Yes that would probably explain the negative correlation. That said, if good CEOs actually controlled the direction of the company and were compensated higher

Yes that would probably explain the negative correlation. That said, if good CEOs actually controlled the direction of the company and were compensated higher to attract talent - I would expect that to override the negative effect for a net positive. Since that's not the case, I think as you suggest that CEOs are "without active influence on success"