I did a few trades, manual again…. yeah, I know…. I have an addiction . At least I didn’t lose to much more, I think 2.5% , so , that is good .
I did get pretty far in peanut butter. The strategy is recoded , easier to read, and thanks to the discovery this morning ( saturday morning ), I think it is much more powerful than I originally thought possible.
So, here is the curve on lightly optimized settings. ( 15 year test, 2001 until now )
Honestly, I was very happy with this. I originally thought of a few variables that might be affecting the drawdown ( which here , it was 38% return with a 67% DD . Not exactly a great setup.
First I tried doubling risk on top-down that showed congruence ( looking at two higher time frames, and if both had MA’s working in the same direction ) and limit risk to half when it was the opposite . No change. I tried flipping this scenario , no difference.
I then thought that the orders made in price that ranged between the 50 and 200 on larger time frame were less probable ,… no change. I tried variant time frames with the same. no good. tried the opposite, still no good.
WHAT THE F!!!!!!
I checked the settings, maybe I had something different loaded. I loaded the saved set, ran it again…. same thing. WTF!
than it hit me, the spread was set to current. With it being the weekend and all, the spread would probably be something retarded, like, 170 . This being a scalping strategy ( now ) , a large spread would kill it. BLAST! I didn’t even f****ing consider the spread for entry/exit.
So , I go and check my brokers average, sure as shit. 1.3 is the average. Put it in the setting and ran it and:
Holy Macerole cereal shits!!!! In case you can’t see it ( you can’t ….. ) , it is a 300% increase with a 7% DD .
So, the system needs to take into account the current spread. Also, it needs to check the spread when a limit is executed and update the stops accordingly. Running a test with a relatively large spread should yield similar results, proving the strategy and my hypothesis that a high spread without change in stops is what is killing it.
Will keep you updated