PsyFi Search

Monday 30 June 2014

CEOs: The More You Pay Them, The Worse They Perform

The Peter Principle
"Every new member in a hierarchical organization climbs the hierarchy until he/she reaches his/her level of maximum incompetence"
The Peter Principle states that everyone gets promoted to a level at which they're incompetent. More generally Lawrence Peter observed that anything that works gets used to do other things until it fails. This is true even of ideas, and sometimes even failure doesn't stop them being promulgated. Consider, for instance, economics as the primary example of an idea stretched beyond its elastic limit.

But at the top of the corporate tree are those executives who've somehow avoided becoming victims of the Peter Principle. Of course, simply becoming a CEO doesn't disprove the idea, it just makes it more likely that we'll have lots of incompetent business leaders who've become really good at blaming other people. And a survey of CEOs suggests that this is exactly true: they're overrated, they're overpaid and they don't deliver for anyone apart from themselves. Why am I not surprised?

Promoting Incompetence

The Peter Principle elucidated an idea that anyone who's ever worked in a hierarchical organization instantaneously recognises: people get promoted because they're very good at what they do, but the skills they needed in their old role are often completely unrelated to their new one, so it's entirely random as to whether they're any good at the next level. As they ascend the corporate ladder eventually they get to a position that they have no competence for whatsoever. And, at that point, because they're incompetent they can't get promoted but no one ever gets demoted, so there they stick until or unless they're unexpectedly offered the opportunity to pursue their careers elsewhere.

Under the Peter Principle the middle managements of organizations are staffed with myriads of people who aren't very good at their jobs, and a few people who will continue to rise. Finally, of course, if you're good enough and lucky enough you make it to the top, attaining the heady heights of CEO. On the Peter Principle, lots of people promoted to CEO are, once again, reaching beyond their abilities but by this stage, after a fulfilling career of undiluted success, it's hardly likely that they're going to accept that.

Managing Upwards

In fact, anyone who's really worked in a hierarchical organization will be slightly sceptical of the Peter Principle. I know lots of incompetent people who've got promoted. In fact I once worked with someone who was so incompetent my manager promoted him to get rid of him. Eighteen months later, he was back as my manager's manager. And some people are simply gifted at managing upwards and appearing to do a good job, while in reality they're parasites on the efforts of other people. But arguably, those are the skills of a top, top manager ...

If this combination of the Peter Principle and networking skills does lead to the rise of incompetent people to the leadership of our corporations we should expect to see this in the only way that really matters - financial performance. Hypothetically you'd expect poor CEOs to demand more money and produce less performance (while, of course, demanding exactly the opposite of their employees - power begets hypocrisy, as we saw in Born Rich, Born Greedy). Which is exactly what Michael Cooper, Huseyin Gulen and Raghavendra Rau found in their paper Performance For Pay? The relation between CEO incentive compensation and future stock price performance.

Aligned Like Hell

In Incentives Gone Bad I looked at a famous paper from 1976 on executive compensation by Michael Jensen and William Meckling that argued that corporate executives needed to be healthily remunerated, especially with future based rewards like stock options, in order to align their interests with those of their shareholders. The idea, of course, was that this would ensure that executives were fully committed to increasing profits and revenues. In fact what it did was get them focused on pumping up the share price, often by dubious and sometimes illegal means.

(Actually, what I find really interesting is that corporate executives never really pay much attention to this type of research. But they did to this paper. Funny that).

The use of stock options has recently come under scrutiny because it rewards executives for perverse behavior to push up stock prices in the short term, rather than focusing on longer-term compensation. Given this the general trend is to look at alternative compensation plans with longer term components, but while these avoid some of the issues with share options they have more similarities than differences.

Chief Executive Optimists

These new plans, of course, still assume that the old meme of alignment holds true. They expect that if you give a CEO more compensation they'll give more performance. The paper from Cooper and colleagues suggest that this may be overoptimistic:
"We find evidence that CEO pay is negatively related to future stock trends for periods up to three years after sorting on pay".
Yes, you read that correctly. The best paid CEOs deliver negative performance against market norms. Presumably having been aligned they can't help but spend their time figuring out how to spend their inflated compensation packages.

The research further suggests a reason for this problem, and it's one we've seen before. Unsurprisingly, the Peter Principle has delivered unto us a bunch of  CEOs who are supremely confident in their abilities. Supremely overconfident, to be precise:
"We show that highly paid CEOs exhibit firm investment and personal portfolio choice behavior that is consistent with being overconfident and that firms with the highest paid and most overconfident CEOs earn lower future returns relative to other CEOs".
Positive Incentives, Negative Returns

The highest paid CEOs are the most overconfident and they demonstrate this bias by engaging more frequently in value destroying actions - things like mergers - and see their firms produce "significant negative abnormal returns". And this performance continues to deteriorate over time. But interestingly it doesn't seem to be normal pay that has this effect - it's the size of the future component of remuneration in whatever form this may come. Indeed, the more you actually pay a CEO in real money in real time the better the performance of their corporations.

Which, on the face of it, suggests that this isn't the Peter Principle at all, but a simple repudiation of the Jensen-Meckling idea of aligning executive rewards via long-term compensation. However, there's a sneaking suspicion that the CEOs who get rewarded like this are the ones that game the system most effectively. It may be that it's the incentives that make the CEOs behave badly but, given that the best paid ones appear to be the most overconfident, it's at least as likely that it's the terrible CEOs that seek the incentives. 

Random Stars

Overcoming the Peter Principle is, for obvious reasons, quite difficult. Promoting the best people meets peoples' expectation about fairness and reciprocity but it may be that this isn't the most effective approach. In The Peter Principle Revisited the researchers analysed alternative ways of deciding how to promote people and discovered that, statistically speaking, the best approaches were to either promote someone at random or to randomly promote the best or worst team members.

So, there you have it: if you want to avoid the worst sort of expensive, value destroying CEOs, you'd probably be better off promoting people at random. Somehow I don't think it's going to happen, but as investors we should avoid investing in corporations that have star CEOs with huge future pay packets. For every shooting star that shines brightly there are dozens that are burning up shareholder value for no one's benefit but their own.

The Peter Principle added to The Big List of Behavioral Biases.


  1. Mr Smithers thinks this creates other effects as well

  2. Too many CEOs are stupid when it comes to getting returns for their investors. They treat the company as if it were their own personal piggy bank, looting it dry. Then they lay off people as "cost cutting" measures but forget their pay and benefits could be cut too (nope, aint' gonna happen)