Help Needed to Debug Existing EA with proven strategy

Nihilist, I think the profit target breaks the spell, given the way trades are taken the system needs to determine the profit target,D
EDIT
If trades go in a straight line it's a no brainer, hence the volatility Phillip talks about, when this dips the price wanes to open multiple positions to cover, D

agree 100%

yes,but as we reduce the profit target we reduce yearly return but we do not necs reduce the draw down. (Thats what I have found ).
I really want to see this working and ive been testing it all day and cant find any working long term solutions. it would be great to see someone show some good results over a year,then we can test a few more.
ill keep testing..

I think with a slow start and all going in our favour we can turn "dines to dollars" with that said, to keep a yearly return buoyant one can start with a higher balance and lots, trading one pair as Phillip has stated, all grows with time, more money, more pairs, what you think, ? D
EDIT
Going to eez off posting until we hear from Phillip, just to make sure we're singing from the same hymn book, lol, D

not sure a lot of people will have more than 10'000 in order to trade 0.01 lots..Thats the harsh reality as the settings stand at the moment.
Im not saying we cant find a way,if its there we will find it

Hey guys, need to some help to understand something


As you can see, the drawdown is really big, 80%, but nor the equity and balance graphs fall under my deposit. I know i am missing something but don t know what. Any idea?

edit: if the drawdown is just the big fall in the graph, it s not a problem since it doesn't fall under my deposit

Can someone please check.I do not think the candle filter is working correctly. by increasing the filter value we should be able to stop ea trading, we cant seem to do that.

a guess, and it is a possible one, buy and sell trades are taken as the price fluctuates, now depending on the profit target as I have highlighted, the more profit looked for, the more the fluctuations of the price and trades are opened, a simple 10 points could be had with one trade, but 100 points, the waves in the market could open many positions, today my account was maxed out with open trades, no room for anymore, until pointed out, my profit target was the default, then I changed it to 5 meaning $5, £5 etc,, and low and behold, the profit was hit, just by chance, so a possible answer to an 80% drawdown could be max trades open, I hope this sense, at no point did my level of profit go that low, however, all the open trades put me in the red, D

Thanks, that's a bug and I'm correcting it now.

Hi, Swingman and Philip, what about set TP as a percentage of balance? If so, it is possible optimize. Greetings

Very promising thread.

Thanks Philip for sharing your ideas and thanks to Swingman for coding the combined EA.

I'll help testing and optimizing the settings.

Regards.

Hi Guys,

Thank you guys for the numerous feedbacks.

I really do not have the time right now to answer/reply to the numerous suggestions put forward but these are my general opinion/views/observations:

1. Let's concentrate on getting the core logic/strategy working first without wasting time on extraordinary items which are nice to have but non-essential at this moment in time eg money management and other suggested extras - these can be added at a much later date when the basic core logic/strategy is proven and weaknesses minimized.

2. I am beginning to think that what Detector had suggested, namely, limiting the number of opened trades is a good and essential feature which ought to be coded in - this alone can remove the only weakness in Hybrid Mode - but then again, how do we deal with the 'hanging' opened trades??? This is not a real issue with experienced user who can easily manually trade out of the position - the challenge here is with newbies who rely 100% on the EA to trade for them. Any suggestions??? My observation so far is that, so long as you respect the $1,000 per 0.01 lot traded, the EA will always get you out of the deadlock, you just need to be patient to wait for the breakout which as I reported the longest took 40 days to happen with Tic-tac-nur's EA (hybrid mode without the filters). Another possible approach is to cut losses and simply close up all the opened trades at a loss at certain level - by setting the Hybrid Mode stop loss much lower - say if we target profit at $50, we set stop loss at say $300 to $500 which represents about 3 to 5 opened trades. I will try back test on this today. What do you guys think?

3. I personally would not trade this EA on any other pair than gbpjpy.

4. I have thought about this before Swingman came into the scene (when we only had Tic-tac-nur EA to work on) - I had observed from the extensive backtests in hybrid mode using Tic-tac-nur's EA (without the filters) that if we could implement a 'one-single order only martingale' strategy would resolve 80% of the dead lock situations. Let me explain. Don't get scared off by the dreaded word "martingale" - I am only suggesting one single order where martingale is applied at the very beginning of a cycle in order to prevent the cycle degenerating into a dead lock cycle. The logic is as follows - tweak our settings in such a way that it does not need more than 1 opened order to hit the TP (if 1 full lot is used and depth set at 10 pips, then TP must be set below $100, say $50 to $80) - in other words, when at the very beginning of a 'deadlock' situation when the price reversed after having triggered the one single trade & before it hits the TP (if we have tweaked our settings as above - we would only need to deal with ONE single opened trade) - so, I think its very effective to implement a martingale strategy at this early stage to remove the 'cancer' before it gets out of hand. Arrrhhh...my two years old son just got up and disturbing me all the time...i will post this first & continue in a short while...

Just for fun, here is the 3.2-Rocket Version

The settings are identical with my #Post 467
https://www.forexfactory.com/showthr...3#post10991933

New is the MoneyManagement with 0.01% from my $10,000 capital, to have at begin 1 lot. At the end there were 3.8 lots.

The win is $ 28,220 (were before $ 23,700)

I still have some programming issues, so I'll post the version 4.0 with this changes later.
For the Version 3.2-Rocket is in addition only the Money Management, with the lots number calculated as percentage of balance, without reference to a stop loss.
We can also calculate the lots as a percentage of the equity.

I am suggesting only ONE single martingale order in the very beginning of a potentially dead-lock cycle as described above - namely:

If the First Order has been triggered and subsequently price had reversed and the first subsequent order is triggered - which will be an OPPOSITE ORDER - namely, eg if the First Order is a Buy order, the first subsequent order would be a Sell order (it could not be another Buy order as the TP of the First (Buy) Order would have been hit well before the next Buy order is triggered if we had tweaked our settings as described above).

May I suggest that the following logic is coded into the EA in that if the first subsequent order is an OPPOSITE order of the First Order, the EA immediately implement a martingale OPPOSITE order with adjustable lot size (from my calculation it would need to be 3 times the First order) - in the event that the martingale order failed to kill off the 'cancer'/First Order (which will happen around 20% of the time from my observation from the backtest results) - then the EA will immediately revert back to the normal size lots and continue with the normal logic/strategy - in other words the EA only implement one single martingale order - and this should NOT in any way whatsoever, increase the risk of the basic logic/strategy. So, if the above is coded in the EA (and we have tweaked the settings as suggested above and assuming we have set the lot size to 1.0 lot), the EA will trade in the following sequence:

1, 3, 1, 1, 1, 1....etc

The first two orders would be OPPOSITE to one another...the following subsequent orders could be anything... The above would only happen in the event that the martingale order fails to kill off the First Order. If the martingale order is successful, we would only have two orders of 1, 3...and the whole cycle would have been closed in profit, hence, preventing a potentially deadlock cycle situation.

The above should NOT increase the risk factor in any manner whatsoever as by definition, a deadlock situation would result in almost a hedge-out position wherein the one single 3 lots order would make no difference whatsoever in as far as the risk factor is concerned. I hope you guys understand what I am trying to say here.

What do you guys think???

Ha ha ha Swingman, you had moved much faster than me (which is VERY good) - since you have already coded in money management, I am all for it - was just trying to save you some burdens from 'unnecessary' suggestions at this point in time but since you had already done the work - then I am all for it. Again, WELL DONE Swingman!!!

I always need some time to understand your logic ...
But there is an interesting idea for martingaling which could perhaps be combined with your entrys: https://www.forexfactory.com/showthr...8#post10982048

I also have to think about this logic, even though I analyzed the EA.

Need to go to work...will be a busy day for me today ...chat with you guys later.

In the mean time, please give us your feedback on my above suggestion.

I personally believe that with the filters now implemented and if the one single martingale order feature can be coded in as well, we should be able to further drastically reduce the incidence of deadlock situation.

Cheers guy!!!

I think the idea is good, because in the Rocket version you can see that the given TP with bigger lots is reached faster!

Last night I downloaded v 3.2 from post #449, and attached the EA to a GBPJPY,M15 chart with the settings displayed below (note: being too lazy to figure out what all of the settings are for, I simply used the same settings as in SwingMan's Strategy Tester report in post #466). As price fell the EA placed 3 sell orders at approx 10 pip intervals, and then as price reversed back upward to the middle of those 3 orders, it suddenly placed 59 buy orders at approx the same price within approx 15 seconds. At which point I immediately switched AutoTrading off to stop it from generating any more buy orders.

The good news is that GBPJPY rose overnight and I'm up around $2,900 on a 10k account -- but somehow I don't think it's supposed to be doing this.

I'm running Pepperstone demo build 1090 and Windows 7 Home Premium Service Pack 1.

If this strategy thrives through volatile periods would it make sense to filter trades and allow the EA to trade only when ADR reaches a certain #? Say for example - Only allow the EA to trade pairs that match ADR 160+ for the last 7 days.