probability of making a good decision that will swing huge amounts of money one direction or the other are exceptionally valuable.
For instance, assume a large company must make one big strategic decision per year that will increase or decrease profits by $1 B (a very reasonable scenario for a big public company). If CEO A has a 1% better probability of making the 'right' call, it is worth $20 mm per year for the company.
They don't get paid too much.