With stock options increasingly out of favor, cash bonuses are making a comeback. A recent survey of 350 companies by Mercer Human Resource Consulting found that the percentage of overall CEO pay coming from bonuses grew from 13 percent in 2001 to 19 percent last year. At the same time, the contribution of long-term incentives such as options and restricted stock fell from 71 percent to 63 percent, while the proportion stemming from salary rose slightly (see “Cash, Please,” at the end of this article).
“The size of bonuses is increasing exponentially as stock options fall,” says Paul Dorf, managing director of Compensation Resources, a consulting firm based in Upper Saddle River, New Jersey.
But if the form of incentive compensation is changing, the grounds for earning such pay may not be. And that could be a problem. Many critics of executive compensation complain that even after the end of the 1990s stock-market bubble, companies still cling to incentive schemes that encourage undesirable behavioraccounting shenanigans, earnings management, and value-destroying tactics.
“We’ve seen the ruination of many firms” as a result of compensation practices that encourage earnings management, says Michael Jensen, a renowned finance professor at Harvard Business School and a principal at Monitor Group, a consulting firm in Cambridge, Massachusetts. And while the incentive of choice during the market bubble was stock options, Jensen contends that few companies have responded by tailoring their bonus programs to discourage abuse. Doing so “takes a very deep understanding” of the pitfalls involved, he says, adding that such an understanding still escapes most CEOs and CFOs because of “the pressure they face from the capital markets and analysts.”
For a spectacular example of what Jensen is talking about, consider the most-recent troubles at Nortel Networks, a Canadian telecom-equipment maker that in January reported its first full-year profit in six years. A little more than three months later, the company disclosed that its earnings for 2003 were in fact half as much as the $732 million it reported. Three top executives were fired as a result: Frank Dunn, who was CFO until he became CEO in 2001; CFO Douglas Beatty; and controller Michael Gollogly.
Sure enough, Nortel had instituted a questionable incentive program for top executives in 2003. While executives would receive bonuses if they oversaw a “return to profitability,” they had great freedom to manipulate the results that qualified. “For the purposes of this program,” says the company’s proxy filing for 2003, “profitability is defined as positive pro forma earnings from continuing operations (which excludes acquisition-related costs and certain other items of a nonoperational nature).” That means the bonuses were to be paid for results that did not conform to U.S. generally accepted accounting principles.
Nor are problems with bonus targets limited to companies whose accounting practices have been called into question. Last February, for instance, Juniper Networks announced a $4 billion acquisition of NetScreen Technologies that critics claimed was driven by the company’s short-term incentive plan for top executives. The plan specified that part of the incentive compensation for these executives in 2004 would be tied to their success in “expansion of the business into new growth areas,” according to the company’s proxy. And, of course, the easiest way to do that is through an acquisition — even if it turns out to be a bad fit, as critics predict NetScreen will be.
Not What, but How
Some experts on corporate pay contend that a company’s choice of performance targets is less of an issue than how the targets are established. The fundamental problem, says Jensen, is that most companies base their bonus targets on internal budgets.
In an article entitled “Corporate Budgeting Is Broken — Let’s Fix It,” in the November 2001 issue of the Harvard Business Review, Jensen insisted that the standard practice of tying bonuses to the budgeting process “encourages managers to lie and cheat, low-balling targets and inflating results. It penalizes them for telling the truth.” He added that such an approach “turns business decisions into elaborate exercises in gaming…. And it distorts incentives, motivating people to act in ways that run counter to the best interests of their companies.”
Jensen went so far as to suggest that incentive compensation should be divorced entirely from the budget-setting process. There’s some evidence that companies were embracing Jensen’s criticism even before it was offered. A recent survey of 237 companies by consulting firm Towers Perrin found that the percentage of companies that tied bonus compensation to budgets fell from 69 percent in 1996 to 60 percent in 2001, the latest year for which such data was available (see “Beating the Budget,” at the end of this article).
Even some companies that tie internal budgets to performance targets may disconnect the process if gamesmanship rears its head. One example is DRS Technologies, a defense-electronics manufacturer based in Parsippany, New Jersey, that has grown from $50 million in revenues in 1989 to more than $1 billion today. If managers downplay their expectations in the budget process for the coming year based on the prior year’s results, the company will not set its bonus targets to reflect what it considers to be unacceptable forecast projections, but will instead set bonus targets higher than the forecast.
“We’re not going to pay them to break even,” says Richard Schneider, CFO of DRS. “We’re not going to pay a bonus for what amounts to less than superior performance.”
Some experts argue that gamesmanship can be limited by the use of market-based performance measures. The most widely used, perhaps, is EVA (economic value added), developed by the consulting firm Stern Stewart & Co. But the use of such metrics is no panacea in itself, and there are challenges involved. One, for example, is to determine just how much an individual manager below the corporate level contributes to shareholder returns. Doing so “takes a fair amount of financial work,” says Richard Ericson, a principal in the Minneapolis office of Towers Perrin.
And market-based targets are no less subject to manipulation than internally set ones. Says Jensen: “The game that the CEO and CFO are involved in with the markets is very similar to the budgeting game that goes on inside the company.”
Jensen concedes that the use of at least two performance targets that serve as a check on each other can go some way to addressing the problem. If, for instance, a company uses both sales growth and return on assets or invested capital, channel stuffing to inflate the first would lower the second, while eliminating valuable assets to pump up the second would lower the first. That way, adds Steve Harris, a compensation consultant in Mercer’s Atlanta office, “each measure is well grounded.”
But, notes Jensen, a manager must be given the weights to be applied to each of the measures to calculate an overall score, or there can be wide ranges where the manager cannot make a principled decision or can be subject to the whims of a superior evaluating the actual performance. Yet if multiple targets don’t check one another, they may simply give managers more ways to manipulate results. Bristol-Myers Squibb recently had to restate $900 million in earnings for approximately three years, largely as a result of misleading revenue-recognition practices, although its incentives at the time were tied to target levels of pretax operating profit, as well as cash flow from operations and sales. As it seeks to put its accounting problems behind it, Bristol has dropped sales and operating cash flow as the basis of its bonus plan in favor of pretax operating profit alone. Now, however, Bristol’s managers may be encouraged to cut costs at the expense of revenue and cash flow.
Other companies seem to be moving in the opposite direction. Towers Perrin’s survey found that almost 9 in 10 companies relied on two or more performance measures in 2001, up from 83 percent that used that many five years earlier. Nearly two-thirds of respondents reported that they used three or more performance measures in 2001.
DRS, for its part, has used revenue, operating income, free-cash flow, and bookings as the financial metrics for its bonus (along with three to five nonfinancial personal objectives) for the past four years. To borrow Harris’s term, the use of revenue “grounds” operating income against heedless cost-cutting, while free-cash flow does the same with temptations to inflate revenue.
But Harvard’s Jensen issues a note of caution here as well. If companies encourage managers to maximize an excessive number of competing targets, he warns, there’s a risk of creating a situation where “confusion reigns and there’s no accountability.” For this reason, he takes issue with the use of balanced scorecards, which can involve as many as 25 different performance measures.
DRS tries to avoid the problem by changing its measures when it finds some objectives being emphasized at the expense of others. Not long after implementing its plan eight years ago, for instance, the company switched one of its financial metrics from operating margins to return on assets. It has since switched it again, this time to free-cash flow. However, the total number of objectives hasn’t increased.
DRS does ignore one of Jensen’s recommendations: its plan pays out only 50 percent of a bonus if managers hit only 90 percent of their targets, and no more than 100 percent if they exceed their targets by 120 percent. If a company finds it can’t divorce bonus targets from the budget process, Jensen generally recommends that it at least set its incentive hurdles lower and caps higher than DRS does. Bonus hurdles and caps too close to budget targets, he warns, can encourage managers to inflate revenue or profits when that’s likely to put them over the lower-bound hurdle in a given year. Alternatively, they have incentives to defer them when they don’t think they have a chance to surpass the hurdle, or when they’re already up against the bonus cap.
But DRS has embedded a check on such behavior. The company includes bookings in its target mix because orders in the defense industry typically take at least a year to fulfill. That, CFO Schneider insists, keeps managers from taking actions that help them hit their other targets in one year but hurt the following year’s results.
Ronald Fink is a deputy editor of CFO.
Cash bonuses have been the fastest-growing component of overall CEO pay since 2001.
Composition of CEO compensation
|Beating the Budget
A survey by Towers Perrin of roughly 200 companies in 2001 found that the proportion that tied short-term incentives to budgets had declined to 60 percent from 69 percent five years earlier. But that was still more than eight times the proportion of respondents that based their bonuses on cost of capital.
|Basis of main performance measure||Percentage of respondents*|
|Determined by management board based on business conditions||49|
|Year-to-year growth or improvement||48|
|Company’s cost of capital||7|
|*Percentages total more than 100 due to multiple responses.
Source: Towers Perrin Annual Incentive Plan Survey, 2001