V1andV2 Hedged EA: Beautiful Equity Curve - Page 22

icgremlins · #**211** (**permalink**) 05-09-2007, 01:44 AM

I have been following the thread and have just begun to forward test the new version v1+v2 with out starting a debate i believe forward test's are the only way to really prove a winner and my opinion is based only on my experience

P.S. My Posts Total does not reflect my experieince

Pinggarcia · #**212** (**permalink**) 05-09-2007, 01:49 AM

Very interesting thread. I just stumbled on it today.

But because there are so many versions, and revisions, maybe the starter of the thread or any mods here can transfer all UPDATED versions on page 1. And make some short comments on it.

Just a thought.

darxtar · #**213** (**permalink**) 05-09-2007, 02:31 AM

Hi Monika,

Thanks for the tips. I've just tried backtest it using IBFX meta trader and still get no result. I use EURUSD 5m chart using the latest version submitted today by Gbolla.

I also don't understand what are each input parameters for and how to change the input parameters.
I've read all the post in this thread and found no detail information about how to backtest and how to set the parameter for this EA. Also that when people talking about setting the Maxtrade, I can't find the parameter in the EA.

I believe we need to understand how to set the parameters prior run it live. Otherwise it would be similar to gambling, since we don't know what we are doing. :P

Anyone can help with these information would be greatly appreciated. Realy.
I'm an IT Manager in Citibank. If I can understand and test it properly, perhaps I can contribute to enhance the EA better.

BTW, thanks to everyone who contributes on this EA. It looks promising, but I still need time to understand how it works, how to tune it, and how to test it. After read all the posts, What I get currently is still only the big picture about how it work. Still totally confuse how to get it tested and work properly.

Thanks

JacK

Quote:

Originally Posted by monik7

Darxtar,

You need brokers allowing 0.01 lots to run the mini version, otherwise it won´t work. I´m sure FXDD does not bring that possibility in demo account.
Pls check Alpari, I don´t know that.
Regards

mikejody · #**214** (**permalink**) 05-09-2007, 04:22 AM

I will do so as soon as the V1andV2MiniNew is created.

Quote:

Originally Posted by Pinggarcia

Very interesting thread. I just stumbled on it today.

But because there are so many versions, and revisions, maybe the starter of the thread or any mods here can transfer all UPDATED versions on page 1. And make some short comments on it.

Just a thought.

darxtar · #**215** (**permalink**) 05-09-2007, 07:05 AM

I think I have figure out why I've got no test result so far. It was caused by the parameters was not set correctly.

Now, how to set the parameters correctly would be the next challenge.
Anybody willing to explain how to?

The information posted by Gbolla was helpful but I still unable to set the optimum parameters.

Will do some testing on EURUSD

JacK

gbolla · #**216** (**permalink**) 05-09-2007, 08:29 AM

Hi ladies and gentleman, in attach fix version with a new features.

1) "Pips_Shift": now we can decide if we need a shift of TakeProfit and Distance between two orders for cover better long trend. If you choose 0 don't use shift.(es.: TakeProfit=10 and D=10.......if we use Pips_Shift=1, second order will be TP=11 and D=11, third TP=12 and D=12...and so on)

2) Fix: fix the problem about close open buy order at swith on the EA.

Guys, I have made my job.......now you made yours......hard optimization and backtest!!!
Good luck
Ciao

Bolla

Pecunia non olet · #**217** (**permalink**) 05-09-2007, 10:10 AM

Quote:

Originally Posted by alpha24seven

....I think that an anti-martingale strategy would be more appropriate. That is, you increase your lots after a win and not a loss.

...I've been reading these martingale threads thinking that an anti-martingale strategy might be complementary to the martingales, in a portfolio.

In theory the anti-martingale would do well in trends, and do a better job of truly hedging the martingale positions.

In this scenario I think the challenge would be to scale the anti-martingale so it's not eating too much margin at just the time that the martingale positions are in deep drawdown. The max lotsize of the anti-martingale EA would need to be in a sort of equilibrium with the open drawdown of the martingale.

The idea of just bailing out of too-deep martingale drawdowns mentioned by Davidke in the 10 Points 3 thread probably has an important place in this scenario as well.

In other words: hedged anti-martingale + hedged martingale w/ stoploss in the same account.

Quote:

Originally Posted by alpha24seven

With trade dependency sequence analysis you can analyze trade dependency and create a call to a calculation that will modify your position on the fly. For example, if the system is analyzed to create a sequence of three consecutive wins and then followed by four consecutive losses -- you have a calculation basis on how and where to dynamically modify (trade to trade) position size. It's called positive and negative dependency. It's a moving target but that is dynamic also.

http://72.14.253.104/search?q=cache:...lnk&cd=2&gl=us
http://www.adaptrade.com/depend.htm
http://www.adaptrade.com/Articles/article-dep.htm

- - - -

A proprietary anti-martingale EA that I've seen makes me think that more primitive methods of exploiting positive dependency are also useable.

I think the main factor for the anti-martingale is getting enough positive expectancy in the entry/exit logic to get the geometric equity growth to fully hedge the martingale (while working with lotsizes that complement the martingale).

It is indeed "a moving target."

gmax111 · #**218** (**permalink**) 05-09-2007, 10:33 AM

Quote:

Originally Posted by gbolla

Hi ladies and gentleman, in attach fix version with a new features.

1) "Pips_Shift": now we can decide if we need a shift of TakeProfit and Distance between two orders for cover better long trend. If you choose 0 don't use shift.(es.: TakeProfit=10 and D=10.......if we use Pips_Shift=1, second order will be TP=11 and D=11, third TP=12 and D=12...and so on)

2) Fix: fix the problem about close open buy order at swith on the EA.

Guys, I have made my job.......now you made yours......hard optimization and backtest!!!
Good luck
Ciao

Bolla

I've backtested and backtested and for some reason groups of trades are not being closed resulting in a margin call...

Symbol EURUSD (Euro vs US Dollar)
Period 15 Minutes (M15) 2005.01.03 00:00 - 2007.05.03 03:45 (2005.01.01 - 2007.05.08)
Model Every tick (based on all available least timeframes with fractal interpolation of every tick)
Parameters Name_Expert="V1+V2"; TakeProfitLong=45; TakeProfitShort=45; DLong=20; DShort=20; TrailingStopLong=0; TrailingStopShort=0; Balance=10000; Pct_Lot=3; Pct_Equity=5; Pips_ShiftL=1; Pips_ShiftS=1; Indicator=""; H_level=0; L_level=0; MultiplierL=2; MultiplierS=1.5; Slippage=3; UseSound=false; NameFileSound="alert.wav";

gbudy · #**219** (**permalink**) 05-09-2007, 10:46 AM

This EA does not trade on my live account, why?

Quote:

Originally Posted by gbolla

Hi ladies and gentleman, in attach fix version with a new features.

1) "Pips_Shift": now we can decide if we need a shift of TakeProfit and Distance between two orders for cover better long trend. If you choose 0 don't use shift.(es.: TakeProfit=10 and D=10.......if we use Pips_Shift=1, second order will be TP=11 and D=11, third TP=12 and D=12...and so on)

2) Fix: fix the problem about close open buy order at swith on the EA.

Guys, I have made my job.......now you made yours......hard optimization and backtest!!!
Good luck
Ciao

Bolla

ralph.ronnquist · #**220** (**permalink**) 05-09-2007, 10:56 AM

If you change the first line in function "closeAllOrders" to be

PHP Code:


			
for ( int x = OrdersTotal() - 1; x >= 0 ;x-- )

instead of

PHP Code:


			
for(int x=0;x<OrdersTotal();x++)

as it currently is, then the EA will be more diligent when closing orders.