Friday, October 20, 2006

Mixed Reviews for Krugman Today

In "Incentives for the Dead" (reposted by Mark Thoma for those of you without TimesSelect), Paul Krugman addresses the social and economic aspects of option backdating and repricing. He gets some things right--this is accounting fraud, it is also tax evasion, and the perpetrators should be prosecuted fully. But he gets some of the economic aspects wrong. Consider:

To understand the issue, we need to go back to the original ideological justification for giant executive paychecks.

In the 1960’s and 1970’s, C.E.O.’s of the largest firms were paid, on average, about 40 times as much as the average worker. But executives wanted more — and professors at business schools provided a theory that justified much higher pay.

They argued that a chief executive who expects to receive the same salary if his company is highly profitable that he will receive if it just muddles along won’t be willing to take risks and make hard decisions. “Corporate America,” declared an influential 1990 article by Michael Jensen of the Harvard Business School and Kevin Murphy of the University of Southern California, “pays its most important leaders like bureaucrats. Is it any wonder then that so many C.E.O.’s act like bureaucrats?”

Jensen and Murphy's article does not provide a theory to justify "much higher pay" or "gigantic executive paychecks." The article provides a theory to justify higher pay-performance sensitivities, for any level of executive pay. Apart from the (relatively minor) increase in expected compensation that must compensate the added risk exposure imposed by the pay-performance sensitivity, there is no justification presented for higher levels of pay unless the incentives work, the value of the firm rises, and the CEO's compensation rises via that pay-performance sensitivity. And, obviously, nothing in Jensen and Murphy's paper justifies backdating or repricing--quite the contrary.

Krugman continues:
The claim, then, was that executives had to be given more of a stake in their companies’ success. And so corporate boards began giving C.E.O.’s lots of stock options — the right to purchase a share of the company’s stocks at a fixed price, usually the market price on the day the option was issued. If the stock went up, these options would pay off; if the stock went down, they would lose their value. And so, the theory went, executives would have the incentive to do whatever it took to push the stock price up.

In the 1990’s, executive stock options proliferated — and executive pay soared, rising to 367 times the average worker’s pay by the early years of this decade.

But the truth was that in many — perhaps most — cases, executive pay still had little to do with performance. For one thing, the great bull market of the 1990’s meant that even companies that didn’t do especially well saw their stock prices rise.

The last sentence continues to be something of a mystery in the economics literature--why don't we see more relative performance evaluation? Fair enough. I've done some research on this myself, so perhaps I'll blog about it at some later time.

But the statement just prior to it is extremely misleading, if not simply wrong. If options were being used in this way, then they should have been used instead of cash compensation, not in addition to cash compensation. Furthermore, the data suggest a more limited role for options in generating large pay-performance sensitivities (what Krugman calls a "stake" in the company).
Among CEOs, the median incentives received from options are roughly the same as the median incentives received from direct holdings of stock. However, the mean incentives through stock are much higher than the mean incentives from options, and almost every instance of truly high incentives that I have observed has come from stock, not from options. (The source for these comparisons is the Standard & Poor's ExecuComp data, tabulations of which are in my August 2003 paper in the Journal of Finance.) For example, Bill Gates got rich as the CEO of Microsoft, but not due to options. He's rich because of his original stock holdings and what happened to their value as the company flourished.

So I think Krugman is overstating the link between Jensen and Murphy's paper--or any economic theory--and the particular way that options have been used in compensation contracts. My reading of the data is that the grants haven't been large enough--even if they were not repriced or backdated--to have moved CEO compensation away from the "bureaucratic" model. (Although see the article by Brian Hall and Jeff Liebman in the Quarterly Journal of Economics for a different assessment.)

So my revision of Krugman's argument, at least as it pertains to the economics, is not that there was a bait-and-switch, but that there was too little bait on the hook to begin with. Why did this happen? In my view, the explanation is simply that CEOs have been pushing their boards of directors to grant them pay increases, the tax code and accounting regulations have favored options as the easiest and most advantageous way to do that, and corporate boards have acquiesced.

Ultimately, all problems of corporate governance derive from inadequate monitoring of the managers by the shareholders. Corporate boards are supposed to do this. Some are obviously failing.


Anonymous said...

Krugman seems correct when you take salary + bonus + deferred comp + stock-option-gain together as "Executive pay".

The thing I think Klugman was dodging was the fact that Clinton is to blame for a lot of this. An unintended consequence to be sure but...

In response to Executive Pay That Was TOO High...

Options did grow in popularity especially after the 1993 Clinton legislation limiting the deductibility of top executive salary when such salary exceeded $1 million and was not "performance based".

The stock options were thought to be a cool answer because they "aligned" investor interests with executive interests and would only bear fruit if the company did well.

Of course the options offered only upside potential to the executives while shareholders could lose their investment.

At that time granting options had no effect on company P&L statements or the Balance Sheet.

Backdating of option grants - usually no farther back than the start of the current month but sometimes to any time in the current year (fiscal or calendar).

Structuring options to set below-market strike prices. This was done by inceaseing the "risk" that the option could ever be exercised.

There was also a Qualified Option which the holder could exercise but not owe tax UNTIL the stock thus purchased was sold. (I think these last 2 particular flavors of option were stomped out by the year 1998).

And of course stock options increased the incentive to cook the books to keep the stock price growing.

bncthor said...

Stock option schemes as a means for encouraging CEO and employee performance are a bad idea. Reasonable cash bonuses for corporate managers and employees would accomplish the same goal - improved management and employee performance - and not leave the door to the corporate cash stash open to abuse. And, no problem of how to treat such bonuses for tax purposes would ensue. Unearned rewards, like temptation, must always to be guarded against.

Barry said...

One important, and highly overlooked reason, is the structure of CEO's/Chairman/Boards of Directors. The stockholders get to vote on slates of directors offered by the Chairman and current Board of Directors. A plurality of shares voted wins, so apathy helps the candidates offered by the current Board.

Individual stockholders have 0 power; even if they got together with many others, they'd still have 0 power. A coalition holding a majority of shares would still face obstacles such as only one director position up for renewal each year, making changing a majority of the Board a multi-year project. Institutional funds, who hold large chunks of stocks, face similar problems, plus conflicts of interest - many of them would like to run investments (401K's, etc.) for the companies whose shares they hold. With Enron, we saw pretty much a 100% failure of supposedly 'neutral' parties, because they were all co-optable.

Finally, information. The Chairman and Board have the lists of shareholders, and can propagandize them at company expense. Everybody else has to persuade voters on their own nickel.