4 ms·

The headline is probably true in the statistical sense. The sample size is undoubtedly small and economic data are often tainted by the efficiency effect--the '

The headline is probably true in the statistical sense. The sample size is undoubtedly small and economic data are often tainted by the efficiency effect--the 'experiment' is uncontrolled and actors might already be choosing the results that are best for them, dramatically reducing the apparent statistical effects of their decisions. Finally, CEO pay is often negotiated before performance becomes apparent, and for something like CEO I'd imagine it's really hard to tell how well they'll do in advance, even if you're trying to pick the best people for the job. It's not like there's a CEO aptitude test, it's not like the skills picked up being CEO of Pepsi are necessarily transferable to being CEO of Apple Computer (for example). The headline might also be true in the economic sense. I'm not especially qualified to judge this. The article, however, is quoting the apparently private data of one obscure financial research firm. So I'm filing it under Not News. (I also just noticed that they call Warren Buffet 'underpaid'. The guy earned fifty billion dollars being CEO of Berkshire Hathaway. Calling him 'underpaid' seems like a bit of a technicality.)

JumpCrisscross 15y ago
I think the point is the board should structure pay so that when the company does well the CEO automatically earns more and when it doesn't he doesn't.
- cynicalkane 15y ago
  CEO pay is frequently structured like this already. CEOs will often get half or more of their pay in restricted stock and options, and it's not unheard of for their salary to be a token value like $100 or $1. They also have the incentive of not being fired, and being re-signed if they're on contract. The alleged problem is that the vesting/restriction periods for these stock and options is too short, like a few years. I have a pet idea that CEO pay should be mostly in stock that vests quickly but is restricted for a very long mean time, like 20 years. But that's just a pet idea and I bet it would be a very hard sell to potential CEOs. I also think pay for performance is overrated. It won't help if you have the wrong guy in the first place.
  - jandrewrogers 15y ago
    You are forgetting that this compensation is then discounted for time. And 20 years would incur a very substantial discount. Every CEO knows how to do that math. Radically increasing the vesting time means that you would need to grant several times the number of options that you would need to grant for shorter vesting terms, just to maintain parity. The potential dilutive effect for other shareholders (including employees) would be rather large.
  - JumpCrisscross 15y ago
    I also think pay for performance is overrated. It won't help if you have the wrong guy in the first place. True, but with pay-for-performance hiring the wrong guy means you paid less (since he didn't perform). Further, I don't know of a viable alternative. Don't pay for performance? Now you're dealing with a problem of adverse selection - you're more likely to land with a lemon. The chief problem with CEO pay is that it exhibits much optionality without compensation to the shareholders for that - if the CEO does well he experiences a gain, if he doesn't, he just gets fired (unless the failure is fraudulent or very public).
  - yuhong 15y ago
    And also that the wrong incentives can do more harm than good.

JumpCrisscross 15y ago

I think the point is the board should structure pay so that when the company does well the CEO automatically earns more and when it doesn't he doesn't.

cynicalkane 15y ago
CEO pay is frequently structured like this already. CEOs will often get half or more of their pay in restricted stock and options, and it's not unheard of for their salary to be a token value like $100 or $1. They also have the incentive of not being fired, and being re-signed if they're on contract. The alleged problem is that the vesting/restriction periods for these stock and options is too short, like a few years. I have a pet idea that CEO pay should be mostly in stock that vests quickly but is restricted for a very long mean time, like 20 years. But that's just a pet idea and I bet it would be a very hard sell to potential CEOs. I also think pay for performance is overrated. It won't help if you have the wrong guy in the first place.
- jandrewrogers 15y ago
  You are forgetting that this compensation is then discounted for time. And 20 years would incur a very substantial discount. Every CEO knows how to do that math. Radically increasing the vesting time means that you would need to grant several times the number of options that you would need to grant for shorter vesting terms, just to maintain parity. The potential dilutive effect for other shareholders (including employees) would be rather large.
- JumpCrisscross 15y ago
  I also think pay for performance is overrated. It won't help if you have the wrong guy in the first place. True, but with pay-for-performance hiring the wrong guy means you paid less (since he didn't perform). Further, I don't know of a viable alternative. Don't pay for performance? Now you're dealing with a problem of adverse selection - you're more likely to land with a lemon. The chief problem with CEO pay is that it exhibits much optionality without compensation to the shareholders for that - if the CEO does well he experiences a gain, if he doesn't, he just gets fired (unless the failure is fraudulent or very public).
- yuhong 15y ago
  And also that the wrong incentives can do more harm than good.