[abraxasz]

Ok, so I do believe that CEO's are usually overpaid with regard to their performance. That being said, I am not convinced by the methodology used in the study, specifically the period used for measuring the performance: 3 years. It sounds like a lot of time, but actually, I would expect a truly great CEO to undertake projects with a 5 years or more horizon. There's this great interview of Bezos discussing the point of focusing on quarterly results vs long term. So an alternative explanation for the findings could be: the highest paid CEOs invest in 5-10 years projects, so their results after 3 years are below average.

Again, I repeat that even before reading the study, I intuitively agreed with their conclusion, but I'm not so sure about their argument to prove it

[Spooky23]

Normally I would agree with you, but the average tenure of a Fortune 500 CEO is less than 4 years. If you raise the bar for period of performance, you're filtering out the bozos who go away quickly.

The way I read it that if you're not a superstar, you're investing lots of time horse trading and kissing babies to get the big bucks. That time would probably be better spent doing work productive to the business.

[CWuestefeld]

Looking at the first 3 years seems almost doomed to failure. It seems to me that it's pretty likely that a new CEO is going to be taking a company in a direction at least slightly different.

That means cutting out some projects may have been just about to pay off, while at the same time embarking on new ventures that won't bear fruit for some time.

One might interpret these results to say, "the highest-paid CEOs are engaging in the most expensive change for their firms". It's only natural that significant change is both forgoing some income that would have been realized soon, and creating risks for what's in the short-to-medium term while they look to the horizon.

This is all to be expected, and doesn't necessarily say anything about the amount of value they create in the long term.

[cpeterso]

The study should also consider the company's performance before the CEO takes over. If a CEO is coming in to rescue a company that's in trouble, they might expect to be compensated for the additional risk they are taking on.

[Spooky23]

Stock prices are based on the market's view of the future. You don't evaluate a CEO like a middle manager. If he cuts some critical program for strategic reasoning, the market will either punish him or reward him based on the future outlook.

[voronoff]

Stock markets also consider risk, so when a new CEO pushes for a new direction, that's penalized initially even if it's the correct long term approach because more is unknown.

[TheOtherHobbes]

Stock prices are animal herd logic, not science. See e.g.

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=670404

If you believe that business should have long-term horizons and not simply be a machine for organising the maximum possible quarterly profit, stock markets in their current form are one of the least efficient ways to organise labor and capital.

Inflated CEO pay is only one symptom of this.

[CWuestefeld]

Stock prices are animal herd logic, not science

That's not how I understand your link. The study suggests to me that analyst recommendations are unreliable predictors of short-term gain. And it certainly does not show that changes in a company's overall market capitalization are based on a herd mentality.

Note that it says that those rebalancing according to advisors achieved higher terminal wealth, but lower risk-adjusted return. Most importantly, they did do better in the end. But adjusted for risk, it's worse. This suggests to me that the analysts usually do pretty well, but when they flub it, it's a doozy.

[njharman]

A couple issues. Poor CEO's probably don't last 5 years. I'd suspect a strong correlation to performance and CEO's "horizon". In that those CEO's focused on next years or next quarters numbers are likely to thave poor (longterm) performance. Superstars such as Bezos or Musk are probably anomalies and not useful in studying CEO performance in general.

[tjradcliffe]

Bezos and Musk are also founders. The basic problem with CEOs is that they pose a huge agency issue: they are supposed to be appointed as agents of the owners who are themselves represented by the Board of Directors. However, no one has ever found a way to provide effective oversight to ensure that CEOs act in the interests of the owners rather than in their own interests.

Attempts to tie CEO interests to owner interests by tying compensation to the company's performance have thus far failed, mostly because a) the fraction of compensation isn't sufficient to overcome the CEOs self-serving motivation or b) the specific tying system can be gamed to the CEO's advantage.

Furthermore, there is an asymmetry in the degree of interest between the parties: CEOs have a very large interest in maximizing their own compensation, while shareholders see CEO compensation as one expense amongst many others, and as such are less interested in it.

These issues have been known for a long time, and again: no one has come up with a viable answer. Anything anyone who is posting here thinks of has almost certainly been thought of before, tried, and seen to fail.

[georgemcbay]

"Attempts to tie CEO interests to owner interests by tying compensation to the company's performance have thus far failed, mostly because a) the fraction of compensation isn't sufficient to overcome the CEOs self-serving motivation or b) the specific tying system can be gamed to the CEO's advantage."

and/or c) extrinsic motivators aren't effective for non-mechanical work.

http://www.ted.com/talks/dan_pink_on_motivation

[gjm11]

What Dan Pink describes is that for non-mechanical work, extrinsic motivation is ineffective in improving performance: how quickly or accurately or insightfully or imaginatively one accomplishes a particular task.

But the question with CEO pay is a bit different, and I don't think the studies Pink refers to say anything about it: what can you do to encourage CEOs to act in the best interests of the companies they manage?

This isn't a matter of, e.g., whether they notice opportunities to attack a previously unaddressed market -- which is the kind of thing for which Pink finds extrinsic motivation isn't good at improving. It's a matter of whether, having noticed such an opportunity, they further notice that taking it would be good for the company but bad for them personally because, e.g., it would have short-term costs that would drive down the value of the shares they were hoping to sell next year to buy a new house.

In which case, the findings Pink describes might actually be good news: the prospect of personal gain may be ineffective in making unscrupulous CEOs good at spotting opportunities to screw their company over to make a quick buck.

[yuhong]

I wonder if corporations that uses entirely non voting stock exist.

[alexeisadeski3]

Who would pick the directors?

[YZF]

You'd be surprised how long poor CEOs can last. I worked for one large company where the market cap went to a 1/10th of what the company has spent on acquisitions alone during the CEO's reign and a continuous sequence of questionable business decisions that included laying off pretty much anyone who could have been instrumental in helping the company recover. I think there are many other examples.

A CEO can always blame the economy, blame the market, blame the weather, blame their predecessors or otherwise spin things in a favorable way. They can sell a story of tough times needed to make a turnaround. Sometimes there will be some secret scratching behind the back (I'm on your board, you're on my board). There's the cognitive dissonance on the board side of things, we've selected this guy therefore he must be good. etc.

As others have said, the incentives are set in a way that a CEO looks primarly after his own interests. It's safe to assume that CEOs that make a lot of money are even better at looking after their own interests. They are also very good at convincing others that they're not looking after their own interests.