A good CEO can create massive value, but the present value of any particular choice of CEO -- which should be an upper bound to the performance-independent pay they are able to negotiate -- needs to be diminished according to the uncertainty in the evaluation. That's where the leverage excuse falls apart: uncertainty is through the roof. Everybody's throwing darts at a dart board and everybody knows it, so if the story about rational value estimation and leverage were the whole truth we would not expect to see CEOs reliably paid stratospheric salaries for mediocre or failing performance, yet that's exactly what we see. We should therefore consider alternative explanations.