Same strategy for different accounts. Can anyone code that?

 

Hi - what I want to do is in the below search:

https://gemini.google.com/share/2473aceac1e9

I want an expert advisor which trades different accounts with the same strategy. Same risk for all of them. 

If someone codes such an EA, how can I backtest hundreds of different periods automatically without involving ourselves?

Thank you.

 
Ed:


I want an expert advisor which trades different accounts with the same strategy. Same risk for all of them. 


You need to elaborate on this. if you mean "same strategy" being as "same ea", then yes it is possible with a copy trading ea. This copy trading ea would be coded to adjust the lot to suite different margin requirements of each account.

see the freelance site to open an order to create your trading ea for a small cost; and there is several free copy trading eas on codebase that may suite your needs; and other copy trade eas for free on google.

Ed:

If someone codes such an EA, how can I backtest hundreds of different periods automatically without involving ourselves?


yes, but that would need a big computer server with 100gbs of ram and 1 core for each period you want to test. VERY EXPENSIVE TO SETUP.

 
Michael Charles Schefe #:

You need to elaborate on this. if you mean "same strategy" being as "same ea", then yes it is possible with a copy trading ea. This copy trading ea would be coded to adjust the lot to suite different margin requirements of each account.

see the freelance site to open an order to create your trading ea for a small cost; and there is several free copy trading eas on codebase that may suite your needs; and other copy trade eas for free on google.

yes, but that would need a big computer server with 100gbs of ram and 1 core for each period you want to test. VERY EXPENSIVE TO SETUP.

Do you believe such a system is more efficient? Splitting the EA with different accounts?
 
Ed #:
Do you believe such a system is more efficient? Splitting the EA with different accounts?

In your OP, you mention "hundreds of different periods" which are timeframes in MT5. The maximum number of charts that MT5 can run is 100. Assuming that you're not trying to circumvent anti-hedging rules (your AI chat mentions hedging), using different accounts simply complicates accounting because you'll have separate account balances to balance. However, you will need multiple instances/installations of MT5 to backtest all of those chart periods simultaneously. Either way, you'll need a trade copier that is capable of running in the MT5 Tester.

As Michael already said, the setup will be pricey. You could go with a local beastly server workstation (150,000 USD or so) and locally copy the trades across multiple MT5 instances, or you could go with multiple VPS's (just imagine the monthly fees) and remotely copy the trades across the MT5 instances there, or you could use a combination of both.

 
Ed #:
Splitting the EA with different accounts?

once again, i am unsure of your meaning. need more details in your questions.

But it is ALWAYS more efficient to add a signal ea to charts of 3 accounts, than to copy trades from 1 account to the other 2 due to copying trades between accounts almost always creates slippage and/or lag that also often results in the 2 slave accounts opening trades at different prices than the 1st account. This difference in opening prices can be significant on many pairs, moreso if the market session is of "high volatility". Hopefully i have understood your meaning of your question?

 
Michael Charles Schefe #:

once again, i am unsure of your meaning. need more details in your questions.

But it is ALWAYS more efficient to add a signal ea to charts of 3 accounts, than to copy trades from 1 account to the other 2 due to copying trades between accounts almost always creates slippage and/or lag that also often results in the 2 slave accounts opening trades at different prices than the 1st account. This difference in opening prices can be significant on many pairs, moreso if the market session is of "high volatility". Hopefully i have understood your meaning of your question?

If by saying "signal ea" you mean trading EA, then I get what you mean. Yes, avoiding additional network communication on top of communication with the trade server is usually optimal. In the OP's case, cross-account hedging might be more nuanced. If each EA on an account has to hedge the other EA's trades on other accounts, then all EA's must be "aware" of the other EAs' trades in the other accounts. As only one account can be traded at once in each MT5 instance, additional network communication is required.

Of course, local copying is always going to be faster than remote copying. As long we're all pretending to be willing to sling money like it's organic fertilizer, I'd prefer the local server workstation setup.

 
Ryan L Johnson #:

 As long we're all pretending to be willing to sling money like it's organic fertilizer, I'd prefer the local server workstation setup.

me too -- if only to be able to boast having such a setup!

Ed #:
Do you believe such a system is more efficient? Splitting the EA with different accounts?

This would depend on how many trades that you think that you will have open. As Ryan said, in the Gemini page it mentions hedging; which often means many trades. The more trades you have open, the bigger lag and network/www connection issues can occur.

then, it would be the communication between the accounts and the broker that will be the bottleneck, 3 vps's at different geolocations would be something else to consider which will have total of 3 connections with the www -- to avoid many different issues that can be caused by this style of trading thru a single connection to the broker. This may also be cheaper option for trading, but you wont be able to backtest.

I think if you want more help then we need to get more details of what you are thinking. I recommend that you write it out on notepad, digital or paper before posting. and rewrite it 3 times, each time expanding on each point in each rewrite. Start with a description of why you think that you will need 3 trading accounts.

 
Thank you to all of you!...
 
Michael Charles Schefe #:

me too -- if only to be able to boast having such a setup!

This would depend on how many trades that you think that you will have open. As Ryan said, in the Gemini page it mentions hedging; which often means many trades. The more trades you have open, the bigger lag and network/www connection issues can occur.

then, it would be the communication between the accounts and the broker that will be the bottleneck, 3 vps's at different geolocations would be something else to consider which will have total of 3 connections with the www -- to avoid many different issues that can be caused by this style of trading thru a single connection to the broker. This may also be cheaper option for trading, but you wont be able to backtest.

I think if you want more help then we need to get more details of what you are thinking. I recommend that you write it out on notepad, digital or paper before posting. and rewrite it 3 times, each time expanding on each point in each rewrite. Start with a description of why you think that you will need 3 trading accounts.

To explain what I exactly I want: the gaps between the accounts will be large. So - a sudden movement in prices will not be loss until we reach the new account - we need to reach the new account after long price gaps. The accounts will not be immediately one another. To reach the new account there should be a distance between them. Not immediately one after another. So it won't be same with single account. I guess it sounds like a good strategy? We can even use grid with that system?
 
Ed #:
To explain what I exactly I want: the gaps between the accounts will be large. So - a sudden movement in prices will not be loss until we reach the new account - we need to reach the new account after long price gaps. The accounts will not be immediately one another. To reach the new account there should be a distance between them. Not immediately one after another. So it won't be same with single account. I guess it sounds like a good strategy? We can even use grid with that system?

I think there seems to be a linguistic misunderstanding. What on earth do you mean by "account"?

For us, an "account" is a trading account offered by a broker. It is simply where one places our funds for trading. There is no "price" or "gaps" in trading accounts.

 
Ed #:
To explain what I exactly I want: the gaps between the accounts will be large. So - a sudden movement in prices will not be loss until we reach the new account - we need to reach the new account after long price gaps. The accounts will not be immediately one another. To reach the new account there should be a distance between them. Not immediately one after another. So it won't be same with single account. I guess it sounds like a good strategy? We can even use grid with that system?
like the moderator mentioned, you and I have different meanings of the word "account". I have no idea what you mean, so this will be the last msg i will post here.