What's a CEO worth?

Marc Hodak tries to use Mark Hurd's resignation as a way of estimating how much value Hurd contributed to HP:

Lost in the fog of this debate is the presumptive basis of pay in a market economy, i.e., that what one gets paid by a company should roughly reflect what one is worth to the company. So, a relevant question in this debate that is never asked: What is a CEO worth?

Mark Hurd’s sudden, surprise resignation at HP offers a rare glimpse into the answer. In the minutes after the announcement of his firing, HP’s stock declined by over 8 percent in after-hours trading.

Ladies and gentlemen, that’s over $9 billion dollars in market cap.

Keep in mind that the stock market did not assume that HP would go without any CEO at all following Mr. Hurd’s departure; that value decline is the market’s assessment of HP with Mark Hurd as CEO versus someone else as CEO. ...

In other words, in the apolitical judgment of equity investors, the only people with the incentive to make this collective judgment correctly, the company would have been better off paying about $2 billion a year for the next five or six years to keep Mr. Hurd than to lose him.

I agree that trying to figure out what a CEO is worth is an important element of the executive compensation debate. I'm not convinced that the Hurd example makes a good case study. There are too many confounding issues:

• Maybe the incident convinced investors that Mark Hurd is dumber than they thought (this seems to be where Tom Smith came out)
• Maybe investors are concerned that HP's internal controls took so long to find Hurd's misconduct.
• Maybe investors are concerned that HP's board is doing a good job of overseeing management.
• Maybe investors think the bad publicity will hurt HP's business prospects.
• Maybe investors worry that when a new CEO is chosen passed over internal candidates will quit in a huff.
• Maybe investors are very easily spooked these days.

What you need to really get at this issue is a cleaner event with less confounding factors. Like an unexpected death. Of course, even death would have its confounding issues. If a highly entrenched CEO of an under-performing firm died unexpectedly, one might expect the stock price to go up as investors reason that anybody would be better than the late boss.

It's just really hard to disentangle all the variables, let alone control for them.

Update: Larry Ribstein runs a different set of numbers to (a) raise questions about Hurd's tenure and (b) take a slap at SOX and federal criminalization of agency costs.

Comments

[MHodak]

I certainly agree that there are too many variables to treat any single stock drop as an event study. As I mentioned in the piece you kindly quoted, the investors got new information about Hurd and their company with this disclosure. But it's still legitimate to compare the value of the HP the investors thought they had on Thursday with the company they realized they had after Friday to get at least a hint of what a more intelligent and straight-laced Hurd might be worth.

[politicalfootball]

It's interesting that this sort of measure is proposed to evaluate the value of a CEO, but isn't used to measure the value of Labor. When Labor issues an effective threat to withhold its services, nobody seems to argue that the resulting hit to the stock price is a measure of the actual value of Labor, and should therefore result in higher pay for Labor.

But to return to the topic at hand, I'll add another confounding issue to the professor's list:

-Maybe the sudden, unplanned departure of a chief executive is an inherently de-stabilizing, damaging event.

I'd argue that the professor's speculation about the death of a CEO, however, actually works in favor of the hypothesis that we should value CEOs according to stock price. In theory, the improvement in stock price upon the death of a bad CEO would be a measure of how bad that CEO was - a measure of how much the company would be improved by a merely average CEO.

[Todd Henderson]

What about looking at the market's reaction to the death of a spouse or child? This might be cleaner than CEO death, which would include judgments on succession planning, disclosures, etc. The loss of a wife or child might tell you more about the short-term value of a CEO, who might be out of work for a period of months, depending on the circumstances.

[MHodak]

The impact of the death of an underperforming CEO was observed with the 1983 demise of conglomerate builder Charles Bluhdorn, which was immediately followed by a jump on Gulf+Western's stock price.

[J. Bogart]

One would need to include consideration of investors betting on what other investors will do.