I am trying to get the seond system working since yesterday and I must say... At least this one is not so easy to CODE (for Ninja im my case). It seems to work pretty well, but it is not the easy "buy as x crosses" thing.
It is really easy for a person, but I am fighting mostly API wise. Strategy two involves identifying an intra bar and then setting up a number of potential trades
SO far so easy.... Sadly there are up to two trades on each side, running possibly multiple days. Once a trade on one side triggres, the other side is cancelled (long breakout, cancel the short breakouts)
What gives me creeps (not in the strategy itself, mind you - basically in Ninja) is that if an intrabad happens again and the first trade is still open (and has a position)... The new trades are added. Basically instead of managing "positions" one deals with "trade setups". Something that seems to be out of the concept of most automation systems
Well...
Anyhow, from what I can see so far it has prety good results. I am still fighting on even the breakout (stupid ninja - no debugger, and I sometimes just do not see the trades happening, and its hard to debug. So, it may not be proper to work with this thing automatic yet.... but I think I will set up an indicator that automatically identifies and warns on a chart.
At least the intrabar breakout possibly can be a LOT more profitable if one adds a third trade on reversal (stop is 68,1% a speficic range - may be sensible to add a third trade on a 50% reversal), and has a trailing stop strategy instead of fixed targets. All that would really really really need some automatic testing....
Life sucks.
Anyhow, this makes things possibly a lot more profitable but also a lot more complicated. Given as they are they are really easy to deal with. A little excel magic helps (adding high and low of the last 2-3 bars), so that the levels for breakout, target and stop are automatically calculated.
Amazingly simple.