On executive pay, incentives have limits

01 November 2017

Boards will always need to apply judgment and a sense of proportion.

What can a person do that is worth a third of a billion dollars? A retirement announcement from Kenneth Chenault, chief executive of American Express, prompts the question. This month he said he would leave the company next year, after 17 years as chief. His compensation as CEO to date runs to $370m in shares and cash, an average of $22m a year.

The sum is all the more stunning in that it is not unusual for someone running a big company (American Express’s market capitalisation is over $80bn). The average S&P 500 CEO earned more than $10m in 2016. These vast sums raise questions of fairness. Is it possible that one person in a very large organisation can generate that much value? Can we tell if they do?

American Express is a private company. What it pays its employees is for its board to decide and its shareholders to approve. Regulating upper limits on pay is a poor fit with a liberal society. But as with any price, it is worth asking whether companies, and the wider society, are getting value.

The best argument in favour of high pay is that even a small movement in the value of a large company can be very significant. If the head of a $100bn company increases its value by 1 per cent a year, then she is creating $1bn in value. At $10m a year, she is a bargain.

The argument has two problems. First, it is hard to see where the argument stops. Apple’s value is approaching $1tn. Should Tim Cook, then, be paid $10bn a year? If not, why not? (he took home $9m last year). In a world increasingly dominated by huge companies, the argument from corporate size justifies sums that would make the steeliest capitalist blanch.

Second, it is very hard to isolate the contribution of CEO skill when setting pay. Links to share performance means CEO pay will depend on market cycles. Benchmarking stock performance against peers helps, but only within limits. American Express shares, for example, have handily outperformed the broader financial sector over Mr Chenault’s tenure. But that is largely down to the fact that it is not a bank (bank shares were hammered in the crisis). American Express shares have badly underperformed those of the credit card networks MasterCard and Visa. But its business model is very different from theirs. Benchmarking will always leave room for judgment.

Using profit metrics instead of share performance, similarly, exposes CEO pay to the whims of input prices, exchange rates and technological advancements.

A final problem is that - again because at large companies even huge pay packages are small relative to total company value - shareholders have a limited incentive to push back on pay. Board members, too, have little reason make a fuss about a high-pay culture that has also served them well.

None of this is to argue that CEO skill is not important, or that it is completely invisible. Nor is it to argue that pay is totally arbitrary. The components of sensible pay structure are increasingly well understood: pay the boss largely in stock that must be held for many years. De-emphasise options and short term performance incentives. Do all of this transparently. This should increase fairness and improve corporate performance. But it does not ensure that CEO pay fully corresponds to value creation.

The lesson is that while proper incentives and full transparency are important, an element of judgment and a sense of proportion will always be required on the part of boards. There is such a thing as too much. Rules, guidelines and procedures cannot absolve us from recognising it when we see it.

Copyright The Financial Times Limited 2017



(c) 2017 The Financial Times Limited