Buffett and Coke's Sweet CEO Pay
Warren Buffett opposes (though he didn't vote against) Coca-Cola Co.'s new employee stock plan as excessive. The problem, which Buffett didn't mention, starts with the chief executive officer, Muhtar Kent. Although Kent is far from America's most overpaid CEO, his compensation demonstrates everything that is wrong with executive pay.
Coca-Cola's proxy statement jumps through hoops to demonstrate fidelity to the corporate governance ideal of "pay for performance." In reality, Kent's pay is rigged to remain at very high levels, with modest downward adjustments in bad years and potentially astronomic rewards in good years.
Kent isn't, to repeat, the poster child for compensation abuse. (That prize goes to Oracle Corp.'s pathologically greedy Larry Ellison, who collected $78 million last year.) Coke's compensation policy even has some admirable features, such as inhibiting the incentive to take short-term risks.
Yet the numbers show a disconnect between Kent's compensation and shareholder rewards. At Coke, a consumer icon struggling to re-energize a passe product, stock performance has trailed the Standard & Poor's 500 in each of the last one-year, three-year and five-year calendar periods.
What's more, the degree of underperformance has been significant: Coke's stock sported roughly half of the Standard & Poor's 500 Index advances of 30 percent and 46 percent over the past one and three calendar years, respectively. Meanwhile, Kent's annual compensation over the past five years totaled $95 million. (Disclosure: I own stock in Berkshire Hathaway Inc., which is an investor in Coca-Cola.)
How do you earn $95 million over five years for consistently underperforming the benchmark? Mostly, Coke does it through stock and option grants. CEO options are supposed to provide an incentive for truly superior performance over the long haul. (Remember, CEOs already get boodles of dough from other parts of their pay package.)
However, at most companies, including Coke, options provide a reward for any increase above the price level at the grant date over a period of 10 years. To put it mildly, this is a very easy mark -- one that can result in a fortune for mediocre performance.
Last year, Kent got 1.9 million options exercisable at $37.61 a share. If Coke's stock advances at a 4 percent annual rate over the options' 10-year life -- a very modest clip -- his grant will be worth $35 million.
Rather than pay for performance, this is "pay for showing up to work in the corner office." If the stock advances at a merely respectable 7 percent clip, his grant will be worth nearly $70 million.
The other serious problem is that Kent gets a boatload of options every year. Large annual grants corrupt the nature of the incentive. Instead of having to perform over a single, pre-designated period, serial option recipients need only preside over a rise in the stock price over any of a series of rolling time frames.
For example, were the stock to fall in one year, the next year's options would be that much cheaper and more attractive.
And since the grants at Coke are so large, Kent only has to cash out once to become stupendously rich. Keep issuing giant grants and one of those awards, retrospectively, will likely prove a winner. But the opportunity to earn hundreds of millions should not be recalibrated every year. (Shareholders, after all, cannot re-price their investments every year.)
Coke has issued so many options and shares to Kent and other employees that it is running out of authority to issue them. That is why it pushed through a new stock plan last week. The plan authorizes up to 500 million employee options over four years -- which combined with existing plans could lead to dilution, Coke says, of 14 percent. David Winters, a mutual-fund manager who opposed the plan, says, "It materially dilutes the owners of Coca Cola if fully implemented."
Coke pats itself on the back for not re-pricing options when its stock falls, but granting new options every year potentially achieves a similar result. Its proxy statement urges shareholders to celebrate its supposedly enlightened policies. "We heard that there were certain elements of our executive compensation programs that some shareowners believed we should consider doing differently," it says. "We listened and have incorporated this feedback into several enhancements."
Coke asserts that its compensation process is better and clearer. However, it still resorts to self-interested spin by downplaying conventional metrics such as earnings per share and highlighting its own, more stylized yardsticks in which performance appears better. Coke also presents its pay process as rigorous. Yet it lists no fewer than 27 factors that the board may use to determine pay -- everything from cash flow to "service level." When you pick from 27 factors, you can get any result you want.
Give Coke some credit: Kent's non-stock bonus is sensitive to performance, and in 2013, a disappointing year, it was cut by almost two-thirds. However, in another sorry hallmark of American compensation plans, Kent's total package is an aggregate of five distinct pay categories. One bracket will rise as another is cut. In Kent's case, though the bonus fell, three other categories increased. The total package fell only 16 percent, to $18.2 million.
In other words, in a year in which Coke's performance, by the company's reckoning, failed to meet some of its targets, Kent earned enough to make him one of the richest people on the planet. Now, the company has authorization to pay him even more.
This column does not necessarily reflect the opinion of Bloomberg View's editorial board or Bloomberg LP, its owners and investors.
(Roger Lowenstein is writing a book on the origins of the Federal Reserve System.)