ON FURTHER REFLECTION
Do EVA and other value metrics still offer a good mirror of company performance?
By Bill Birchard and Alix Nyberg

The first thing Jim Rutledge did when he became CFO of Baldwin Technology in January 2000 was throw out the performance measurement system. Not that he was shunning the idea of measuring performance. Far from it. The problem was the confusion that surrounded the economic profit program that Baldwin had designed with the help of a consultant, Vanguard Partners. "There was a lot of mysticism around that," says Rutledge of the value-based program that had been the guiding light for Baldwin's FInancial management for three years.

Employees had "no crisp understanding" of how their behavior affected the US$200 million printing equipment, controls and accessories maker's "economic profit" - generally defined as the net operating profit after tax of a company's businesses, reduced by subtracting a charge for the cost of capital. In particular, Baldwin executives had to make adjustments for such items as transfer pricing to calculate the numbers used to determine, for instance, the size of their incentive compensation.

Adding to its value-metrics problems, returns at the Connecticut company became uneven - whether measured by earnings per share or economic profit - thus slashing the bonus potential. And the difficulties brought on operational changes, including a revamp of product lines and a move toward globalization, along with downsizings and divestitures, that made keeping track of the shareholder-value calculations all the more burdensome. In short, says Rutledge: "Moving all the way to an economic profit model was too much."

In jettisoning its program, Baldwin was hardly alone. Indeed, between 40 and 50 percent of all companies trying value-based metrics abandon them between the third and fifth year of implementation, according to Jim Knight, a partner at US-based SCA Consulting, which helps companies structure value-based metrics. Among other companies dropping their economic-profit approach, or subordinating it to more-traditional metrics such as earnings or return on capital employed, have been AT&T, JC Penney, Tenet Healthcare and Armstrong Holdings. Some of these used a form of economic value added (EVA), the program trademarked by Stern Stewart, the leader in the economic-profit consulting field.

Many decisions to discontinue value-based measurements are made because plans are "incorrectly designed to begin with, and don't reflect the business strategy of the companies," suggests Knight. Stern Stewart, which has emphasized the training of rank-and-file employees recently, says that only about 5 percent of its 200 full-fledged clients have actually discontinued EVA. It acknowledges, though, that a number of companies adopting value-based systems may keep them in name only after failing to implement programs properly - for example, by choosing not to tie compensation to the metrics.

No matter what the exact size of the exodus from shareholder-value metrics - or the reasons for it - questions remain. Does using economic profit help a company deliver on its performance promises over the long term, or does it harvest the low-hanging fruit of cost reductions for more of a one-time boost? Is the bloom off the value-metrics rose when bonuses evaporate? And how can a company boost shareholder value year after year?

The Enthusiasm Fades

To be sure, some mechanics necessary for improving the benefits of EVA or other value metrics are well known: continual training, lots of communication, enduring CEO support, meticulous accounting and careful design of compensation terms to provide true incentives. These can help counter the two standard criticisms of economic-profit programs: that they discourage managers from investing in the business, and that they require inordinately complex calculations among business units and divisions that share corporate services or assets.

But a look at why some value-metrics users have chosen to either drop, change or maintain their programs suggests that results hinge on some less-understood challenges. These include managing the metrics during times of corporate transformation; confronting employee concerns about perceived bonus inequities; and identifying the right "drivers", the individual performance indicators that become targets for employees' efforts to boost the company's shareholder value.

For assembly-line workers at a manufacturing company, drivers might include working-capital ratios and output per employee, while customer satisfaction might fit into the formula for a plant manager, for example. "This is the blocking and tackling that makes economic profit work," says Roy Johnson, a partner with Vanguard, based in Connecticut. "If companies don't do this, they're not getting at the root causes" of the failure to create shareholder value. (Baldwin, Johnson says, suffered from a failure to maintain its system using techniques Vanguard provided.)

"You Have to Tend to It"

At AT&T, the inability to adjust the program to reflect a drastically changing company seemed at the heart of the problem. The company implemented EVA in 1992 and 1993, and extended an EVA bonus plan to its entire white-collar workforce of more than 100,000 people. But then Ma Bell "struggled to reset EVA targets after Lucent Technologies and NCR were spun off and AT&T Capital was sold" in 1996, according to Stephen O'Byrne, president of Shareholder Value Advisors, in New York, and a former Stern Stewart consultant (see box below).

Behind the restructuring and associated problems, though, AT&T was looking at value-based metrics as a "panacea", he says. Company managers "came to EVA with tremendous enthusiasm, but no specific understanding of how EVA was going to help them." O'Byrne, who co-authored the book EVA and Value Based Management, says AT&T also "suffered from a lack of commitment to EVA as the sole basis of their nonstock compensation," thus diluting the power of the model to deliver results. In the face of challenges, AT&T managers "found it easier to take alternative routes, like setting new goals or adopting new measures," he says. Recently, AT&T replaced EVA with various expense-to-revenue ratios, along with EPS, in bonus calculations. (AT&T declined to comment on EVA's discontinuation.)

It was largely a driver-related problem that killed value-based metrics at Baldwin Technology. Upon his arrival, Rutledge found managers puzzled about how the drivers they were using - improvements in inventories, receivables and cashflow, for example - would work to boost economic profit. "They debated and wondered if they were right" in trying to adjust their behavior to get a certain result, he explains.

And then there was the dwindling-bonus syndrome. From a 1998 peak of US$9 million in earnings for its fiscal year ending June 30, Baldwin earnings plunged to US$4.8 million in fiscal 2000. At the same time, the economic profit number sharply reduced bonuses, which had been hefty in the first year. "It did give rise to a lot of ill feelings," admits Rutledge, when bonuses collapsed in 1999 and 2000.

Earnings seem to be on the mend, with first-half profits up 40 percent over the same period last year. But any bonuses will be based on a combination of cashflow and EPS, not economic profit. "You really have to tend to it," says Rutledge of the value-metrics program.

Today, Baldwin's finance department calculates cashflow and working capital goals for each unit, and managers understand those conventional numbers more easily, says Rutledge. He describes the new process as "almost like breaking down the [shareholder-value] formula." So far, the feedback is positive. "They love it," he says.

A Different Algebra

Embracing value-based metrics fully means taking the time to adjust their inner workings to fit your company's needs. At Briggs & Stratton, the Milwaukee-based small-engine maker, president John Shiely believes that adjustment of the drivers is often necessary to make sure the workers actually have the power - something he calls "decision rights" - to improve economic profit with their actions. What sets good value-metrics operators apart is a company's ability to "match the performance metric and the bonus with the decision rights" of employees, says Shiely.

That isn't always easy. A decade ago, for example, managers at Briggs & Stratton, a longtime EVA user, measured the performance of the company's big, unionized engine plant with EVA. But, eventually, they realized it was wrong for the factory, and started using productivity instead. The reason: most of the plant's 1,000 workers couldn't affect any decisions relating to capital expenditures.

At the company's nonunion foundry, though, managers found that EVA was so relevant that they could measure it directly, and not even bother with intermediate drivers. Most of the 100 workers there believe they can affect decisions that relate to capital spending, and Briggs & Stratton has found that it can train them to understand just how that occurs. "The algebra is different" in such a small plant, says Shiely. The managers themselves have identified EVA's drivers at the foundry: molding efficiency, uptime, scrap rework and attendance. Finding the drivers that can visibly help employees measure their performance "is the big issue as you push [EVA] down in the organization."

Thinking Long Term

Still, that other question - how to keep managers when their bonuses dry up - is a biggie, too. At Briggs & Stratton, the approach is to try preparing people for how organizational, product and strategy changes will affect EVA and the resulting compensation, and to hope this information keeps them motivated.

The first challenge started in 1995, Shiely recalls, five years after EVA had become a fixture and the company had captured the easiest gains, chopping out excess capital and improving capital efficiency. Top managers received significant bonuses back then. But the next step to growing EVA was to embark on a strategy to build three small "focus factories". And that led to a problem: according to forecasts, EVA would plunge for two years, wiping out EVA-based incentive pay.

Hardest hit were managers building the product base within those new plants. The investment dollars lavished there burned their EVA and their bonuses. "They knew they'd have to bite the bullet short term," says Shiely. But he didn't consider changing the reward formula, although he was well aware that managers would not like the new growth strategy. In keeping with the Stern Stewart model, EVA bonus targets are set higher than each previous year's actual results. To help discourage short-term thinking - and eliminate the prospect of a no-bonus year - the model also spreads out bonus payments over several years through a "bonus bank".

Shiely says that, except for normal retirements, Briggs & Stratton didn't lose a single top manager during that down period. The information strategy, he says, appealed to a management team peopled with long-term thinkers. "If you're the kind of company that goes back to the incentive program that everyone else has, you get the kind of people everybody else has," he says. "The guys who want an automatic bonus [every year] go to work for another company, and that's fine with me," adds Shiely. Of course, many of those managers had enjoyed big EVA bonuses before, and figured - correctly, as it turned out - that they would again.

Still, bonus evaporation is often seen as the Achilles' heel of value-based metrics - and a major cause of plans being dropped. At companies he has worked with, says O'Byrne, people loved the early years of higher capital efficiency and rich bonuses. But "when the shot in the arm wore off, they went off to look for the next one." O'Byrne's research shows that the companies that stuck with economic profit tended to adhere to strict compensation practices - calculating bonus targets from an automatic formula that requires increases in economic profit each year, leaving payout rates uncapped, and using the bonus-bank approach.

Bonus banks, though, can have their own downside. Michigan-based US$1.9 billion furniture maker Herman Miller uses EVA to determine bonuses for both executives and rank-and-file workers, under two separate formulas. In the past two years, Herman Miller has authorized heavy capital investments, much like Briggs & Stratton. And it knew that when it decided to develop a new line of office furniture - featuring a "no-panel, 120-degree work system" of desks, separating screens, and lighting - the spending would punish EVA and the EVA-based incentive compensation.

When it told workers that fiscal year 2000 would be "bonus-free", saving the company US$42.6 million, "no one was happy," says CFO Beth Nickels. But employees were especially irked because executives still got incentive pay for the year, flowing from their three-year EVA pooling. (The employee plan had no provision for pooling bonuses.) The bonus omission was further galling to employees, because the company had earnings in the strong economy that year of US$140 million, about flat with the prior year.

The employee discontent, in turn, made corporate directors worry that EVA bonus provisions might be demoralizing to workers. Nickels had to educate the board. "It's doing exactly what it was supposed to do," she told them. When new investments started to pay off, EVA bonuses would surge, she asserted.

Even with the bonus-free year, total employee payouts over the past several years have yielded more than they did under the company's former profit-sharing plan, replaced when Herman Miller installed EVA. The alternative would have been to cut jobs to contain costs.

While directors were mollified, Herman Miller still made a concession, issuing employees a one-time, 100-share option grant. Indeed, since then, EVA-based employee bonuses have come back.

No Reward for Accuracy

But while economic profit is often designed to insulate executives from a three-year downturn, it may not be able to help them dodge a bigger bullet, as Armstrong Holdings recently had to do.

The Pennsylvania-based floor-and-ceiling materials maker replaced return on assets with Stern Stewart's EVA program in 1995, and at first scored consistently positive EVA returns, topping its 11 percent cost of capital, until last year. But calculations didn't include the hundreds of millions of dollars in asbestos-related lawsuit liabilities - a number expected to climb as high as US$1.4 billion by 2006 - and a reorganization charge that it capitalized. As the stock quote plunged, management decided to choose new metrics to stem the tide.

Last year, Armstrong added profit to EVA, and this year, it replaced EVA entirely, substituting cashflow. It chose profit and cashflow, says CFO E Follin Smith, "because we want to reward growth and accuracy and meeting budget commitments, which EVA doesn't capture." She says that management's drive now is "the need to create a meet-the-numbers culture, a meet-your-projections culture."

Armstrong filed for protection under Chapter 11 of the federal Bankruptcy Code in December. But Smith maintains that the filing had nothing to do with the move away from EVA. "Different behaviors," she says, "have to be encouraged at different points in a company's life."

Bill Birchard is a contributing editor, and Alix Nyberg is a staff writer, at CFO in the US, CFO Asia's sister publication.