I’ve speculated before that one of the main reasons that CEO pay has risen quickly since the 1950s might be that CEOs have become more important since then. The question is very important because people who worry about executive pay use this rise as evidence that CEOs are overpaid now – if they were only paid ten times what the average worker was in 1965, how could they possibly be worth two hundred times that now? A new paper tests this hypothesis. It looks at what happens to firms’ values after CEOs die unexpectedly. (To be precise, they look at “cumulative abnormal returns”, or the difference in firm value compared to what was expected, over a five-day period.) If CEOs are important and difficult to replace, firm values should move a lot when they die. If they aren’t important or are easy to replace, they should move a little. Note that we shouldn’t expect firms to always become less valuable – a bad CEO dying should make the firm more valuable, just as sacking a bad CEO does.
Interestingly, they find that the average change in firm value doesn’t change over the 1950-2009 period. That is, the good and bad CEOs cancel each other out. But the variance – how big the changes in value are – has changed a lot. Over the course of 60 years, “the shift in market value caused by an unexpected CEO death increased by approximately $65 million (in 2009 dollars)”.
What that implies is that CEOs have indeed become more important to firms over the past sixty years. The good ones are more valuable, the bad ones are more costly. I find that quite intuitive. Markets seem more competitive now than before, and technological change seems to be moving more quickly. That means that the strategic decisions that a CEO will help make matter more.
It also undermines the claim that CEOs are paid more than they’re worth. For sure, some of the cost to firms will be risk-based – the process of finding a new CEO is costly even if the old one was no good. But the fact that the cost is much greater for some firms than others – and that some firms do better when their CEO dies – seems to be good evidence that CEOs matter, and matter much more now than they once did.