Drastically different results with 2 different brokers.

New comment
 
I was just wondering if it is normal for an EA to perform differently with 2 different brokers? I have been studying forex and learning for a while now, and have done several different demos. Not until recently have i had any success with an EA. I had my EA set up good with a broker on a demo account with 98.8% winning profits, profit to loss ratio in dollars is 15/1, and .14% drawdown. So i decided to open a nano micro account with this broker just to run my EA live with chump change. I have to say it is doing well, although not enough time has passed to say it is perfoming up to the demo backtest results. Anyways, i had run across some info on my broker. The information is not good. I have seen many a reviews saying that the broker i am using has stophunted in the past. I have seen several legal issues with this broker on the web. Some dude has made an EA just for this broker that pings there server to prove some bad things about this broker when they do there shady stuff. So i searched for a new broker and made a new demo account. All 3 of my successful EA's that performed well on my previous account do the complete opposite on the new demo account. Is this normal or is my EA tuned to this shady broker which is drastically different then the others?
 

An EA's performance will vary across the live feeds of different brokers

In general, EA's working purely on support & resistance will work with the least variation across brokers

Breakout EA's transfer quite well, especially if on H1 and above timeframe

Indicator-based EA's vary the most, from great profit to great loss between brokers

Also an EA entering mid-bar will show more variation than one entering on close of a bar

The same general points apply to some extent when moving from a demo account to a live account of the same broker

So...

As ever, lets be careful in the Forex

FWIW

-BB-

 
Thanks, i am thankfull for your knowledge. Seems true when my EA is indicator based. I will have to learn to set up an EA that works with in support/resistence, breakout, and indicator based. That would probably create less trades. Well, back to the drawing board. Are there any EA's in this website that are based on support/resistence that i can study and learn from?
 

Wouldn't a EA running well with a bad broker be considered better than an EA running badly with a good broker? So perhaps you don't need to change brokers? :)


But back to your point - yes, the same EA can perform quite differently on accounts from different brokers. Try a few more different brokers and they are likely to be all different. And it means that your EA is finely tuned to the bad broker - which makes the EA dangerous, as things can and will change.


As well, EAs can perform to a particular pattern consistently for a certain period, and then change behavior completely after a certain point - all on the same account from the same broker. Something to do with fitting the curve nicely, but then the curve changed shape.

 

Support & Resistance lines are so easy to do by drawing on the chart but are harder to do with code

See 'Construction of Fractal Lines' for an example

and SHI is brilliant - just watch for the Slope getting too great which would indicate a trend developing

Good Luck

-BB-

PS

My first reply assumes that if your EA uses Start and End Hours for trading, these are changed if the brokers timezone differs

 
blogzr3 wrote >>

Wouldn't a EA running well with a bad broker be considered better than an EA running badly with a good broker? So perhaps you don't need to change brokers? :)


But back to your point - yes, the same EA can perform quite differently on accounts from different brokers. Try a few more different brokers and they are likely to be all different. And it means that your EA is finely tuned to the bad broker - which makes the EA dangerous, as things can and will change.

As well, EAs can perform to a particular pattern consistently for a certain period, and then change behavior completely after a certain point - all on the same account from the same broker. Something to do with fitting the curve nicely, but then the curve changed shape.

Well your 1st point is something i had put some thought into. In my backtesting with the bad broker this EA performs consistently over the year it has in history files which is the previous year. I had also done a back test where it turned a $60,000 profit with $10,000 start up on an aggressive money management in only one month. I am considering saying to hell with them and let it run balls to the wall. I am scared of the possibility of them locking my account. Then again i had put $250 into this account for EA testing just to test live, so I have no attachment to the amount. Plus the balance had dwindled a bit with my previous EA's. With my new EA's it is growing with conservative lots. The balance isnt enough to worry about at the moment. What do you think? Should I believe in my EA and let her rip with this crappy broker because im not even gonna try to withdraw the money as it is now? If i do and it all turns out well it would make for good education studying this.

 
BarrowBoy wrote >>

Support & Resistance lines are so easy to do by drawing on the chart but are harder to do with code

See 'Construction of Fractal Lines' for an example

and SHI is brilliant - just watch for the Slope getting too great which would indicate a trend developing

Good Luck

-BB-

PS

My first reply assumes that if your EA uses Start and End Hours for trading, these are changed if the brokers timezone differs

I can already see its very difficult but my eagerness to learn this is driving me more then anything has ever inspired me. I am military and looking for a way to do good with my savings but not gonna jump in fully with my bank account till i have totally educated myself. This stuff is tuff but im not backing down. That EA i speak of took me 18 straight hours from start to finish. I was motivated and determined, didnt stop till i was satisfied.

 

> let her rip?

Yep go for it - its all good experience :)

Spotting the differences in trading rhythm and equity curve between demo & live is invaluable education

Under any circumstances it can easily cost $1000+ to assess a brokers unsuitability for an EA or class of EA...

-BB-

 

Well being that i had wrote off the money anyways cause its there for live testing and its not my total assets, i will opener her up and let it spin the tires. When i get off work tonight i will crank up the RPMs and see what happens. On a side note, with this EA running on 4 pairs, and 3 other EA's running on several pairs, the account has doubled in a week. Like i said though the balance was extremely low from my previous testing before the past week..

 

Update: I am running multple EA's on this live account. The account in total has grew 40%, profit/loss ratio is 4/1 in dollars, 87% profit wins, in only 5 days. The Limited martingale system is working real good here.

New comment