Major companies use questionable peer groups to set CEO pay

To keep talented CEOs, most publicly-traded companies peg pay packages to compensation at so-called peer groups of firms they say have similar revenue, market value or profits, or are in similar industries.

But an Atlanta Journal-Constitution analysis of the proxy statements of dozens of Georgia companies shows how some big Georgia firms tip the scales in their CEOs’ favor by using peer groups that are not similar. The AJC has been looking into executive pay issues because they have gained increased attention following the financial crisis and Great Recession. Congress has passed legislation giving shareholders more say on executive pay.

Some local companies appear to be cherry-picking bigger or higher-paying firms as peers, or inflating their executives’ pay targets compared to peers:

— Equifax uses a 14-company group in which the typical “peer” has 89 percent higher profits and 68 percent higher revenue than it does. Still, the Atlanta-based credit reporting firm paid its CEO nearly twice as much as his peers last year. CEO Rick Smith got a 20 percent pay raise last year, to $13.4 million. The median pay for his peer group: $6.9 million.

— Coca-Cola Enterprises is now too small to even be a member of the two peer groups it uses, after selling off three-quarters of its operations to Coca-Cola in 2010. But the Atlanta-based bottling company chose to keep using its old peer groups, paying CEO John Brock $10.2 million last year — up to 10 percent higher than his typical peer.

— Coca-Cola, on the other hand, retooled its peer group after the CCE deal increased the size of the Atlanta soft-drink company. Coca-Cola ditched nine beverage and food-industry companies from its old peer group and added four generally higher-paying companies — Apple, IBM, Wal-Mart and AT&T. Only one beverage company remains — Pepsi.

Coca-Cola said it chose companies with well-known brands, global operations and market values above $100 billion. But it kept several companies with highly-paid CEOs that didn’t meet all those criteria, while dropping one firm that did — Anheuser-Busch — whose CEO was paid $5 million last year.

Coca-Cola denied that CEO pay played a role in its choices.

CEO Muhtar Kent’s 2012 pay of $30.5 million was still higher than all but one of his peers, and more than double the peer group’s median pay, or midpoint.

John Bizjak, a Texas Christian University finance professor, has found that companies often create biased peer groups to help boost executives’ pay.

“There’s some monkeying around,” said Bizjak. “That’s maybe led to higher pay raises than there should have been.” And that can cost anyone who directly owns stocks, has a retirement plan, or pays taxes to federal, state and local governments whose employee pension plans are invested in stocks.

Some Georgia companies picked peer groups of firms similar to theirs by most measures. But then they made it company policy to pay their CEOs more — sometimes much more — than their peers.

IntercontinentalExchange, an Atlanta-based financial exchange, sets executive salaries and bonuses to be higher than 75 percent of its peers.

Newell Rubbermaid pegs top executives’ salaries to the median pay for its peer group. But the stock and bonus awards — the bulk of their pay — is pegged to exceed 65 percent of the peers.

CEO pay has skyrocketed far beyond corporate profits and other performance measures in recent years. While the S&P 500 firms’ profits have risen 89 percent since 1989, after inflation, CEO pay at those firms has risen more than 300 percent.

Biased peer groups is one reason for the disparity, said several researchers. But such groups don’t automatically mean firms are padding executives’ pay. Experts see evidence, for instance, that some boards of directors have countered the effects by setting executives pay raises below the compensation targets tied to the peer groups.

Keeping talent on board

Why do companies create top-heavy pay targets?

Many firms argue they need to offer above-average pay to attract and keep talented and experienced managers, and encourage top performance.

A higher pay target “encourages (executives’) efforts to increase stockholder value,” said Newell Rubbermaid in its recent proxy filing to the SEC.

Corporate boards tend to be “risk averse,” said Joel M. Koblentz, a longtime recruiter of top executives. They don’t want to set pay too high or too low, he said, so they use peer groups to figure out what other companies are paying. But they also want to set pay high enough to avoid losing top performers, he said.

“They want to stay in a defend-able position,” he said.

Coca-Cola Enterprises, for instance, kept its old peer groups after selling most of its operations to Coca-Cola because “(maintaining) senior officers’ level of compensation following the transaction was critical to maintaining management continuity,” the firm said in an emailed response to the AJC.

In its recent proxy filing, Equifax said it pegged pay to larger companies to “provide sufficiently competitive pay to attract and retain experienced and successful executives … from companies larger than ourselves.”

Typically, the board of directors hires an outside consulting firm to collect pay data and set up peer groups, usually based on a dozen or so firms that are similar in terms of industry type, revenue, market value, profits and other factors.

But there’s little consistency in how firms do this.

Most companies pick a dozen or a couple of dozen comparison firms. Others, like CCE, pick hundreds.

Some say they peg pay to the median, or midpoint at which half the peer CEOs’ paychecks are higher and half are lower.

Others are vague about how they use the pay comparisons. Coca-Cola, for instance, said its peer group “is not a primary factor” in setting executive pay, which is also based on the company’s and individual’s performance.

“Mr. Kent’s leadership has directly contributed to the company’s strong performance over the last several years, and specifically in 2012, and should be appropriately rewarded,” Coca-Cola said in an emailed response.

Some firms, like Aflac, SunTrust and Equifax, stick with companies in closely-related industries. Others span the economy. Coca-Cola includes makers of drugs, computers, sneakers, software, aircraft and burgers. CCE’s 220-firm group ranges from Delta Air Lines to Google to Walt Disney.

Biased pay groups

Critics and some academics said these widely varying approaches give companies a lot of leeway to create biased peer groups and set inflated pay targets. The result, they said, is a rigged game that ratchets up CEO pay.

“Everyone can’t be above the 75th percentile,” said Bizjak, the finance professor at Texas Christian University.

If everyone’s pay is pegged to the midpoint or higher, that’s going to push pay higher “almost by definition,” he said. Last year’s laggards catch up to the median, pushing the group’s midpoint higher next year.

Bizjak and his co-researchers found that many companies’ peer groups gravitated toward bigger, better-paying, or more profitable firms that nudged pay comparisons upward. The trend was especially true at smaller public companies. Their typical “peers” were 19 percent larger, the researchers found.

There are hints that corporate boards now may be feeling more pressure from investors to re-think the issue. Since 2011, the federal Dodd-Frank financial reform law has required companies to give shareholders a non-binding “say-on-pay” vote on their executive pay plans.

Bizjak said it’s too early to tell how such votes have impacted executive pay levels. But in one study, his group found that corporate boards were awarding smaller raises than dictated by the firms’ peer comparisons.

One corporate governance consultant that advises pension funds and other big investors on say-on-pay votes now creates its own peer groups to check whether companies are being honest.

Institutional Shareholder Services said it devises peer groups of 14-24 firms as similar as possible to the target firm to see how its CEO’s pay compares.

‘Gaming continues’

Despite such efforts, one recent study suggests that companies are manipulating their peer groups even more since 2006, when the SEC began requiring firms to show their cards by identifying the composition of their peer groups.

In a recent study, the University of Maryland’s Michael Faulkender and Indiana University’s Jun Yang compared peer groups before and after the SEC rule took effect.

After the SEC-mandated disclosures began, they said, companies added more firms with higher CEO pay to their peer groups, and deleted lower-paying firms.

“The gaming of peer benchmarking continues,” they concluded.

Such trends prompted University of Delaware researchers Charles Elson and Craig Ferrere to suggest that it’s time for boards of directors to ditch peer groups.

In a 2012 study, Elson and Ferrere looked at whether peer groups are based on the wrong idea — that CEO pay levels at one company are useful for setting CEO pay at another company.

“The only theoretical justification for (peer groups) is that talent is transferable” from one company to another, said Elson. If that’s true, he added, you should see many more companies hiring CEOs who headed other firms, rather than promoting from within, he said.

But after checking an academic database that has tracked executives’ moves at 1,500 companies since 1994, they found few such CEO transfers at healthy companies.

“There were only 27 of those in the entire database,” said Ferrere. The CEOs that did jump from one industry to another often flopped, he said. Usually, he added, most companies promoted CEOs from within their own executive ranks, indicating that deep knowledge of the industry and the company is crucial, he said.

The idea behind peer groups, said Elson, is “flawed at its philosophical core.”