Question: Why do live vs demo accounts give different tester results?

New comment
 

I apologize if this has been discussed before, but I have not yet run across a clear explanation. If someone could simply point me to a previous thread that has actually answered the question, that'd be great.

Specifying the same test period for the same pair and same EA ...

1) I get completely different looking curves which have ended up with profit claims up to 8 times higher when logged into a demo account, than when logged into a live account. When logged in live, the tester return is always lower and there are usually fewer trades. The tester recently showed a gain of $7,000 when logged into the demo account, and a loss of $1,000 when logged in live. And I don't know how to develop EA's with such erratic test results.

2) I can also get a 3rd result from the tester when specifying no account, that is, when I cancel out of the log in box.

Same everything, including the same $10K starting point for each test, except the account currently logged in.

Q: What is the connection between the current account and the historical test?

Q: Is the historical data being accessed, different depending upon how one is logged in?

Q: Or does the tester for whatever reason behave differently for demo vs live accounts?

Q: Which result is 'accurate'?

Thanks in advance to anyone who might help explain this.

reply
 

Q1: Spread, Commission, maybe Volume

Q2: depends, if you downloadet the history, its based on metaquotes prices if you scroll back its data from your server

Q3: No, it uses the Spread/Commision/Swap, basically all variable data which is avaliable at the moment you press start

Q4: Depends on the timeframe you are trading, but since you have such significally differences i assume this is some kind of scalping strategy an so, non of your results is accurate

 

Lou

Use Search and read, see https://www.mql4.com/search/account%20demo%20test%20results%20different

-BB-

New comment