i run the ea on a demo account at the same broker that I want to run it live. If i run it on 2 different brokers, each one can be completely different. I do demos on the same broker that I want to go live with, and do it for only 3 months. After that go live.
Scott David Maclean: Hey guys I was just wondering how you go about preparing your EA for live trading. Obviously you can optimize and do out of sample testing. I usually try and find the settings for the parameters with best profit*recovery*sharpe*trades. Would be interested how you go about it yourselves. Thanks in advance.
Test it on a demo account (or at most a cent account) over a reasonable amount of time to see if it is in fact doing what you want. Only after should you place it on a real account.
Scott David Maclean:
Do not double post!
I have deleted your duplicated topic and moved Fernando's reply to it here.
Hi,
First of all great that you are doing an out of sample test. But I think you're making a crucial mistake.
The out of sample data is very valuable. One should use this data as little as possible for validation. Of course, if you start optimizing in sample, then test the best parameters with out of sample data, and repeat this process many times, you will, if by sheer luck, find a parameter set that works well in both data sets. But what you have done now is almost the same as a normal backtest with all the data. If several variants are tested with the out of sample data, then the out of sample test becomes meaningless. Therefore, this out of sample data is very valuable and should be used as little as possible. It is best to carry out other tests before doing an out of sample test in order to stop the further development of bad strategies directly.
As the others have already suggested, you can of course test a strategy in a demo first. This is also an out of sample test, but you can't fool yourself here. However, I personally do not have the patience to run a strategy in the demo for 2 years or more. Of course, the duration also depends on the number of trades.
Of course, the number of trades in your backtest also plays a very important role, but that is another topic in itself.
All just my humble opinion, hope it helps, ask any questions ;D
First of all great that you are doing an out of sample test. But I think you're making a crucial mistake.
The out of sample data is very valuable. One should use this data as little as possible for validation. Of course, if you start optimizing in sample, then test the best parameters with out of sample data, and repeat this process many times, you will, if by sheer luck, find a parameter set that works well in both data sets. But what you have done now is almost the same as a normal backtest with all the data. If several variants are tested with the out of sample data, then the out of sample test becomes meaningless. Therefore, this out of sample data is very valuable and should be used as little as possible. It is best to carry out other tests before doing an out of sample test in order to stop the further development of bad strategies directly.
As the others have already suggested, you can of course test a strategy in a demo first. This is also an out of sample test, but you can't fool yourself here. However, I personally do not have the patience to run a strategy in the demo for 2 years or more. Of course, the duration also depends on the number of trades.
Of course, the number of trades in your backtest also plays a very important role, but that is another topic in itself.
All just my humble opinion, hope it helps, ask any questions ;D
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Hey guys, I have about 2 years experience with developing EA's using fxdreema. I was wondering how you prepare your EA for live trading.
I have done a lot of tests using sharpe ratio, recovery factor, profit/drawdown, etc. It is difficult to have my in sample tests come up with a decent amount of out of sample results that are good. And even if the in sample and out of sample results are good it doesn't always mean it will work on live account. Any help would be great. Thank you.