I'm willing to believe that CEO pay reflects poor governance practices to some degree, but this study fails to prove anything. As a simple alternative explanation for the data, suppose that companies that are performing badly have to pay more to attract CEO talent since the proposed CEO is taking on reputation risk if they fail to turn the company around. Then you'd expect exactly what the study found; the more highly paid the CEO, the lower the corporate performance.
This report should have looked at how CEO pay in one year affected company performance the next year and vice versa.