The best trading strategies of PRC for Prorealtime – Backtests update & Rankings
-
with some modifications, but still without stoploss
alessiopippo thanked this postHi @fifi743 @reb
This is great; thanks guys for your inputs; this hereunder strategy modified which is better than the first one will be added and monitored on best-trading-algos.comThe backtest is just impressive ^^^; never seen a 99% ratio of win ^^^^
Please continue to post here all the stategies that you want me to monitor scatered in all the threads;
thanks for your feed abck on our work ^^^^;
FYI, this hereunder strategy is already listed:-) it’s a * star (one star) performer unfortunetly//========================================================================================================== // Code: TEST Bollinger Band Engulfing (Long Only) // Source: Inspired by all things trading related // Author: samsampop // Version 1 // 1.4 fifi743 : adding these conditions in the input // and open>low and high>close and (open[2]<close or close[3]<close or open[1]<indicator5[1] or indicator5[10]<indicator5 or high[2]<high or low[2]<low or dopen(0)<close or dopen(1)<dopen(0)or dclose(1)>dopen(0)) // Index: DJI // TF: 15 min // TZ: EU // Spread: 3 // // Next Needs stop loss // Add shorting strategy //========================================================================================================== DEFPARAM CumulateOrders = False // Cumulating positions deactivated Defparam Preloadbars = 1000 //stp = 2000 //Risk Management PositionSize=1 tp = abs(close-average[20,1](open)) //StopLoss=stp //=== Entry Filter === //Only go long when close is above the close of 1 week prior priorweekclose=close[480] bubba = close[480]-close //points diff = bubba/priorweekclose //as percentage //graph bubba //graph diff //Identify an up trend in the market on same timeframe indicator3=average[55,7] //71 indicator4=average[145,7] //174 F1 = indicator3>indicator4 // Conditions to enter long positions Headroom = average[20,1] - 5 indicator5 = BollingerDown[20](close) c1 = (close[1] < indicator5) c2 = open < close c3 = (close > open[1]) c4 = close[1] <= open[1] c5 = close < Headroom IF F1 and c1 AND c2 AND c3 AND c4 AND c5 AND bubba > 0 and open>low and high>close and (open[2]<close or close[3]<close or open[1]<indicator5[1] or indicator5[10]<indicator5 or high[2]<high or low[2]<low or dopen(0)<close or dopen(1)<dopen(0)or dclose(1)>dopen(0)) THEN BUY PositionSize CONTRACT AT MARKET set target profit tp //set stop $loss StopLoss ENDIF // Exit Zombie Trade EZT = 1 if EZT then IF longonmarket and (barindex-tradeindex(1)>= 200 and positionperf>0) or (barindex-tradeindex(1)>= 500 and positionperf<0) then sell at market endif endifThe backtest is just impressive ^^^; never seen a 99% ratio of win ^^^^
Less than one star from me. 😉 And still subtract one star for the lacking SL.
The trades/revenue you see in the 1st attachment could be of any substance. But spread over 1,5 years ? forget it. You can’t even by bread from it. But this is okay of course and to each his own. However, the fun really fades if you see what you need to do for it. That shows in attachment #2.
I post this, not to rain on someone’s parade, but to show to people how to interpret Systems which look to good to be true.
A kind of positive (though subjective) remark : This System is super decent and it doesn’t even require a SL. But on the other hand, that’s exactly where the tradeoff is hidden : it is so decent that it trades so few times that nothing can be made of it.
🙂
OK, again :The backtest is just impressive ^^^; never seen a 99% ratio of win ^^^^
Right !
Now on to the why of the difference. I hope I used the correct instrument on IB ?
Btw, I used the $2 cash here (could not find the $1 quickly).Meta Signals Pro thanked this postIB and IG are totally incomparable. See the first picture for about the same time span. IB starts at 15:30 (Amsterdam) and ends at 22:00. IG starts at 00:00 and ends at 23:00.
In this DJI case there’s even sufficient volatility outside trading hours.
So … Peter should have taken the IB DJI Future because IB’s DJI Index is only active during RTH.
This DJI Future coincidentally does not work out for the best because it runs into the “Zombie” protection (500 bars) once – and in bad shape of the trade. The IG Index also runs into that protection but with a coincidental very low loss.So a SL or other measure is still needed and that will make the win rate lower. Of course you can also make the zombie protection more weak (like 1000 bars) but hey …
I did that in the 4th attachment and we see that it is not quite for the better. So it is still a bit of too good to be true, but you must run unlucky in order to be bothered by that. 😉Meta Signals Pro thanked this post
- You must be logged in to reply to this topic.
Author
Summary
This topic contains 68 replies,
has 15 voices, and was last updated by 
8 months, 1 week ago.
Topic Details
| Forum: | General Trading: Market Analysis & Manual Trading |
| Language: | English |
| Started: | 02/09/2023 |
| Status: | Active |
| Attachments: | 39 files |
