Hi Q Folks,
Thanks for fixing the cvs download with all the risk and performance metrics. Very interesting to analyse!
I'm curious though how it's possible for so many of the entries to go above (and below), sometimes by a lot, many of the contest requirement boundaries? I've had my algos automatically withdrawn several times for only just slightly breach one of the boundaries. Are these outlier 'spikes' and is there a tolerance level for these 'spikes' for all of the contest constraints? Many thanks in advance.
Regards,
Joakim