"The salary of a CEO could do with more reference to the cost of the next available substitute, rather than how much the CEO changes the stock price. That would

"The salary of a CEO could do with more reference to the cost of the next available substitute, rather than how much the CEO changes the stock price. That would require changes in how CEOs are compensated [...]" Would it? I assume that just like in every market, the pay of the CEO is based on exactly this. Supply & demand tend to be the cause of prices. When we talk about the value that CEOs bring, and quite specifically talk about how no one else could've done the job that Marissa Mayer did, we're implicitly talking about supply & demand - the supply of people who can do the job she did is very small, the amount of companies who would love to have a high-functioning CEO that can raise their stock price is very high, and therefore CEO pay is what it is. I assume that if Yahoo could've gotten some other succesfull figure at a fracftion of the price, they would've done so - it's in the shareholder's interests and they more-or-less make the decisions. Of course, I'm not naive enough to think that this is a purely technical matter - I assume a lot of human psychology, precedent and culture goes into this, e.g. the amount that exectuvies make is well-known, which causes other executives to raise their price to "compete", etc. This effectively makes wages higher than they "should have been". But I do assume that none of us are smart or knoweldgeable enough to say what CEO pay really "should" be, even assuming that's a well-defined concept. I think that most people just hear a big number, say "that's way too much" and have this as the bottom-line of their thinking, then come up with all sorts of arguments for why their bottom line is right, rather than questioning the core assumption and realising that maybe, CEOs are worth their price.