Do you have any reason to believe you have a good algorithm to predict who these privileged few will be? Anecdotally, I've seen little evidence that being a good student at age X is reason to invest more heavily in that person going forward.