Great EA in backtest! - page 51

 

try not using auto lot. And set a min amount.

real account on mine affects the auto lot function.

 

also I cann't get sept-dec 05 and 04 to test good with new settings.

big problem is news. But its also an active time with the market. Lots of large moves.

still testing

Dave

 
whitesand:
Aragorn, Your backtest that you ran is with IBFX too like your forward test?

yes it is from the demo acct.

 
cko:
From CD forum, any comments?

Bad results with Meta-Trader v.4.00 - build 1.96

Submitted by davidfx108 on Tue, 2006-09-26 18:05.

"MetaTrader build 196 or higher have BUGS with Cyberia Trader

MT 196 or higher have BUGS with Cyberia Trader:

i am running Cyberia with 4 pair, 3 time frame in 3 broker concorence for FORWARD testing, about 6 week.

when i ugrade MetaTraders from 195 to 196 and then to 197

i have many bad results !!!!!!! in two last week !!!!!"

well that's just peachy considering that I was running build 195 up until today when I decided to try this live and decided to update to build 197 first.

 
Wackena:
Aaragorn,

error 131 for some brokers is lot size below 0.1. I think most brokers for standard accounts have minimum of 0.1 lots. You can test by setting AutoLots=false.

Wackena

I am testing this on my live mini acct. When I change maxlots = 10 I don't have the error any more. I think I'm only allowed 10 lots max on this mini account.

 
xxDavidxSxx:
also I cann't get sept-dec 05 and 04 to test good with new settings.

big problem is news. But its also an active time with the market. Lots of large moves.

still testing

Dave

Hey Dave...

I just ran this with extended alpari data I have on my live account backtester...

My live account only allows 10 lots max so that is what this is. the numbers aren't as smashing but I can't argue with the direction of that growth. Even thru the date ranges you mention. What are you seeing? I do notice the drawdowns are now in the 70% range but that has to be from the period before Aug. 2004. The slope of that line after Aug. 2004 just doesn't have massive drawdowns in it! Least not as I can see anyway. I guess I could do it again starting at that point and see what the drawdowns are after Aug 2004 to now.

 

well I am looking at my actual mini account which only has a little less than $400 in it.

This backtest with maxlots=3 managed to survive the month of august when it started with that amount. However when I did the maxlots=5 it went bankrupt by the time it got to sept of 2004. Based on that I don't think I'm going to allow my live account to trade above maxlots=2 for the time being. Looks like on first glance I can expect an average of 4 trade/day give or take a few.

This is the first EA I will have let touch real money since about last may/june approx.

 

What other pairs besides gbpusd have you tried this on and with what results?

 

Ok Here I am putting what I said back again. I took it out because I did not want to get into an argument.

What makes a BIG difference in Backtests: SWAP is taken from the last signed-in account. Keep in mind that swaps change all the time. This also holds true for some other variables.

Solution: Don't sign in again, once you have your data stablished. This will give you consistent backtest with each other.

To confirm the above: Check the value of swaps by checking the currency's property. Two different places to do it. One is on "strategy builder" and the other from the "Market watch". Normally they are the same, but if you sign-in/out of different brokers then it is different. Make sure it is the same, so to run a backtest, sign-in to another broker with different swaps or just wait a 1 to 5 days and run the backtest again with the new swaps. Yes you will see.

Also Aaragorn: To keep your big data-> From the Menu Tools->Option->Chart and change the "Max Bars in History" to 9999999999 (mean all 9s). Next time you come back it will be 2147483647. It is ok, this is really the max, just easier to put 9s.

 
Aaragorn:
Hey Dave...

I just ran this with extended alpari data I have on my live account backtester...

My live account only allows 10 lots max so that is what this is. the numbers aren't as smashing but I can't argue with the direction of that growth. Even thru the date ranges you mention. What are you seeing? I do notice the drawdowns are now in the 70% range but that has to be from the period before Aug. 2004. The slope of that line after Aug. 2004 just doesn't have massive drawdowns in it! Least not as I can see anyway. I guess I could do it again starting at that point and see what the drawdowns are after Aug 2004 to now.

thats the draw down I was seeing. Its not as bad in 05 but sept,oct of 05 wasn't real friendly either. I noticed there was simmiler price movement as this year. Looking hard at sept-dec of 04 and 05 to get an idea of what this time of year has to offer. Lots of news and lot of sharp moves. If I knew it was all news that was drawing me down I'd feel better. I have 9 news times taken out of my tests. But there others scattered out too. Over these months the account isn't getting out and staying out of the 500-6oo$ range. So I won't go with the agressive settings yet. But in slower market after mid Dec. I might.

I am not using more than .1 lot on real account untill it gets over 1k. Its to easy to lose 500$. If it draws down to 250 that gets scary. I will add some of my manual trades to the test account to get ballance up over 1k faster.

My Dad is gonna open a 500$ account too and let me run CT on it too. And he'll add to it as well as my manual trades to get it above 1k. ones above 1k I'll set auto lot to start with .2 and just let it go.

The returns arn't as impressive as the demo tests, but that was just dreaming. The returns on real account tests are more realistic. But consistant returns 10-20% a month are good #'s.

gbp, jpy havn't given me good tests. I am still gonne try though more tests.

I am considering running agressive settings on sunday for a few hrs to see what happens.

Dave

Reason: