Backtesting/Optimization - page 85
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
Very interesting and helpful thread!!
Data generated by back tester ...
Just as a curiosity : this is how data generated by back-tester looks like when recorded. Compared to real time data it is obvious that the modeling way is rather distant from what actually happens in everyday world. So, because of that too (along with a total lack of real spread, real bid and ask) be very careful when getting some results from the back test. The only reliable way of testing is still the forward testing
PS: there are some periods that look even stranger than this one, but I think that even this period illustrates why back test is not a perfect tool for testing any EA
Problem with time intervall during backtesting and optimizing
Hello,
I've coded a new EA and want to optimize and backtest it. I try this for the time period from 1.1.2008 to now (22.9.2012). But both stop some time in November 2011 even if I refresh my historical data. (Alpari UK and FxPro, same problem) Any idea ??
Juergen
How does it stop?
See the equity curve in the graph tab and if it goes down to 0 try increasing the initial deposit in the expert properties. That is the most common reason why an EA stops trading in back test
Hello,
I've coded a new EA and want to optimize and backtest it. I try this for the time period from 1.1.2008 to now (22.9.2012). But both stop some time in November 2011 even if I refresh my historical data. (Alpari UK and FxPro, same problem) Any idea ??
JuergenIf in back-test it works and then stops it means that it lost all the money
If it was some other problem it would not trade in the beginning and would not behave in such a similar way on two different brokers back-tests
PS: or you have a protected EA (that is limited to work only to certain date) but in that case you should contact your EA developer
PS: or you have a protected EA (that is limited to work only to certain date) but in that case you should contact your EA developer
Thanks to all, but:
- the EA is my own one and not limited anyway.
- balance and equity are ok and increasing
( ~16.000 € )
- it just does not trade anymore.
when I shift the starting time of backtest for example to 1.1.2009 it does backtest until now. Seems as if there is a limit for the time interval for backtesting :-) and the limit is not from my EA.
Regards, Juergen
...
JStein
It seems that you are right
It seems that there is some problem with backtesting (again)
Out of curiosity tested the simple MACD example that exist and comes with metatrader and on my tests it worked from 1971 to May 2010. Then it stopped for no reason at all : no error written out in the journal, nothing. Guess that it is one of the new bugs that came with some of the builds we were having in the last couple of months
Thanks to all, but:
- the EA is my own one and not limited anyway.
- balance and equity are ok and increasing
( ~16.000 € )
- it just does not trade anymore.
when I shift the starting time of backtest for example to 1.1.2009 it does backtest until now. Seems as if there is a limit for the time interval for backtesting :-) and the limit is not from my EA.
Regards, JuergenJStein
It seems that you are right
It seems that there is some problem with backtesting (again)
Out of curiosity tested the simple MACD example that exist and comes with metatrader and on my tests it worked from 1971 to May 2010. Then it stopped for no reason at all : no error written out in the journal, nothing. Guess that it is one of the new bugs that came with some of the builds we were having in the last couple of monthsI also thought about a bug in MT4 with backtesting but I wondered, that nobody else detected this problem before. But now I see that also other people (you :-) ) do have problems. We'll wait for a bugfix.
Just another thought about backtesting:
- most of us do optimize parameters of their EA's with "Long and Short" trades enabled and get some parameter set with max. balance, min. DD or something like this. But think about this: perhaps, in your EA it makes a difference what parameter set to use for short trades and for long trades. So I optimize my EAs with different parameter sets for short and long and I've detected major differences in parameters. I do this by enabling "short only" or "long only" to get the best parameters. And afterwards I run the EA in two different chart windows, one with the best parameters for long trades and one with the best parameters for short trades. This increased my profit dramatically.
Maybe this doesn't work for all EAs but it's worth a try and doesn't cost any money.