Hi Nicolas and other moderators,
I appreciate that the DEFPARAM code isn’t really an indicator but as its having an impact on my backtest result and acting like an indicator I thought it may be relevant for this section of the forum.
I’m still getting wildly different reporting of trades when backtesting and then switching between 100k units (5 Trades & gain £307), 50k units (17 Trades gain £1380) for the same unchanged strategy testing in v11 instead of 10.3 but accept this is just how it is, maybe this will get resolved in V12.
My logic was to try and lock the 50k units results in with a DEFPARAM preloadbars of 50k but then remembered it’s limited to 10,000 units, trying Defparam 10,000 units anyway I noticed that is has a vastly detrimental effect to the backtest figure. Please see attached screenshots.
Can someone please explain the logic of this as it’s hurting my head, the only thing I could think of is that it’s acting like a defparam resctricted hours and cutting the trades short somehow?
Thanks in advance,
Kovit
Post your topic in the correct forum:
_ ProRealTime Platform Support: only platform related issues.
_ ProOrder: only strategy topics.
_ ProBuilder: only indicator topics.
_ ProScreener: only screener topics
_ General Discussion: any other topics.
_ Welcome New Members: for new forum members to introduce themselves.
Moved to ProOrder support.
Thank you 🙂
lol I knew it, no worries thanks Roberto. 🙂
Any thoughts as to why this might be happening, have you encountered it before with your own strategies ?
Might be worth everyone checking their strategy results on a 50k backtest with and without this applied.
I quickly scanned your OP … if you use a variable that uses period, for example, 5000 and you use default Preload bars of 2000 then for the first 3000 bars of running your variable is not calculating … may this be the cause of your Issue?
Thanks Grahal,
I think it’s safer for me just not to use the preload at all I don’t see any benefit. Backtesting is inconsistent and variable enough without adding extra uncertainty into the mix.
Case in point, I just switched back to 10.3 and the same strategy over 50k bars on NAS is showing 61 trades with no losses, instead of 19 trades with 3 losses. I give up. lol
Come on v12, reliable consistent backtest figures that don’t vary ftw lol
Scratch that, I was looking at the same version but with added trailing stop code.
So the strategies were different on each version?
showing 61 trades with no losses
You sure you had tick by tick enabled?
No the strategies were the same, its just the version I backtested on 10.3 had trailing stop code attached to the end of it, which cleared the 3 losses which the strategy would have experienced without it.
Having defaram enabled on all versions either through v10.3 or v11, with or without trailing stop code appears to negatively impact it.
I do have a period reference as GraHal mentioned but my understanding was defaram preloadbars was just a facility to improve speed of loading and backtest rather than something which would affect the gain or number of losses.
I suppose as long as people use shorter term indicators that they are confident would normally display within the 10,000 preloadbars or less it shouldnt be an issue for them to use it.
