Martin’s speech was basically a summary of the key ideas from his new book, Fixing the Game: Bubbles, Crashes, and What Capitalism Can Learn from the NFL. (I mentioned Martin’s book a few weeks ago, in a blog posting called Business, Football, and Incentives.)
Here is a rough summary of what he had to say, paraphrased and condensed:
Prior to the mid-70′s, stock-based compensation for CEOs was rare. But starting especially in the 80′s, it became very common indeed. Martin traces the sea change to a famous paper by Michael Jensen and William Mecklin, called “Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure” (PDF here). The basic idea at the time was that paying senior executives, and especially CEO’s, in company stock or stock options would align their interests with those of shareholders. Shareholders naturally want the value of stock to rise, and paying CEOs mostly in stock gave them a very concrete reason to want stock to rise, too.
It was a fine theory, says Martin, but it didn’t work out well. If you compare the era of stock-based compensation to an equivalent period before, you see that returns went down about 15% and stock volatility went up about 15%. Those definitely aren’t the kinds of results that shareholders were looking for.
And yet somehow people still cleave to the idea that stock-based compensation aligns interests. Why?
It’s clear enough why CEOs themselves are fans of the system. The reason, according to Martin, is rooted in the fact that stock prices only reflect the market’s collective expectations about a company’s future performance. That means in order to boost stock prices (and hence their own compensation) CEOs merely need to boost expectations. So, says Martin, that’s what CEOs have learned to do: manage stock analysts’ expectation, rather than managing actual performance. If analyst expectations are low when stock options are granted, and high when they get cashed out, a CEO stands to make a lot of money, independent of what that variation means in terms of actual performance.
But of course, says Martin, CEOs have realized that you can’t play that game for very long. So, they learned to look for opportunities to play a hit-and-run version of the game: get in, play hard, and cash out. That, he says, is the real reason why the average tenure of CEO is so short these days.
Is this malfeasance on the part of CEOs? Not really, says Martin. It’s just CEOs doing what they are payed—incentivized—to do.
Now, says Martin, compare this situation to the way quarterbacks are payed in professional football. Professional quarterbacks, he says, are paid for real, on-the-field performance. Additionally—and this is crucial—they are forbidden from profiting from outsiders’ expectations of how they will perform, i.e., from gambling on the outcome of the games they are playing in. Why? Because professional football leagues realize that letting quarterbacks gamble would give them all kinds of perverse incentives. The corporate world, it seems, has something important to learn from the world of pro football when it comes to incentivizing key personnel.
In the corporate world, says Martin, the only ones with something to gain from having stock-based executive compensation are CEOs and hedge funds. Both, he says, benefit from volatility of stock prices.
Martin’s prescription: performance-based compensation is fine. But don’t reward CEOs based on stock prices. Reward them based on real performance, in terms of something like earnings or sales or market share—different systems will make sense for different companies with different strategic objectives. But the point is to reward them for something more real than merely meeting the expectations of analysts.
It’s a provocative thesis, and a bold prescription. To say that stock-based compensation is common practice is an enormous understatement. And Martin acknowledges that change, if it comes at all, will not come quickly. But given how widely-agreed-upon it is that current modes of compensation are not working, bold prescriptions may just be what is in order.