<<Do you use stops?>>
I programmed in stops once, but the results did not favor using them. There were 3 or 4 instances (in the backtest) where stops would definitely have helped, but there were lots of times that I (the backtest that is) got stopped out and then a reversal soon took place, leaving me holding the bag.
But I was testing hard percentage stops, like 7% or 10%. It might be a good idea to revisit that testing, but use variable percentage stops that link to Average Daily Volatility or Average True Range.
Market Data 
Markets 
AI
