As long as the target you're optimizing for is different than what those 100 of PhD's (and most of the market) are optimizing for then maybe you've got a chance

As long as the target you're optimizing for is different than what those 100 of PhD's (and most of the market) are optimizing for then maybe you've got a chance. You also don't need to pay their salary. So I agree wrt/ to longer time periods. Most of those PhD's are probably trying to predict minute to minute moves, or daily moves. A lot of them would be out of a job if they lose over 90 day periods. Not sure how much of an edge can be gained over buying and holding some reasonably diverse equity portfolio. If anyone is thinking they'll beat some HFT hedge fund they're almost certainly not going to. If they're thinking of beating a reasonably diverse/optimized portfolio over longer terms that's maybe possible IMO but the difference isn't going to be huge. To amplify the difference requires taking on more risk, for example by leveraging, options etc. If I want to decide whether long term investing in the US vs. Turkey, Greece, Brazil, Russia, or the UK, or gold :) it's not clear machine learning can give me useful insight. The historical data has its limits. Predicting geo-political processes and things like interest rates for the long term seems like a very difficult problem. I would suspect the machine answer to this question would be something like P/E is lower so it's a good investment but who knows. So I agree with the economic/market insight comment as well.