• Registered: 2018-12-16
  • Posts: 362

Topic: complete different results eas and mt5 backtest

Hi,

for some time i realize that the backtest of EAs is different of MT5.

I use ticks from dukascopy tickstory data. i import them to my custom symbol.

i use premium EAS data. When i now backtest the ea in MT5 i get complete different results. E.g on EAs i get Profit Factor of 3 and in MT5 it is 1.1 for example--

can someone explain what i did wrong

  • Popov
  • Popov
  • Lead Developer
  • Online
  • From: Bulgaria
  • Registered: 2006-11-30
  • Posts: 8,151

Re: complete different results eas and mt5 backtest

Hello Roughey,

Thank you for posting this issue.

The difference may come from a particular strategy included in the Portfolio Expert.

A highly possible reason for the discrepancy is the so-called "order flow". It means that if we change the starting date of the backtest, it may produce different results because the first trade is different.

Such strategies are dangerous for trading. Monte-Carlo's "Randomise backtest starting bar" may detect such a case.

I'm planning to develop a particular tool for analysing Order-Flow issues.

Please test the individual strategies in the MetaTrader tester to find where the difference is coming from.
Set the MetaTrader's "Starting date" to the date of the first trade shown in the EA Studio Journal.

Popov's Website