The Financial market is not a complex calculation. It is a derivative of human emotion and psychology. In the end humans will try to manipulate the computers. The best investors such as buffet and soros are not math whizes. They take qualitative and quantitive approach to their investment.
Tell that to Jim Simons, math PHD and multi billionaire founder of Renaissance Technologies- a hedge fund based entirely on computer trading. There is certainly room in finance for quants.
Any data mining driven approach that shows good returns is arbitraged out of existence almost immediately, unless it's niche and thus unscalable to a larger fund. All the blackbox systems end up doing pretty much the same thing and getting low returns. The markets are simply too dynamic for a pure data and algorithms system to work. One needs to use theory to win, and the guy picking the theories in the program is exactly the same thing as a manager in the first place.
The only way to get rich from a naive blackbox is to twist the risk knob up to 11, which always goes badly eventually.
You can sometimes pick out the quant plays on the very short term. If there's a super obvious area of support, the algorithm will force the bid below that point, causing a good amount of stop losses to be taken out, then the market will snap back up when they cover.
Sounds like a major fail to me.