Nice backtest

Attached Images

pipmaker.gif (21.4 KB, 611 views)

__________________
RUN With The BULLS And HUNT With The BEARS - Nothing By Chance

But not so nice DD

I always try to get lowest DD, not higher profit.

You're welcome.

Backtest speed is not entirely dependent on EA tested, also count PC speed, available memory, amount of data to be tested..
Usually EAs using external indicators will run slower and also some indicators require more RAM and CPU cycles than others.

Enforcer, just replace ARSI_trigger= 0.007 with this

ARSI_trigger = iATR(NULL,1,13, 0)*10;

You`ll stay satisfied - i`ve got 40% more profit. The DD is increased a litle. ATR bars=13 is the best, I start making tests one week ago. This is the way to grab only unusual strong impulses compared with current volatility.

Also I`ve tried with separate trailing for Buy`s and Sells. I`ve got nice increased profit about another 5%, but DD looks unacceptable.

The other thing I`ve tryed is using i_Trend, but I have changed the i_TrendDecision (if = 1 becomes = -1 and oposite if = -1 becomes = 1). The idea is simple: i_Trend is looking for main trend in H4 or H1. Same time ARSI in M1 is looking for strong impulses against the i_Trend, but the orders will be placed in the direction of the main H4 (H1) trend because ReverseDirection = true. In this way PipMaker becomes main_trend follow robot and the orders are placed in the best posible place - in the bottom of the chanel if upgoing_trend exist and in the top of the chanel if down _going trend exist. The number of the trades is much less but we`ve got DD=4,21% so we can play with more risk. I post some result from my modified versions using i_Trend.
The DD is 38.94% but note that I play with MaximumRisk=1 and profit is $4084
until end of april. Without i_Trend I`ve got profit about 15`000 but DD are still unacceptable.
I`m still testing...

Attached Images

	PM_modified_for_tests_ATR_and_iTrend.GIF (31.5 KB, 581 views)
	PM_2_modified_no_iTrend.GIF (40.8 KB, 574 views)

@montedoro

please can you post the new mq4?

This is very strange indeed, as I'm running the backtest on some rather highend servers, using indicators that I have tested in other EAs that ran very quickly.

Anyone else noticing the fact that this EA takes a substantial long time to backtest? I wonder if it's just me.

Thanks Enforcer.

-dgv

__________________
Metatrader 4 - EA Live and Demo Hosting Solutions http://www.omegasupreme.com

montedoro, I have some questions...
You wrote
Its true.

But
This is not the case.
If i_Trend signals UP and ReverseDirection = true then EA will take the SELL position, because ReverseDirection = true! (if we take into account only i_Trend)
Here is the code:
So what you have done I don't fully understand.

As I have understood you've done something like this:
if( i_Trend(60 min)==UP && ARSI(1 min) == DOWN ) then SELL
AND
if( i_Trend(60 min)==DOWN && ARSI(1 min) == UP ) then BUY

Am I right?

Thanks.

Didn't had time to look into this but at first look may be a contradiction in indicators signals.
Better post your version to have overall view of code.

Lemyx, I`ll explain again: The order must be placed in the direction of the main trend. So

if( i_Trend(60 min)==UP && ARSI(1 min) == DOWN ) then BUY
AND
if( i_Trend(60 min)==DOWN && ARSI(1 min) == UP ) then SELL

The easy way to make it just for test : change iTrendDecision() each other

if (B1_1 > B1_2 && B2_1 > B2_2) {tradeDirection=-1;} // it was =1
if (S1_1 < S1_2 && S2_1 < S2_2) {tradeDirection= 1;} // it was =-1


I mean we are looking for best position in the H1_trend_channel to place our order. The best result we can expect in pairs usualy in trend and we can play with more risk. The test with GBPUSD was not good example. I`ll test next week with other pairs.

erdenmensch, my changes are not a secret, but I have about 10 raw modified versions of PipMaker just for tests. They will be useless for you. Only I know all the mess I have done with the code inside (actualy my friend did it by my order, I`m not programer and he is already hiding from me because I`m eating all his free time. Untill now the only one sure result I`ve got is with the floating ARSI_trigger. The value of ARSI is rising faster than ATR so this is great way to grab the righ impuls of the price. If Enforcer agree with my tests I expect he to add this in the next versions.

If I'll get positive results, I'll add changes, sure.