BETA
This is a BETA experience. You may opt-out by clicking here

More From Forbes

Edit Story

The Highest-Paid CEOs Are The Worst Performers, New Study Says

This article is more than 9 years old.

Across the board, the more CEOs get paid, the worse their companies do over the next three years, according to extensive new research. This is true whether they’re CEOs at the highest end of the pay spectrum or the lowest. “The more CEOs are paid, the worse the firm does over the next three years, as far as stock performance and even accounting performance,” says one of the authors of the study, Michael Cooper of the University of Utah’s David Eccles School of Business.

The conventional wisdom among executive pay consultants, boards of directors and investors is that CEOs make the best decisions for their companies when they have the most skin in the game. That’s why big chunks of the compensation packages for the highest-paid CEOs come in the form of stock and stock options. Case in point: The world’s top-earning CEO, Oracle billionaire Larry Ellison, took in $77 million worth of stock-based compensation last year, according to The New York Times, after refusing his performance bonus and accepting only $1 in salary (he made a stunning total of $96 million in 2012). But does all that stock motivate Ellison to make the best calls for his company?

The empirical evidence before fell on both sides of that question, but those studies used small sample sizes. Now Cooper and two professors, one at Purdue and the other at the University of Cambridge, have studied a large data set of the 1,500 companies with the biggest market caps, supplied by a firm called Execucomp. They also looked at pay and company performance in three-year periods over a relatively long time span, from 1994-2013, and compared what are known as firms’ “abnormal” performance, meaning a company’s revenues and profits as compared with like companies in their fields. They were startled to find that the more CEOs got paid, the worse their companies did.

Another counter-intuitive conclusion: The negative effect was most pronounced in the 150 firms with the highest-paid CEOs. The finding is especially surprising given the widespread notion that it’s worth it to pay a premium to superstar CEOs like Jamie Dimon of JPMorgan Chase (who earned $20 million in 2013) or Lloyd Blankfein ($28 million) of Goldman Sachs. (The study doesn’t reveal individual results for them.) Though Cooper concedes that there could be exceptions at specific companies (the study didn’t measure individual firms), the study shows that as a group, the companies run by the CEOS who were paid at the top 10% of the scale, had the worst performance. How much worse? The firms returned 10% less to their shareholders than did their industry peers. The study also clearly shows that at the high end, the more CEOs were paid, the worse their companies did; it looked at the very top, the 5% of CEOs who were the highest paid, and found that their companies did 15% worse, on average, than their peers.

How could this be? In a word, overconfidence. CEOs who get paid huge amounts tend to think less critically about their decisions. “They ignore dis-confirming information and just think that they’re right,” says Cooper. That tends to result in over-investing—investing too much and investing in bad projects that don’t yield positive returns for investors.” The researchers found that 13% of the 150 CEOs at the bottom of the list had done mergers over the past year and the average return from the mergers was negative .51%. Among the top-paid CEOs, 19% did mergers and those deals resulted in a negative performance of 1.38% over the following three years. “The returns are almost three times lower for the high-paying firms than the low-paying firms,” says Cooper. “This wasteful spending destroys shareholder value.”

The paper also found that the longer CEOs were at the helm, the more pronounced was their firms’ poor performance. Cooper says this is because those CEOs are able to appoint more allies to their boards, and those board members are likely to go along with the bosses’ bad decisions. “For the high-pay CEOs, with high overconfidence and high tenure, the effects are just crazy,” he says. They return 22% worse in shareholder value over three years as compared to their peers.

Yet another surprising finding: The high-paid CEOs did poorly for themselves when it came to cashing in their options. Among the bottom-paying firms, 33% of the CEOs held onto their options when they could have cashed them in for a  profit, which the paper calls “unexercised in-the-money options,” while more than twice as many high-paid CEOs, 88%, held onto their options when they could have made money selling.

What can be done about all those negative numbers? The paper doesn’t venture to say but Cooper notes that some finance experts have suggested so-called claw-back provisions. In a CEO pay contract, there would be an item that says, if the firm does poorly compared to its peers, the CEO loses a share of his compensation. “That proposal hasn’t gone over real well,” says Cooper. “There is another school of thought, that CEOs are just too highly paid, period,” he adds. “The U.S. is pretty egregious as far as the ratio between median pay and what the CEO makes.”

Though four years ago the Dodd-Frank law instituted a requirement that firms divulge the ratio between CEOs and median pay, the SEC has yet to issue a final rule ordering it, and companies have been less than forthcoming. But Bloomberg compiled data last year showing that the average multiple of CEO compensation to that of rank-and-file workers was 204, up 20% since 2009. At General Electric, with its star CEO Jeffrey Immelt ($28.2 million in 2013), the ratio was 491, according to Bloomberg.

The Occupy movement, labor unions and some members of Congress have pushed companies to divulge more information about pay ratios, and complained about excess CEO pay, while boards have pushed so-called say-on-pay provisions that would allow them to vote on executive compensation packages. Now those groups have some new empirical evidence to support their positions.