Hello fellow traders!
I started this thread because i firmly believe that a PROPER backtest is a backbone of any robust strategy. On the other hand, i think that backtesting is one of the most underestimated topics in trading (in general). Some traders don't care at all for it, and its hard to blame them considering all the problems and contraversy around it.
We all know mt4 is very limited, even with imported history data (thank god for Dukas and Birt's TickSuite, before we had to import it manually). And still you have problems when testing portfolio, even bigger problems when trying to statistically evaluate it, ... . Mt5 has better tester, but impossible to import any data, so it is practically useless for more complex strategies. Programming in R like @integracore2 suggested solves some speed problems, but others stay.
So we are forced to use a little bit mt4, then patch it with QuantAnalyzer (with its own limitations), excel, and other solutions ... at the end, most time goes for that shitty work in stead of backtesting itself.
I started backtesting thread some time ago on Forex Factory, but the responces were more negative than positive (no surprise actually). At the end i had to almost appologize why i even consider backtesting valuable, hehe. Here on Darwinex, the community is on much, much higher expert level. I saw that some commentators are very well informed about it. So maybe we can have a serious discussion here.
So, lets start with a question if someone knows any better way to execute a proper backtest? Share your experience.
How much importance you give to backtesting and its results?
Is it possible to expect some kind of "standardised" backtesting module/procedures here on Darwinex in future? That would be great IMO. Because, if the procedure is valid also results should be a lot more relyable. Maybe it could even be possible to "compensate" a little bit of experience level with it ....?
I mean, the field of backtesting has a lot of potential, so why not look into it more?