No. of Recommendations: 8
FWIW, so long ago that I can't remember details, I looked at shorting those falling below SMA200 (or was it a momentum strategy?).
IIRC, the 'bad' choices don't necessarily do badly (which would be good for shorting), instead they just don't do so well (bouncing around, so not great for shorting).
If eyeballing behavior of some low ranked choices isn't sufficient, the 'pro' version of portfoliovisualizer allows you to upload your own return time series. Could backtest by uploading the negative return series of the thing you want to short. Wouldn't include fees of shorting, but if it's not looking good before fees then it certainly won't after fees.