[digz]

I haven't read the paper, just the article. It's possible some of these are directly addressed, but I would want to know more before any credibility is given to the conclusions.

Some alternative explanations:

- The 'lowest paid CEOs', per their definition, are typically the ones who have the most skin in the game. When you have many CEOs out there with only $1 compensation, and just stock (and not stock grants as the author mention), they will inevitably be on the bottom. Think Zuckerberg, Google, Apple w/ Jobs, etc. One would expect this highly underpaid group to outperform.

- The highest paid CEOs are often times the ones dealing with the most troubled companies. If you were a shareholder of Kodak in 2000 and saw that digital cameras were coming, would you want to pay for the best CEO possible to ensure you could harvest the most out of the company? Kodak would still underperform the market, but maybe they would have underperformed the market more with inferior management.

- This is a little technical, but with an experienced manager pulling in big dollars, it's more likely that this is a well established company and manager that are well understood by the market. This means that the risk premium demanded by investors would be smaller resulting in a higher stock price. This means that the stock has less to move. And on the other hand, more unproven companies with cheaper managers will have higher risk premia demanded by investors, resulting in lower stock price and therefore have more room to go up over time as some of the unknowns are answered.

Ideally we'd need include other variables such as market size and expected growth (P/E Ratio), CEO share ownership, ex-ante company distress, etc. into the analysis. Based on the described methodology, it doesn't sound like that was done.. but again, didn't read the paper itself.