Quantopian's community platform is shutting down. Please read this post for more information and download your code.
Back to Community
long-short multi-equity algo

Here's an example perhaps of interest. One thought would be to see if this could be applied to it:

https://www.quantopian.com/posts/machine-learning-on-quantopian-part-3-building-an-algorithm

2 responses

Here's the same backtest, but with:

    set_commission(commission.PerShare(cost=0.0, min_trade_cost=0))  
    set_slippage(slippage.VolumeShareSlippage(volume_limit=1, price_impact=0))  

I've been able to run a backtest back to 2010 (reportedly the time frame typically used for evaluating Q fund algos). It looks pretty good, but is a pig that can't be loaded onto the forum, and into the research platform. I'm working with Q support to understand why.

One curiosity is what a hedge fund is supposed to do in the first place, for its customers. Presumably, when the market (SPY) gets choppy, the hedge fund portion of an overall investment portfolio should smooth out the returns. But when there is a bull market, then flat returns on the hedge would be o.k., since it is a hedge against the inherent volatility of the market (and perhaps if one tries too hard to get the hedge returns up during a bull market, it is self-defeating).

Feedback and improvements welcome. As I mentioned above, I'm particularly curious if the alpha combination can be improved, with ML or otherwise.

isn't your volatility custom factor just going to weigh on high priced to low priced stocks?