Many firms have done their own internal studies though. IIRC Google undertook a big study and found no correlation between gpa and job performance. I'm sure financial firms have probably looked at the link between gpa and alpha, if only because it's so easy to measure.
A more interesting thought experiment is to ask yourself the opposite question: For students who have jobs before going to college or grad school, would you expect those who made more money to later get a higher GPA?