Profit Generator EA - page 66
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Ok, we all know I had to have my 50 LWMA filter and thanks to Nich, I got it. At first look I was rather disappointed while doing backtesting. Then I stumbled onto something incredible...Fuzzy Logic! I input a -20 ma pips level instead of trying to follow the logic I originally intended. In backtests the results were incredible, for example, with the EUR/USD doing TP-10, SL-30, using obsolete set to 1 minute, TF1hr, it reduced the number of trades from 490 to 90. At the same time, it increased win/loss ratio to over 95%, it reduced drawdown to just .5%. Also, previously untradable pairs showed a win/loss of 85%+. Today I watched the trades it made with these settings and it was incredible how it worked, so intelligently, it was like watching a human trade. Plus, no time filtering and it's only been trading during the proper hours all by itself!! I even had 3 winning USD/JPY trades today with no lossers. I have an idea to take this fuzzy logic filtering a step further using the pivot levels, yes there it is again.....More later..
You lost me there Nova.
Is that TP= -10 (minus ten?),SL minus thirty and a minus twenty MA pips level?
Weird
Sada
You lost me there Nova.
Is that TP= -10 (minus ten?),SL minus thirty and a minus twenty MA pips level?
I'm confused too. Are you suggesting these settings on H1? (with other settings default):
extern int stoploss=30,takeprofit=10;
extern bool UseObsoleteMethod=True;
extern bool OMwhenLossOnly=false;
extern int ObsoleteMinutes=1;
extern bool Use50maFilter=true;
extern int MApips=20;
extern bool UseTE=false;
extern int badtrades=2;
extern int ReEnter=2;
Sorry
the TP=10, SL=30, 50 MA filter=true, MA pips=minus 20, obsolete=true, only when profit=false, obsolete time=1 minute, no time filter, everything else was default.
The other thing I forgot to mention was I had no contradicting trades today, automatically!
Let's get sick with this thing
Please see post #523 as a primer to this discussion: https://www.mql5.com/en/forum/173795/page35
Ok, here is my new proposal. First, let's forget the whole trade pivots strategy for now and instead I would like to focus on using the first two pivot level filters along with the 50MA filter which we already have now.
We use the "pivot Power" and "Headroom" filters created with the fibopivot; i.e.; daily range indicator. The 50MA and daily range are the holy grail in forex trading. So we will have these important levels working in our program giving it what would be a rigid logic but we will introduce a randomness around all these levels by using negative inputs instead of positive inputs, or a combination of both. This will keep us working around these levels but give the program breathing room to work it's magic.
So, just like I input a minus 20 for the 50MA filter, we could skew the logic with the pivot power and headroom filters in the same way. Or put in a combo of positive and negative numbers with the three filters to create some serious fuzzy logic while staying in the ballpark of the trading range.
We will be playing with some high level math then, I can assure you.
the TP=10, SL=30, 50 MA filter=true, MA pips=minus 20, obsolete=true, only when profit=false, obsolete time=1 minute, no time filter, everything else was default. The other thing I forgot to mention was I had no contradicting trades today, automatically!
Thanks for the quick answer Nova,
I assume these settings are for the latest 3.3.1 version?
Sada
Thanks Nova. Got it now:
extern int stoploss=30,takeprofit=10;
extern bool UseObsoleteMethod=True;
extern bool OMwhenLossOnly=false;
extern int ObsoleteMinutes=1;
extern bool Use50maFilter=true;
extern int MApips=-20;
By the way, you are right--these settings backtest beautifully! Of course, we seem to continually be disappointed in the disparity between backtests and forward tests. But you may have something here. We definitely must forward test. Good work!
Novafx99, I appreciate your thinking outside the box, I find that I do that often in other areas of my life but am a sheer novice doing it at trading and yet I'm aware that like anything else it just takes effort plus practice. I am going to assume that in addition to using fuzzy logic part of the positive pip production came from using TP 10 pips, I have been toying around with grabbing lower pip TP with this EA myself. Great brainstorming.
Well, forget about profit and pips. What the backtester shows is an improvement in win/loss ratio, drawdown and margin requirements due to the seriously reduced number of trades. Reducing our trades and increasing our win ratio while decreasing drawdown is what good filtering should accomplish. Seriously, do you really think you want to take 490 positions in the euro over a 30 day period? This is ludicrous. Our equity will develop just fine with reduced trades if they are better trades. It worked great today in live trading. We did have a decent trading day for the system but I think the filtering ideas will prevent us from getting killed during the breakouts. Plus, it seems that the EA can run more or less unattended instead of introducing the time filters. Just a convenience I guess. Wasn't that the point? To be able to let it run 24/7?
The latest version is posted here: https://www.mql5.com/en/forum/173795/page40
Here is the latest verison. It prevents contradicting trades on the 3 majors. Thanks for contributing nova.
Nich,
This doesn't seem to have Nova's new srttings with the -20 pip thing...
Can you make one with his exact settings?
Thanks,
Billy