Hello,

I developed the algorithm below and submitted it for review (albeit through support not the contest) a couple of months ago and didn't hear anything back. I then submitted it via the contest over a week ago and haven't heard anything back either.

Has anyone else experienced this? The Quantopian support people are splendid but this extended period of silence seems really strange. I'm not expecting anything in particular (the algorithm doesn't pass a couple of the contest requirements - single name exposure and number of stocks traded). However I, like many others, have spent months and months on the research, coding, testing, recalibrating, etc. So hearing nothing is ...well a bit annoying. I know this is just a hobby but still, one that people take quite seriously.

On the algorithm itself, it back-tests ok, with good returns and what to me seems very low volatility (comments always welcome). Its premise is that there are groups of securities where their price levels are cointegrated, returns correlated, and returns multivariately normally distributed. I've got some code in research that takes c3-4 hours to run that finds groups that have these properties. In an ideal world I'd put this in the algo but I'm not clever enough to think of a way of doing this within the time constraints of an algo. The algorithm then mines these groups to find price dislocations (where the position of the combination of prices for a given group [works best with just pairs]) is at the tail of the groups' multivariate normal distribution.

Feel free to share your thoughts on the algo or whether you've experienced similar periods of silence. I realise this post is part rant, part sharing of ideas.