I appreciate Quantopian, a lot, and I understand what it's trying to do. What I don't understand though, is if Quantopians aim for the fund is hundreds of independent returns streams, all running off of different strategies and alpha sources. Why are we constrained to a set Universe?
Surely if all the algorithms are constrained to trading the same stocks, or similar stocks, you're going to get a lot of similarity in the returns that you wouldn't see if you had algorithms trading small cap stocks, high cap stocks and anything else that's available on the platform. I understand the need for all the other constraints, you want to make sure that algorithms, which appeared to be un-correlated return streams aren't going to suddenly turn around and move in tandem with each other. In my mind constraining eligible algorithms to a specific number of stocks can only be detrimental to this aim. How many different return streams can you get off of the S&P500? A very limited number. Expanding that out to the QTradableStocksUS Universe there's still going to be a limited (albeit much larger selection). You're also increasing the chance of algorithms from the fund treading on each others toes, taking long positions and short positions in the same stock at the same time.
If an algorithm can function on stocks which aren't in the QTradableStocksUS Universe, and can find liquidity to trade. Why is it automatically disqualified from the contest? Surely the ability to do that shows that it's an un-correlated return stream in and of itself and massively different from anything else that's going on.
It seems to me at least that everything in the Risk Model is designed to make our algorithms more likely to be un-correlated return streams, except the QTradableStocksUS requirement, which in my mind does the opposite.
I would be really interested in why Quantopian decided on this constraint?
(Apologies if this came off confrontational or anything, it's meant to be anything but. I just struggle a lot with conveying tone over the written word. I am genuinely interested in Quantopian's rationale behind this decision.)
