Warren Buffett has repeatedly used his “letter” to Berkshire Hathaway’s shareholders to complain about pay. The “boardroom atmosphere almost invariably sedates [directors’] fiduciary genes,” he observed on one occasion. “Collegiality trumps independence.” In 2003, with the scandals of WorldCom and Enron still smouldering, the great investor issued a challenge to directors across the country. “In judging whether corporate America is serious about reforming itself, CEO pay remains the acid test,” he wrote. “To date, the results aren’t encouraging.”
One sign that something is amiss comes from studies that seem to point to a failure of governance. Take research by Marianne Bertrand of the University of Chicago and Sendhil Mullainathan, a Harvard economist. Reasoning that executives should be paid for their own accomplishments but not for sheer good fortune, they investigated whether chief executives were in fact paid for luck.
In the oil industry they found that chief executives’ pay always benefits when the oil price is high, but does not necessarily suffer correspondingly when the price is low. Across a large collection of companies, they looked at the effect of changes over which managers have no control, such as shifts in exchange rates. They found that the typical firm rewards its chief executive as much for luck as it does for good performance. The effect cannot be explained simply by the increase in the value of managers’ share options: it also shows up in their base salaries and bonuses, which are directly controlled by boards. Revealingly, the rewards for luck were smaller in companies that were judged to be well governed.
The most compelling evidence that governance matters comes from putting aside the abstractions of economists and examining what really happens inside companies. The way that new chief executives are hired, studied in forensic detail by Rakesh Khurana, a professor at Harvard Business School, is especially flawed.
Can You Keep a Secret?
For a start, information flows poorly, because both buyer and seller are at high risk from leaks. The executive does not want to let his employer know that he has been talking to another company — probably a competitor — until the deal is done. The hiring company is scarcely less comfortable. It does not want the identities of its favoured candidates to be known because it might end up hiring someone who is obviously the second or third choice.
That is what happened to John Walter as head of AT&T in the 1990s. The unfortunate Mr Walter was clearly the runner-up to Michael Armstrong, who had turned the job down. On news of the appointment, analysts spoke of “a mistake”. CNN, Mr Khurana recounts, called Mr Walter “a puzzling choice”. AT&T’s stockmarket value fell by $4 billion. Mr Walter lasted less than nine months, and when Mr Armstrong did eventually take over there was general euphoria — until he proved no more able to restore the company’s fortunes than Mr Walter.
As that cautionary tale shows, the choice must satisfy not only the board itself but also stockmarket analysts and the financial media. The need for external approval is all the greater if the company is in the throes of some crisis, as it so often is when the board feels compelled to go outside for a new chief. Imagine the circumstances: the previous boss has left in disgrace, the share price is falling, commentators are dragging the company through the mud, the employees are anxious and disillusioned. The board feels compelled to act swiftly.
Hence the list of potential candidates is a short one. Some three-quarters of chief executives appointed from outside between 1985 and 2000 were already chief executives or presidents elsewhere. That sort of background helps to establish their credentials, but the companies they come from also have to be seen by analysts and by the press as successful (even though nobody knows precisely what part the bosses played in that success).
The whole process is carefully choreographed by the headhunters, who act as matchmakers between the two sides. By the time the board and the candidate meet — and they may meet only once — the union is all but decided. With billions of dollars of market value at stake, the board may be inclined to overlook a few million dollars of pay. “The nightmare scenario for the board at this point is having to restart the process,” one candidate told Mr Khurana. “So you ask. And you get.”
Which is why outside candidates are offered the sun and the moon — and why they demand that the stars be thrown in as well. They are probably leaving behind a valuable options package that will have to be “made whole”. The chances are that they have swapped a secure future for a risky (but potentially very remunerative) roll of the dice at the new company.
Understandably, before they sign the contract, new executives bargain to make their bonuses and options more secure. This is the moment when the board caves in and offers the most offensive ingredients of executive pay. The oxymoron of the “entrepreneurial manager” is shown up by safety devices such as guaranteed bonuses, severance pay, hugely generous provisions if the company changes hands, perks, accelerated vesting of options and massive pensions contributions. But boards may think that is the price of getting their man or woman.
“Once they have employed the guy, they want to police the CEO,” says Jay Lorsch of Harvard Business School. “But they have given away the store. It is worked out at the beginning. Having made the deal, how do you contain this thing?”
The board’s inability to stand up to the incoming chief executive is an example of a more general spinelessness documented by Lucian Bebchuk and Jesse Fried, of Harvard Law School and the University of California at Berkeley. Boards are agents, too, and Messrs Bebchuk and Fried believe that their interests are more closely aligned with those of powerful executives than with those of the owners they are supposed to represent. As a result, the bargaining over pay is not at arm’s length and boards conspire with executives by providing all sorts of “stealth pay” that disguises the true extent of their rewards.
The director’s role is ambiguous, because he is charged with counselling the chief executive as well as monitoring management’s performance. The duty to offer friendly advice inevitably interferes with the duty to admonish and if necessary to sack. The conflict has intensified as boards have been loaded up with responsibilities. As Richard Schmalensee, dean of the MIT’s Sloan School, puts it: “You ask hard questions to prevent dumb decisions. If you pretend to be an auditor without the time to be one, who is going to listen to your advice?”
Messrs Bebchuk and Fried argue that the chief executive can co-opt the directors in all sorts of ways. Although executives no longer decide who is nominated to become, or remain, a director, they still have a large influence over the pay and composition of the board. Business dealings between executives and directors should be declared and must not be large, but executives can determine whether the company gives money to a director’s favoured charity, and they control special payments. In addition, many directors were once chief executives themselves and are likely to think that others in that job deserve to be well paid. Boards face some constraints from investors and predators, but as long as they do not attract notoriety they have plenty of leeway.
To support their arguments, the two academics looked for evidence that weak boards pay powerful chief executives more. They found, for instance, that high pay is associated with large boards (which are easier for the chief executive to dominate), companies that have no outside shareholder to monitor them, and companies protected by anti-takeover provisions. The researchers highlight some cases in which the board’s behaviour seems to have run counter to shareholders’ interests. When Jill Barad left Mattel’s top job because the toymaker was performing poorly, she was forgiven a $4.2m loan, handed $3.3m to pay the tax on the forgiveness of another loan and had her options vested. None of that was in her contract, which did, however, provide for a payment of $26.4m, which was duly made. Perhaps Ms Barad’s allies on the board were willing to see her go only if she was showered with money. But would the directors have been so generous if they had been paying it out of their own pockets?
The great irony is that the most imaginative scheme to mitigate the agency costs of the public company ended up creating agency costs of its own. Options were supposed to make managers think like owners. But as Michael Jensen and Kevin Murphy explain, in an overvalued stockmarket options can make managers more short-termist than ever. Bob Monks, a veteran corporate-governance reformer, calls them the “enablers of excess”.
To see why, think of a manager who owns options in a company that is heavily overvalued. He knows that the company will not make enough money to justify its valuation. Yet many people’s prosperity — including the manager’s own — rests on the share price remaining comfortably above the options strike price. The temptation is for the manager to use his inside knowledge to buoy up the share price in the short run. He can make acquisitions, delay investments or manage earnings using questionable accounting in the hope that some unexpected good news will turn up or that he will win enough time to make money from his options. In one survey, well over half of the chief financial officers being questioned said that they would cut a value-creating investment rather than miss an earnings target set by the market. And sometimes managers will resort to earnings manipulation and fraud. Anyone who lost money in Enron knows how that can end.
To make matters worse, options were often granted on over-generous terms. Consider an example from a Fortune 500 company described by Messrs Jensen and Murphy. When the company’s share price was $57, managers were granted options “at the money”, that is with a strike price also of $57. They told the board that they expected the share price to rise to $100 in five years. The company’s cost of capital was agreed to be 15%, and a dividend of 2.5% of the share price was being paid. The shareholders would therefore need a share-price rise of at least 12.5% a year if they were to cover their cost of capital. On that arithmetic, the share would have to rise to $102.72 over five years. If the markets believed the managers, today’s share price would be too high and investors would sell. Messrs Jensen and Murphy calculate that the share price would fall by about $1.50. On the other hand, if the managers achieved their modest $100-a-share target they would be $43 a share better off.
In the bull market boards tended to hand out set numbers of options each year, regardless of the share price. That seems odd, because if the firm had doubled in value in a year the options grant would automatically be worth twice as much. Executives were often able to sell shares as soon as the options had been exercised. That was good for the executives, who could diversify their investments. But it was bad for the companies, because the executives had less incentive to behave like shareholders.
At the peak of the bull market, the boards of some new-economy companies were giving away more than 7% of the equity a year, heavily diluting the original shareholdings. Most pay experts think that board directors mistakenly believed the options they were giving away were free, which is economic nonsense.
In sum, the market for chief executives is flawed and boards find it hard to stand up to executives. Options were handed out too easily and, in the internet bubble, encouraged short-term behaviour. There has been plenty of misgovernance.
Clearly all this had an effect on pay. The question is, how much?