Carefully read what other people wrote here before. Unfortunately, almost all complaints are true - inexperienced management (yeah, knowledge of how to make money doesn't automatically entail knowledge of how to run business), promises without realization, under-compensation, treating people as slaves.
Important thing that many speakers have mentioned - tense and destructive competition. The problem is that the company does standard statarb everybody does nowadays, and doesn't want to expand its areas of expertise, but expands it research stuff. As the result, all simple and medium-difficulty ideas are already explored, and new researchers have no room to show their talent (and earn bonuses). But management just doesn't see this problem.
Ok, put aside money/atmosphere, if you just want to get a bit of experience. What about major - research - side of the business? Bad news for you too: you're going to work on ancient c++ infrastructure, without usual stuff you have probably got used to (linear algebra, econometrics, statistics, machine learning, etc.) Have you seen their websim? Actually, that is the level of infrastructure you're going to work with. Forget about gpgpu and other modern technologies - old school c++ and a couple of threads are your friends here, welcome to the 20th century, son!
Another problem here is that due to tense competition people don't have time for self-development and education, which further flattens learning curve.
