If the company hits it big despite them, the stock is awesome.
If the company does poorly they might miss a bonus but they'll still get their inflated salary.
If the company ends up on the rocks, they'll get their salary cut "till company performance improves" as their pay is usually tied to the performance of the company as an incentive.
But what invariably happens is that, even with their pay tied to company performance, the moment it gets cut, they high tail it out of there and land a new job with their top-10 MBA and another couple years of job experience...saying "I'm looking for growth opportunities" over and over again in their interviews.
I don't know if it's still true, but Facebook used to filter out these types by making them do a version of an engineering interview. Except instead of regurgitating algorithms you learned in college, MBAs had to regurgitate stuff from MBA school. Like derive the time-value of money, or discuss the TPM or whatever. The ones who coasted through MBA school got filtered out really quickly and the ones who took the subject seriously might pass the gauntlet.
The truth is, nobody knows how to bottle "success" and reproduce it -- but in a sense that's what management school is trying to do. With just the right management approach, applied in just the right ways, you can get Instagram instead of Color, or Google instead of Cuil.
The problem is that it's very easy to look at the failures and pick apart the problems. But it's very hard to look at the 1% that succeeds and figure out why -- and when it's attempted it usually overlooks the cases where the company succeeded despite having many of the problems the failed companies had.
Most companies succeed because of 90% dumb luck and 10% business strategy.
(hell, business is so screwed up people still don't understand how to define success. You see it here all the time that a "successful" startup is one that raises a huge round, not one that's profitable and growing...even BusinessWeek does this)