The paper itself covered a lot of ground (including "firm size and prior stock performance, especially over the past three years, are significant predictors of both excess cash and incentive compensation"; the results confirm than recent stock performance is positively correlated with CEO pay). It didn't however give any detailed consideration to the obvious other hypotheses:
(i) Far from being overconfident about their decisions, top paid CEOs (and the boards that negotiate their pay packet) are actually more conservative than the market about the future prospects for the stock, so are willing to agree bigger stock-based incentive packages as a result. CEOs generally should have more insight into the workings of their firms than Wall Street analysts, of course...
(ii) If a CEO's pay packet at time t predicts good stock performance since t-3 and bad stock performance up to t+3 might it be that their ability to prevent the firm's stock price reverting to the mean is limited (even if they're actually still doing a good job... and receiving acknowledgement in the form of pay packets more focused on that difficulty)
There's also
[iii] The decisions the CEO makes is not even slightly affected by their pay structure but high CEO compensation packages is a good proxy for lack of board oversight
but that's a lot closer (and to a large extent compatible with) the authors' hypothesis about overconfidence
(i) Far from being overconfident about their decisions, top paid CEOs (and the boards that negotiate their pay packet) are actually more conservative than the market about the future prospects for the stock, so are willing to agree bigger stock-based incentive packages as a result. CEOs generally should have more insight into the workings of their firms than Wall Street analysts, of course...
(ii) If a CEO's pay packet at time t predicts good stock performance since t-3 and bad stock performance up to t+3 might it be that their ability to prevent the firm's stock price reverting to the mean is limited (even if they're actually still doing a good job... and receiving acknowledgement in the form of pay packets more focused on that difficulty)
There's also [iii] The decisions the CEO makes is not even slightly affected by their pay structure but high CEO compensation packages is a good proxy for lack of board oversight
but that's a lot closer (and to a large extent compatible with) the authors' hypothesis about overconfidence