sorry my English…

After one year dived total in forex and harvesting much information in this forum I also want to contribute:
Results of my EA (one of the thousands that I modified and until I created) of the year of 2006.
I feel myself happy because it had 3 bad months: (October, November and December) what it gave much fear to me… This already also earns in these.

Clearly that it is not total guarantee… but something is something…

Followed principles: In favor of having it the always underlying trend… and to all the processes.
Basic concern with the exits and not with the entrances. Very very important
And very small Lots. So small as that for 5000, one only 0.1 position.
As they know, this of the management of the capital, makes the difference.
And the ambition kills very this way…
I only make a relative exception, in the entrances of recovery of losses.

An idea of a participant of this fórum… Thanks.

I use the pointers iTrend_s and the Fisher_m11
EAs where I was felt inspired… Sincerely, in this height, mine it is not looked like none in particular…
Although it has not invented the wheel, clearly… I did not invent nothing

I find that the majority of good the EAs sins for ambition…
During much time, also I thought about a wage leaving of a capital of 5000… Me it does not seem possible.
But it has other forms… Appositive very in the EAs, therefore I am convinced the great principle that circulates in this way: without system, if it cannot be successful in forex. E I believe that for systems, the PCs is better that we.
Seeing my EA and I am not only made use to share the things and how much very some code… All we know the hunting to the other people's work that exists in the Internet… Already I know that that is arguable… Clearly that yes…

My idea to gain money, is, as already in this way, to rent the EA and to charge part of the profits that this to generate to the month. It seems me just and feasible.

The tests are, clearly, 90% and to I oppose it of the miracles that to the good times if place here also give very result in the tool of analyze Monte Carlo tool
http://www.mt4tools.com/Html/main.php?file=home.php

I am not satisfied still…
I will place the EA in ex4 here
that it will not function in live (safe engineering reversa eh eh) Risks…

Thanks and Good Entered and Better Exits… AF

Anos de 2000 a 2006


Tool:


resume:

ok you have EA and make backtest but any fowardtest?
Foward test on demo then on mini live account?
When you think EA works on demo on live account sometimes dont works or never works and i sugest to test it on small account(mini account)
I know many want EA for live trading but i really dont trust in EA and i never see EA to works on almost brokers.
but who knows maybe we will find it in future and for that we have a forum..

Have the best and i hope to see something new
Happy trading and many + pips

__________________
http://WWW.FOREXSTAGE.COM/
Your next stage to forex trading
Free 900 indicators
Free 900 Expert advisor

Thanks
My first mini, I lost it with an ambitious EA…

But it was a good learning…

Now I placed more $ and I go to open a standard one in the FXDD
That is I am much more confident today.

By the way, as it is known well: To make a mistake very and very fast is important

Will bring the results here. clearly…

Happy trading and many + pips

To test the EA i9-5M
It has 2 filing-cabinets:
EXE with setup
http://www.i9trader.com/publico/FX-setup-i9-5M.exe
ZIP p who pos security not to execute the EXE.
http://www.i9trader.com/publico/i9-5m.zip

the EXE is enough to exec and to select the folder of the Metatrader to use.
the ZIP, after decompress to lead for the Metatrader in the same folders.

It uses c/ PAIR EURUSD in M5. In the first tests, inputs for defect leaves.

For demos and backtesting the EA is free.
For real accounts it is necessary number it of account. If somebody wants to test in account mini of moment is gratis. In standard month is gratis 1.º month If it charges in following an equal percentage to the one of the profit gotten until a 45% maximum.

For minis IBFX a adjustment is necessary still.

I know that demon and Real are not the same. E less bactesting.
But tb is truth that serves to give a idea of the EA.
And is sufficiently good. the idea.
Because I put a EA here that I think to be of success?
Because the market varies immense and the opinions can enrich the EA.

2006

This is my entrance favorite that use in the I9-5M. He is of many of the EAs here (Terminator for example).


int fTrendRsiSto()
{
int myOrderType =0;
double Sinal1_1 = iCustom(NULL, 0, "i9_Trend", 0, 0, 0, 20, 2, 13, 300, 0, 0);
double Sinal1_2 = iCustom(NULL, 0, "i9_Trend", 0, 0, 0, 20, 2, 13, 300, 1, 0);
double Sinal2_1 = iCustom(NULL, 0, "i9_Trend", 0, 0, 0, 20, 2, 13, 300, 0, 0);
double Sinal2_2 = iCustom(NULL, 0, "i9_Trend", 0, 0, 0, 20, 2, 13, 300, 0, 1);
double Sinal3_1 = iStochastic(NULL, 0, 8, 3, 4, MODE_EMA, 1, MODE_MAIN, 0);
double Sinal3_2 = iStochastic(NULL, 0, 8, 3, 4, MODE_EMA, 1, MODE_SIGNAL, 0);
double Sinal4_1 = iStochastic(NULL, 0, 8, 3, 4, MODE_EMA, 1, MODE_MAIN, 0);
double Sinal4_2 = iStochastic(NULL, 0, 8, 3, 4, MODE_EMA, 1, MODE_MAIN, 1);
double Sinal5_1 = iStochastic(NULL, 0, 8, 3, 4, MODE_EMA, 1, MODE_MAIN, 0);
double Sinal5_2 = 70;
double Sinal6_2 = 20;
double Sinal7_1 = iRSI(NULL, 0, 5, PRICE_CLOSE, 0);
double Sinal7_2 = iRSI(NULL, 0, 5, PRICE_CLOSE, 1);

if (Sinal1_1 > Sinal1_2 && Sinal2_1 > Sinal2_2 && Sinal3_1 > Sinal3_2 &&
Sinal4_1 > Sinal4_2 && (Sinal5_1 > Sinal5_2 || Sinal5_1 < Sinal6_2) &&
Sinal7_1 > Sinal7_2) { myOrderType = 1; } // Sell

if (Sinal1_1 < Sinal1_2 && Sinal2_1 < Sinal2_2 && Sinal3_1 < Sinal3_2 &&
Sinal4_1 < Sinal4_2 && (Sinal5_1 > Sinal5_2 || Sinal5_1 < Sinal6_2) &&
Sinal7_1 < Sinal7_2) { myOrderType = 2; } // Buy
return(myOrderType);

}



It is optimized in relation to the original below that it has code repetitions.


int OpenOrdersBasedOni_TrendRSISto()
{
int myOrderType=3;
double Buy1_1 = iCustom(NULL, 0, "i_Trend", 0, 0, 0, 20, 2, 13, 300, 0, 0);
double Buy1_2 = iCustom(NULL, 0, "i_Trend", 0, 0, 0, 20, 2, 13, 300, 1, 0);
double Buy2_1 = iCustom(NULL, 0, "i_Trend", 0, 0, 0, 20, 2, 13, 300, 0, 0);
double Buy2_2 = iCustom(NULL, 0, "i_Trend", 0, 0, 0, 20, 2, 13, 300, 0, 1);
double Buy3_1 = iStochastic(NULL, 0, 8, 3, 4, MODE_EMA, 1, MODE_MAIN, 0);
double Buy3_2 = iStochastic(NULL, 0, 8, 3, 4, MODE_EMA, 1, MODE_SIGNAL, 0);
double Buy4_1 = iStochastic(NULL, 0, 8, 3, 4, MODE_EMA, 1, MODE_MAIN, 0);
double Buy4_2 = iStochastic(NULL, 0, 8, 3, 4, MODE_EMA, 1, MODE_MAIN, 1);
double Buy5_1 = iStochastic(NULL, 0, 8, 3, 4, MODE_EMA, 1, MODE_MAIN, 0);
double Buy5_2 = 80;
double Buy6_1 = iStochastic(NULL, 0, 8, 3, 4, MODE_EMA, 1, MODE_MAIN, 0);
double Buy6_2 = 20;
double Buy7_1 = iRSI(NULL, 0, 5, PRICE_CLOSE, 0);
double Buy7_2 = iRSI(NULL, 0, 5, PRICE_CLOSE, 1);
double Sell1_1 = iCustom(NULL, 0, "i_Trend", 0, 0, 0, 20, 2, 13, 300, 0, 0);
double Sell1_2 = iCustom(NULL, 0, "i_Trend", 0, 0, 0, 20, 2, 13, 300, 1, 0);
double Sell2_1 = iCustom(NULL, 0, "i_Trend", 0, 0, 0, 20, 2, 13, 300, 0, 0);
double Sell2_2 = iCustom(NULL, 0, "i_Trend", 0, 0, 0, 20, 2, 13, 300, 0, 1);
double Sell3_1 = iStochastic(NULL, 0, 8, 3, 4, MODE_EMA, 1, MODE_MAIN, 0);
double Sell3_2 = iStochastic(NULL, 0, 8, 3, 4, MODE_EMA, 1, MODE_SIGNAL, 0);
double Sell4_1 = iStochastic(NULL, 0, 8, 3, 4, MODE_EMA, 1, MODE_MAIN, 0);
double Sell4_2 = iStochastic(NULL, 0, 8, 3, 4, MODE_EMA, 1, MODE_MAIN, 1);
double Sell5_1 = iStochastic(NULL, 0, 8, 3, 4, MODE_EMA, 1, MODE_MAIN, 0);
double Sell5_2 = 80;
double Sell6_1 = iStochastic(NULL, 0, 8, 3, 4, MODE_EMA, 1, MODE_MAIN, 0);
double Sell6_2 = 30;
double Sell7_1 = iRSI(NULL, 0, 5, PRICE_CLOSE, 0);
double Sell7_2 = iRSI(NULL, 0, 5, PRICE_CLOSE, 1);

if (Buy1_1 > Buy1_2 && Buy2_1 > Buy2_2 && Buy3_1 > Buy3_2 && Buy4_1 > Buy4_2 && Buy5_1 < Buy5_2 && Buy6_1 > Buy6_2 && Buy7_1 > Buy7_2) { myOrderType = 2; } // BUY
if (Sell1_1 < Sell1_2 && Sell2_1 < Sell2_2 && Sell3_1 < Sell3_2 && Sell4_1 < Sell4_2 && Sell5_1 < Sell5_2 && Sell6_1 > Sell6_2 && Sell7_1 < Sell7_2) { myOrderType = 1; } // SELL
return(myOrderType);

}

But this is not important.
What I want to point out is that use changing BUY/SELL that becomes it much more effective for a short strategy (scalp ?).



and changing BUY/SELL:



i9-5m.zip

afernandes

After much evaluation I have come to the conclusion that 5 min and higher timeframe backtests on MT4 are very inaccurate, even with 90% modelling quality -

i.e M5, M15, M30, H1, H4, D1, W1 and MN have a tendency to produce highly inaccurate backtests even with 90% modelling quality!

Try backtesting your system on the 1min timeframe. If it produces good results here then the chances are that you may have a good system.

I don't think the picture of the backtest is realistic at all due to the following:

1. The balance line is very linear in its appreciation.

2. There is litle deviation from the balance and the equity line. If you analyse Martinage sytle EAs in forward or live testing you would know that as your lots start increasing then your equity starts decreasing. In the MT4 backtester the Balance line is the blue line and the Equity line is green. We can see no green line in your backtest picture. I am not saying that this is your fault, just pointing out a flaw in the mt4 backtester.


Do you have any forward testing results?

Attached Images

2006-i9-5M.jpg (116.7 KB, 321 views)

Hello and
Thanks a lot anticipated...

Question: scalp is when closes the orders in short times all the time? Or when this is not systematic if it can also consider scalp?

In my EA it close when I reach pips. The minimum of broker when the Trend is contrary… and this causes that it has orders with 3 times of mn, 2 mn, and up to 1 mn. But this is not strategy… (This I say more below…)
I like this broker, http://www.fxclearing.ca/eng/ but as is announced clearly the opposite to scalp…
Thanks…

http://www.i9trader.com/publico/h4c profitloss.htm

My strategy…

In the ‘combat to `scalp' step of 5M to H4….
Well, but the strategy is similar:
I open one alone position.
Latch reaching objectives of PIPs.
If trend is favorable, 9 or 12
If it is not, the minimum of broker (MarketInfo (Symbol (), MODE_STOPLEVEL))
If the position enters significant loss, I look to a situation some indicators where are in agreement for opening a strengthened position and closing loss.
If this situation delays one hour more than, allows to open one 2.º position, not to waste the time that is to pass. As soon as it obtains to close the Loss, I come back to allow one alone position.

This to have the risk edge very low. With the risk in 0.03, a time that in H4 if few positions open, I obtain a reasonable profit

Well..
The tests are of backtesting, but the EA in demons already demonstrated that it is reasonable.
It lacks the test of the 9 of live that it goes to still enter this month.
My previous mini, I lost for placing an aggressive EA.
Not this EA.
EA: setup or zip

Thanks and good profits...
AF