thus it is dangerous to closely link the vesting or award of equity to the stock price at a given point in time, because capital market dynamics don’t always align with changes in the company’s intrinsic value.

In using equity-based mechanisms, boards have to think through the what-ifs of the market and the business cycle, on the upside and the downside. If the market booms as it did in the late 1990s, is the CEO rewarded for underperformance? If the market busts as it did in 2000, is the CEO punished because of external factors such as a looming recession? If so, the board could find itself paying a CEO much less for leading under dramatically more challenging conditions. Boards need to discuss how the CEO should be compensated in those scenarios, and be comfortable with the what-ifs of equity awards.

These days, there is much discussion of the form of equity granted to CEOs: stock options, performance share units, restricted shares, and innovations such as premium priced options or caps on options gains. Boards such as those at Microsoft and General Electric have taken steps to distance themselves from stock options for executives altogether. Experimentation with the delivery of equity-based awards is likely to continue, but the ones that work best are likely to incorporate more than one objective and use a long-term orientation.

Take the case of General Electric, which in September 2003 announced changes to its compensation policy for its CEO, Jeff Immelt. The objectives the board set forth reflect its compensation philosophy. Notably, Immelt was granted 250,000 performance share units (PSUs). Half of those units would vest as shares of common stock in five years if GE’s operating cash flow increased an average of 10 percent or more per year. The other half of those units would vest in five years if GE’s shareholder return outperforms the S&P 500 total return for the period.

GE’s board decided that two objectives—operating cash flow increases and five-year stock performance relative to the broader market—provide the proper incentives on which to reward Immelt over the long term. Operating cash flow, of course, is not only a clear and concrete indicator that GE is executing, but also an important element of GE’s precious AAA credit rating. In addition, GE’s stock was expected to outperform the broader market over five years, a long enough period of time that short-term volatility would smooth out.

There is another element to GE’s long-term incentive program, as announced in February 2003. Immelt can earn a maximum of 2.5 times his salary if the company achieves specific goals from 2003 through 2005 for, as its proxy statement reads, “one or more of the following four measurements, all as adjusted by the Committee to remove the effects of unusual events and the effect of pensions on income: average earnings per share growth rate; average revenue growth rate; average return on total capital; and cumulative cash generated.”

Every board must work out the details for its own company, and perhaps change the mix over time. GE is a broad-based company; thus the S&P 500 comparison for its PSUs makes sense. Other boards should carefully select their own comparison groups and their own measures of performance that reflect long-term performance.

The boxes of the framework contain the specific objectives. In that way, the framework shows how the objectives link with the specific forms of compensation. Think of the retailer I described earlier. That board defined nine objectives that operationalize its philosophy:

The first five items are shorter-term financial and operating objectives that match well with an annual cash bonus. The last four items have a longer-term orientation. Differentiating the brand against Wal-Mart, for example, is a multiyear strategic building block. It can be broken down into critical tasks and milestones that must be met year by year. This year’s milestone might be to make progress in developing new store formats—an assessment the board will have to make qualitatively. Because the task contributes to a longer-term objective, the reward might be similarly long term, namely a portion of equity that vests only after retirement.

To assess the objective to upgrade the pool of store, regional, and merchandise managers (in the category of human resources), the board could look at progress on several initiatives. For example, the Compensation Committee might review the steps the CEO has undertaken in the recruitment and training of store managers. Home Depot in the past three years has considerably strengthened its store management pool by recruiting some 420 officers from the armed forces, creating a dedicated store manager training program, and partnering with the AARP to recruit employees over fifty years old.

Exhibit 7.1 summarizes the framework for this company’s CEO compensation. It illustrates not only the balance in the objectives and in the compensation components but also the linkage between them. The framework gives a clear, quick picture of the cash and equity components of compensation. It gives the board a base to discuss what might happen to the equity component if the stock market booms (or busts), or whether the deferred salary program might create what one director calls a “Grasso Effect.” Compensation deferred over many years can accumulate into a lump-sum payout large enough to inflame public passions.

Reda notes that many boards are moderating their use of miscellaneous forms of compensation, such as supplemental retirement plans, loans, post-retirement health insurance, and the use of company planes. If such forms of compensation are in place, the board must include them in the framework. One board, after reviewing the total compensation package with the CEO, found it could eliminate a few perks, including a country club membership that the CEO said he never used. It’s an important discipline, and one that makes boards more comfortable with the notion of transparency. They need to be sensitive to public reaction.

On rare occasions, the board will need to make midcourse corrections to ensure that compensation is in fact promoting the right behaviors. What if the strategy needs to change in response to new opportunities, competitive moves, or a change in the business cycle? Maybe the old incentives are causing the CEO to postpone major moves for fear of missing short-term targets. Keeping the incentives relevant and working has a huge impact on the business. If the board has a robust process and a framework for looking at both pay and performance, adjustments can be made on a forward-looking basis. But boards should avoid adjusting awards that have already been made, for instance by repricing underwater options or accelerating the vesting of shares.

When the board has moved beyond mechanical formulas, it has to allow ample time and thought for considering how well the CEO has performed, particularly on the relative and qualitative measures. Boards are comfortable exercising judgment on qualitative factors as long as there is rigor in the process and in the collection of data. “Considering qualitative factors affecting financial results does not necessarily make the process arbitrary,” stresses Harvey Golub, retired Chair and CEO of American Express and currently Chair of Campbell Soup and THLee Putnam Ventures, as well as a director of Dow Jones. Compensation decisions made in the past, he concedes, “were too often idiosyncratic. Compensation Committees would say, ‘We thought the CEO did a good job so we gave him a package worth x dollars—at the 75th percentile of comparable CEOs.’ Executives certainly prefer the linear certainty that if you do x, you get y, preferably with y based on budget. But to my mind, the absence of judgment implied in that approach is just as bad.”

Accounting measures are only a small part of the picture. Getting a full picture of whether the earnings were really “earned” usually means looking at them relative to those of competitors and apart from uncontrollable or one-time influences. As Golub says, “Consider how the financial results match up not merely against the budget, but rather how the results compare to what competitors were able to do. That’s how a committee can assess the difference the CEO made. If EPS increases 10 percent but the industry went up 12, even though in absolute terms that may be good performance, in relative terms, it’s not.

“Then look at the quality of the earnings,” explains Golub. “For example, if your company makes its EPS targets, but misses EBITDA because currency exchange rates changed, obviously results are not as good as making EPS without those effects.”

Relative measures can be complex in practice. It’s not easy to compare when there are fifteen competitors, some of which are part of multibusiness companies, and each uses different accounting methods. Management can help, but boards are likely to want some objective advice. Reda says he has seen a distinct trend in boards hiring their own outside advisers to cut through the details and ensure that the measures are sound.

He even notes an emerging and intriguing practice: to conduct a compensation audit. He has seen a growing number of boards asking an audit firm to certify the results that the long-term payout is based on.

There are times when the board should consider unforeseen factors beyond the objectives it had earlier defined, and outside management’s control. Successfully navigating a crisis, for instance, usually merits reward.

Boards should make transparent not only the value of compensation and perks, but also the criteria used to assess the CEO’s performance bonus. If boards are up front about what constitutes performance, the public will see the judgments the board is making. Even if some don’t fully agree with the board, the public at large will see the diligence embedded in the process.

The payment is largely in the form of a signing bonus, but some of it is deferred. Then when the person leaves, the vesting is accelerated. The board is simply fulfilling its contract, but to the outside world, it looks like a handsome reward for dismal performance.

Disney’s Ovitz problem illustrates the public relations danger of lucrative contracts when severance is triggered. Ovitz’s fifteenmonth tenure earned him the maximum possible value that he could have received from his five-year employment contract. Disney wanted Ovitz as a potential successor, but in order to hire him, the company had to replace the pay package he owned as head of Creative Artists Agency. Unfortunately, when the news came out that Ovitz was being let go and receiving proceeds from the contract in full, shareholders went wild. Lawsuits are still pending at the time of this writing.

Compensation Committees and boards as a whole have to think through the contingencies in such contracts when they make the hire. If it doesn’t work out, how much will it cost? Is it worth the risk? Tax efficiencies shouldn’t drive decisions. And if necessary, boards should consider putting the money in escrow and defining the conditions under which it can be accessed. Above all, the total package, including the signing bonus, should be transparent to shareholders, so there are no surprises.

The second place to turn is compensation consultants. For some period of time, however, there has been an overreliance on compensation consultants to set CEO pay. Too often the consultants were selected and paid by the company and its CEO, not by the board. That’s a dangerous situation if the consultant’s recommendations are not placed in proper context.

It’s the board’s job to set the CEO’s pay. The board owns this process. It’s okay to have a consultant involved, but it’s imperative for the board itself to discuss the philosophy and behaviors, define objectives that link to the intrinsic value of the corporation, identify the correct peer groups of companies against which performance should be examined and median salaries compared, and evaluate performance.

Along the way, a consultant’s input can be very valuable. Compensation consultants typically have a wealth of knowledge of ways to structure a package within a context of desired short- and long-term goals, and of the tax consequences of various elements of compensation. Tax consequences are subsidiary to the compensation philosophy, but they are a consideration.

Consultants also have databases of current CEO pay that provide important context. After all, in most cases, a CEO’s base salary must be competitive to prevent poaching by other companies. But it is up to the board to define the unique problems that face the company, the ones the CEO will be compensated for solving. The board, not the consultant, has the knowledge to define that context.

Most companies use benefits consultants to support the CEO and the HR department. Boards should hire a different firm to advise them and support their own work. In this era when outside observers are scrutinizing independence, both the perception and reality of a potential conflict of interest are important considerations. Thus, the Compensation Committee should use a consultant that is independent of management and that can give the board the support it needs.