Why does it seem that it's always Christmas in corporate boardrooms? And how can investors tell whether those glittering pay packages are worth the cost?

The answer sounds obvious: Pay the boss more for good results now, and you should get even better results later. But the evidence for that is surprisingly weak, and two new studies even suggest that when chief executive officers get paid more, shareholders end up earning less.

The first study, led by corporate-governance expert Lucian Bebchuk of Harvard Law School, looked at more than 2,000 companies to see what share of the total compensation earned by the top five executives went to the CEO. The researchers call this number—which averages about 35%—the "CEO pay slice."

It turns out that the bigger the CEO's slice of the pie, the lower the company's future profitability and market valuation. "These CEOs," says Prof. Bebchuk, "seem to be trying to grab more than they should."

Finance professor Raghavendra Rau of Purdue University and two colleagues looked at CEO pay and stock returns for roughly 1,500 companies per year from 1994 through 2006. They found that the 10% of firms with the highest-paid CEOs produce stock returns that lag their industry peers by more than 12 percentage points, cumulatively, over the next five years.

Companies at the top of the pay pile, Prof. Rau concluded, award their CEOs an annual average of $23 million—but leave their shareholders poorer (relative to other companies in the same industry) by an average of $2.4 billion per year. Each dollar that goes into the CEO's pocket takes $100 out of shareholders' pockets.

Prof. Rau's team is still validating these findings, so regard them as provisional. But both studies are a reminder of a profound point: There may be no way to prove that paying CEOs more money leaves outside shareholders better off.

"Compensation changes so quickly, in so many ways, that it's one of the least stable variables in finance," says David Yermack, a finance professor at New York University. "And stock prices gyrate for reasons no one understands. These are such noisy variables that it's very hard to determine direct causation. You really need to know what you don't know."

Given the vast array of factors that affect business results and stock prices, it is extraordinarily difficult for the directors on a compensation committee to know whether the past results they are rewarding came from luck or skill. Nor can they reliably tell how much of the company's future performance should be attributed to those rewards. The truth, says Prof. Yermack, is unknown and may never be knowable.

Further muddying the waters: Many CEOs expect to be bought out of their interests in their former company when they are recruited to join another firm, as Bank of New York Mellon Corp. CEO Robert Kelly recently sought when he was a candidate for the top spot at Bank of America.

Mr. Kelly ultimately decided to stay at BNY Mellon. But when a CEO does get bought out in order to jump to a new ship, that "creates an expectation or baseline for future pay that's uncorrelated to the success of the new firm," says Harvard Business School management professor Rakesh Khurana.

Years ago, the great investor Benjamin Graham pointed out that directors shouldn't merely be independent, but also "businesslike." They must have an arm's-length relationship with management; they also should combine "good character and general business ability" with "substantial stock ownership." (They should have purchased most of their shares outright rather than getting them through option grants.)

The independent directors, Graham believed, should publish a separate annual report analyzing whether the business is "showing the results for the outside stockholder which could be expected of it under proper management." Executive pay should be part of that analysis.

The annual proxy statements that report on pay still fall short of what Graham advocated in 1951: a kind of interrogation in which directors are "called upon to justify…the generous treatment they are asking the stockholders to approve." Added Graham: "The stockholders are entitled to be told…just what are the excellent results for which these arrangements constitute a reward [and] by what analogies or other reasoning the board determined that the amounts accorded are appropriate."

Surely the financial crisis should have taught us all that we must acknowledge the extent of our ignorance. What's urgently called for in the CEO pay process, says Prof. Yermack, is "a great deal of humility." It's high time for corporate compensation committees—and investors—to start doubting whether the lavish pay packages they endorse actually work.

Write to Jason Zweig at intelligentinvestor@wsj.com

Copyright 2009 Dow Jones & Company, Inc. All Rights Reserved

This copy is for your personal, non-commercial use only. Distribution and use of this material are governed by our Subscriber Agreement and by copyright law. For non-personal use or to order multiple copies, please contact Dow Jones Reprints at 1-800-843-0008 or visit


More In Investing

About Jason Zweig

Jason Zweig writes The Intelligent Investor every Saturday for The Wall Street Journal. He is the author of Your Money and Your Brain, on the neuroscience of investing, and the editor of the revised edition of Benjamin Graham's The Intelligent Investor, the classic text that Warren Buffett has described as "by far the best book about investing ever written." Before joining the Journal, Jason was a senior writer for Money magazine and a guest columnist for Time magazine and CNN.com, and he also spent a year studying Middle Eastern history and culture at the Hebrew University in Jerusalem.