It's similar to several other EA's I've coded and tested over the years, except it places trades upon a cross into a channel created by the high and low of a moving average, whereas all of my designs entered trades when price exited the channel.
At first glance at the chart generated by the back tests (not shown), I was impressed that it seems to capture trend changes very early.
The system listed on Forex Factory calls for a 5 period moving average on the daily chart to form the channel, whereas this robot uses a 10 period EMA on the 4 hour chart.
I ran about 5 back tests with various settings (no full scale optimization) and got some positive results. This is generally a good sign. Of course, I have tested only on EURUSD and only for a limited period of time. Still, when robots perform this well with very little modifications of the initial (wildly guessed) settings, it intrigues me.
Back test on 4 hour EURUSD from 2011-1-1 to 2012-09-05
Fairly nice equity curve, 25% return, 10.95% draw-down using an automatically set lot size (back test was with a $10,000 account and initial lot size was 0.1, adjusting automatically with equity).
Consecutive loss count can get scary at times (11 maximum constitutive losses), however that appears to be the worst statistic from this result.
This is a pretty good result. But I'm too scared to test it further, because it will undoubtedly be a failure! However, if these limited automated results can be this good, I'm sure trading this system manually will be even better. It's also an extremely simple system and wouldn't take much time to set up and monitor trades.
