Fabio,
This is exactly the same results I am showing, and the same thing it did all last week. The EURUSD has been trading in a wide
range 
for the past month and somehow the robot can detect this and respond accordingly.
I currently have 3 sell and 2 buy trades, opened in order, sell 1.2667, sell 1.2666, buy 1.2667, sell 1.2646, buy 1.2663. As it hits some targets and opens new orders the
spread between orders will widen.
If you look at the chart, the profit targets on these are all right in the trading
range . Its a gamble to predict which will get hit first, or which way the
currency will break from here. I suspect that the lower ones will get hit first, but that the upper ones will eventually follow. As long as it maintains this
range this could go on for a while. When it does break though, we will have to take a loss. Better if we do so quickly and aggressively, rather than waiting for the program to stop out.
Watch out for orders that are opened near the limits of the
range without an easy target. It might be best to close orders like this if they dont seem likely to have a quick return, or at least set a stop and a realistic target. I had one that opened when I first turned on the robot that lasted all week and cost me in missed oportunities and eventually a force sell.
Once the price forms a
trend we should see the trading direction align. On their test results it appears to work that way. It also shows some losses that I believe are avoidable, or at least can be reduced with a more active management aproach. This is something I've been saying for a while with the Steinitz, and most any EA program, mechanical system, newsletter recomendation, etc. Any situation where a general arbitrary setting is used for entry or exit can be improved upon.
I'm psyched that the Steinitz is working with this method. I was concerned that the
indicator is too lagging to produce a profit without a strong
trend . It does lag badly. Often if you look at the color changes and the actual price, it seems to get in and out at the worst possible moments. Thats why the method of taking repeated small profits was working well. Unfortunately that last one would always get caught without a good exit strategy.
5% compounded weekly is great money. I like this method, it locks in the gains and also means your generaly out of the market before the weekend. I also like the idea of using the
indicator signal as a stop. Do you still use the small take profits along with the
indicator exit? What do you do if the robot doesnt hit the 5% target in 1 week?
Be smart! Read the 
.... But FWIW you *can* run it without the HAS indicators loaded - as they suck up the most... or just run the limited version of the 
Linear Mode
