Backtesting/Optimization - page 18
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
I think I forgot to say that realData is a qoute data dumper. So go to files directory to obtain the dumped datasets.
I think I forgot to say that realData is a qoute data dumper. So go to files directory to obtain the dumped datasets.
forward testing vs back testing
in back testing result is good
but when we do forward testing the results is bad
y this is????????
can in forward testing the ea make decision (on 4 hour chart )as example
once every 4 hour not every tick
thanx
Backtesting
Hi everyone,
I hope in this thread we can discuss the pros & cons of backtester, I have made many efforts to get my backtester as reliable as possible, 90% modelling quality etc ... however after upgrading from v198 to v201 I get different backtesting results. It also seems important to backtest offline. What does everyone think ? Is there an alternative available to backtest mq4 experts which may be more reliable ? If we could backtest accurately this would be very valuable aid to developing accurate EAs. I have seen experts which backtest badly but forward test well - and its annoying to have to forward test for months and months would be much better if we could sort out MT4 bugs - any input would be appreciated
Maybe you use oscillators that repaints.....so in backtest work wel but in real time......
hi
no i dun think so it is cyberya indicator
Let us know when you figure it out.
hi
when i used forward testing i get very very bad results
i used cyberia allmost one week on 50k account test of course )
Got up to 53700 in a few days and in only one day went back to 50400 with a pending order of -500$
To understand difference of forward and back tests, anyone should know the term : "CONTINGENCY"
I will describe contingency:
something happened, but these things also could have been also happened differently. there is no necessity.
you set up a backtest,
than you find best result for one strict time interval,
but this will not work for other intervals,
WHY?
because this settings are for that time interval.
if your expert has more criteria than the back test result will fly but forward test will fail.
how can we avoid from this problem?
- never use complex Experts. never make complex back test. test one criteria at a time, by this way you can avoid contingincy problem. Things/happenings will vary, but the strategies will not change.
- the strategy should be visible, you should explain what is your aim. by this way you can easily avoid from contingency roblem.
- Backtest feature is for understanding the expert behaviour versus criterias. that is not for finding a best and unique rule by settings. nobody can find a primary rule to make money there is more than one way to make money.
- Expert advisors are for sending orders as you want they cannot think what should you do. So find a good strategy first, than code it than test the criteria to understand the strict trend free points of your strategy, than make forward test, than make a real small money test. do not invest big money, if you find a good strategy be sure that you will be rich
I'm not and expert coder or mql programmer, I m only a trader and I can say that I can find some rules making money but without a strict system like a computer program I know that I will not obey my rules. the expert advisor is for obeying your rules and strategies and I think that it is an extraordinary way.
I'm sorry for my terrible English,
I hope this explanation will be helpful for some members...
..