Mother of Dragons trading strategy…

nonetheless · Tue, 21 Jan 2020 10:47:18 +0000

Well, I had to call it something, right? Attached is the first code I’ve put together that ‘seems to work’. Survives all manner of WF tests: 70-30, 50-50, 30-70 – however you want to chop it. Also does well in Vonasi’s robustness tester, Random 10-10. I tried hitting it with a very big hammer but still performs well. Uses a combination of HullMA and Supertrend on the 2h TF to define the primary trend, then another HullMA + trailing stop on a 5m TF to control entry and exit. The 5m TF doesn’t give us much in the backtest but it does well in clear uptrends and downtrends. Just that choppy bit in March 2019 where it couldn’t cope. I hate to see whole month where a strategy doesn’t come good, but one out of 16 is not too shabby. It runs around the clock so I’ve allowed for a spread of 3.3, calculated for IG as: (6.5 x 2.4 + 6.5 x 1.6 + 2 x 9.8 + 9 x 3.8)/24 All backtesting was done with MM disabled but I ran the final optimization with it turned on just to demonstrate the boost it gives. Starts with position size = 1 and by the end of the run when equity has trebled, it’s placing position size of ~3.4, so more or less in keeping with standard practise. Also reduces position size in bad patches. Credit where credit is due: Vonasi wrote the Money Management part and coded the HullMA for me. The trailing stop I believe is Nicolas’ work – big thanks to both! But the really cool thing about this strategy is that the equity graph draws a perfect profile of Daenerys Targaryen! You might have to squint a bit, but if you can’t see Daenerys then you haven’t optimized correctly. All comments, suggestions, mods etc welcome! // DJI 5m DEFPARAM CumulateOrders = False // Cumulating positions deactivated DEFPARAM preloadbars = 5000 Capital = 10000 MinSize = 1 //The minimum position size allowed for the instrument. MM1stType = 0 //Starting type of moneymanagement. Set to 0 for level stakes. Set to 1 for increasing stake size as profits increase and decreasing stake size as profits decrease. Set to 2 for increasing stake size as profits increase with stake size never being decreased. MM2ndType = 1 //Type of money management to switch to after TradesQtyForSwitch number of trades and ProfitNeededForSwitch profit has occurred TradesQtyForSwitch = 15 //Quantity of trades required before switching to second money management choice. ProfitNeededForSwitch = 2 //% profit needed before allowing a money management type change to MM2ndType. DrawdownNeededToSwitch = 8 //% draw down from max equity needed before money management type is changed back to MM1stType. DrawdownNeededToQuit = 25 //% draw down from max equity needed to stop strategy Once MoneyManagement = MM1stType Equity = Capital + StrategyProfit maxequity = max(equity,maxequity) if equity < maxequity * (1 - (DrawdownNeededToSwitch/100)) then enoughtrades = 0 tradecount = 0 moneymanagement = MM1stType endif if equity < maxequity * (1 - (DrawdownNeededToQuit/100)) then quit endif if not EnoughTrades then if abs(countofposition) > abs(countofposition[1]) then tradecount = tradecount + 1 endif if tradecount > TradesQtyForSwitch and maxequity >= Capital * (1 + (ProfitNeededForSwitch/100)) then EnoughTrades = 1 MoneyManagement = MM2ndType endif endif IF MoneyManagement = 1 THEN PositionSize = Max(MinSize, Equity * (MinSize/Capital)) ENDIF IF MoneyManagement = 2 THEN PositionSize = Max(LastSize, Equity * (MinSize/Capital)) LastSize = PositionSize ENDIF IF MoneyManagement <> 1 and MoneyManagement <> 2 THEN PositionSize = MinSize ENDIF PositionSize = Round(PositionSize*100) PositionSize = PositionSize/100 // Size of POSITIONS PositionSizeLong = 1 * positionsize PositionSizeShort = 1 * positionsize TIMEFRAME(120 minutes) Period= 520 inner = 2*weightedaverage[round( Period/2)](typicalprice)-weightedaverage[Period](typicalprice) HULLa = weightedaverage[round(sqrt(Period))](inner) c1 = HULLa > HULLa[1] c2 = HULLa < HULLa[1] indicator1 = SuperTrend[5,21] c3 = (close > indicator1) c4 = (close < indicator1) TIMEFRAME(5 minutes) Periodb= 60 innerb = 2*weightedaverage[round( Periodb/2)](typicalprice)-weightedaverage[Periodb](typicalprice) HULLb = weightedaverage[round(sqrt(Periodb))](innerb) c5 = HULLb > HULLb[1] c6 = HULLb < HULLb[1] // Conditions to enter long positions IF c1 AND C3 AND C5 THEN BUY PositionSizeLong CONTRACT AT MARKET SET STOP %LOSS 2.1 SET TARGET %PROFIT 1 ENDIF // Conditions to enter short positions IF c2 AND C4 AND C6 THEN SELLSHORT PositionSizeShort CONTRACT AT MARKET SET STOP %LOSS 1.2 SET TARGET %PROFIT 1 ENDIF //trailing stop function trailingstart = 84 //trailing will start @trailinstart points profit trailingstep = 3 //trailing step to move the "stoploss" //reset the stoploss value IF NOT ONMARKET THEN newSL=0 ENDIF //manage long positions IF LONGONMARKET THEN //first move (breakeven) IF newSL=0 AND close-tradeprice(1)>=trailingstart*pipsize THEN newSL = tradeprice(1)+trailingstep*pipsize ENDIF //next moves IF newSL>0 AND close-newSL>=trailingstep*pipsize THEN newSL = newSL+trailingstep*pipsize ENDIF ENDIF //manage short positions IF SHORTONMARKET THEN //first move (breakeven) IF newSL=0 AND tradeprice(1)-close>=trailingstart*pipsize THEN newSL = tradeprice(1)-trailingstep*pipsize ENDIF //next moves IF newSL>0 AND newSL-close>=trailingstep*pipsize THEN newSL = newSL-trailingstep*pipsize ENDIF ENDIF //stop order to exit the positions IF newSL>0 THEN SELL AT newSL STOP EXITSHORT AT newSL STOP ENDIF //************************************************************************

This topic has 522 replies, 50 voices, and was last updated 3 years ago by LaurentBZH35.

Viewing 15 posts - 421 through 435 (of 523 total)

← 1 2 3 4 5 … 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 →

10/30/2020 at 8:44 PM #149057

OboeOpt

Impressive to see MOD get better and better and that people share their findings!

@nonetheless
-Why do you choose to divide the NAS version into a separate L and S algo? (worse results in a combined one or just for testing out values and combine later?)
-Is there a reason that the NAS version is based on v4, not v5? (I saw that you wrote that v5 should be a good template when testing on other markets)
-Ever tried any of the versions of MOD on forex? (I guess there has to be some changes made to make it even work)

Thanks in advance, Oboe

OboeOpt

Participant

Topics: 11

Replies: 101

Been thanked: 7 times

10/30/2020 at 8:52 PM #149058

Paul

mmm probably it was easier using only this row

condbuy=c1 and c3 and c5 and c7 and c9 and c11 and c13 and c15 and c19 and c21 and c23
condsell=c2 and c4 and c6 and c8 and c10 and c12 and c14 and c16 and c20 and c22 and c24

1 2	condbuy=c1 and c3 and c5 and c7 and c9 and c11 and c13 and c15 and c19 and c21 and c23 condsell=c2 and c4 and c6 and c8 and c10 and c12 and c14 and c16 and c20 and c22 and c24

and keeping the reset conditions to 1. Something to check.

1 user thanked author for this post.

nonetheless

Paul

Participant

Topics: 40

Replies: 812

Been thanked: 660 times

10/30/2020 at 9:43 PM #149061

nonetheless

Thanks Paul, not entirely sure if i get it (cuz I’m a bit simple) but I’ll play around with it at the weekend. 👍

@OboeOpt, yes the Nasdaq being so much more volatile it benefited enormously from splitting the code into long and short. The Dow not so much so I left it as one.

NAS v4 and DJ v5 are the same structure, the Dow has just gone through more iterations, so diff number.

Personally I’ve never had much joy with forex so no idea if that would work… you should have a go!

2 users thanked author for this post.

josef1604,

OboeOpt

nonetheless

Participant

Topics: 83

Replies: 1237

Been thanked: 978 times

11/01/2020 at 11:59 AM #149137

volpiemanuele

Hello, I have create my topic for my preferite strategy…if you want to aid me !!! Thanks

https://www.prorealcode.com/topic/lunch_bias_dax/

volpiemanuele

Participant

Topics: 35

Replies: 189

Been thanked: 49 times

11/01/2020 at 12:04 PM #149142

robertogozzi

@volpiemanuele

Do not double post. Ask your question or state something only once and only in one forum. All double posts will be deleted anyway so posting the same question multiple times will just be wasting your own time and will not get you an answer any quicker. Double posting just creates confusion in the forums

Thank you 🙂

robertogozzi

Moderator

Topics: 69

Replies: 16469

Been thanked: 3643 times

11/01/2020 at 12:15 PM #149144

volpiemanuele

@robertogozzi Thanks….just wanted to share It’ s all clear.

volpiemanuele

Participant

Topics: 35

Replies: 189

Been thanked: 49 times

11/02/2020 at 1:08 AM #149236

StingRe

@BobOgden

My DJI 5.1 also went short (live on forward testing demo) with the attached parameters.

I have another account where it went short on the back test at the same time but it did not take it on the LIVE demo.

I’ve seen this happen a number of times before and investigated it with IG & PRT. They state that the demo accounts are not entirely accurate and reflective of market action so that there will be discrepancies from time to time.

I’ll also note that I have had this happen on a LIVE $$ account where the trade was taken and it should not have been. I’m currently disputing it with PRT who have admitted fault and I’m in the process of getting the lost funds back.

It is important that we keep PRT and IG honest always and dispute and discrepancies we see collectively. This is the only way they will be motivated to improve the accuracy of the system.

The moral of the story here is that PRT and IG don’t entirely have this sorted out yet and you will see anomalies on both Demo and Live accounts that can not be explained.

My theory is that PRT has a resource limit on the number of cycles it will run against demo accounts and therefore some trades will be taken and others won’t.

More concerning for me is the inaccuracy on the LIVE accounts.

2 users thanked author for this post.

swedshare,

saint

StingRe

Participant

Topics: 3

Replies: 20

Been thanked: 9 times

11/03/2020 at 4:25 AM #149340

saint

Thanks to all the contributors here,

May I ask those based in Europe or elsewhere do you change the time frame in PRT to New York time zone when loading the DJI code?

Or should I be doing this differently please?

Thanks Saint

saint

Participant

Topics: 0

Replies: 1

Been thanked: 0 times

11/03/2020 at 10:54 AM #149362

nonetheless

the DJ MoD runs 24hrs a day, you don’t need to do anything to your time settings.

The NAS version takes a 1hr break between 23:00 and midnight GMT, so you should adjust the Ctime values for whatever that equates to down under. It should adjust automatically for US daylight saving.

2 users thanked author for this post.

josef1604,

saint

nonetheless

Participant

Topics: 83

Replies: 1237

Been thanked: 978 times

11/04/2020 at 2:33 AM #149424

Glums

Thanks for sharing all these codes, that is amazing!

I tested NAS version both Long and Short and I am getting weird results, see attached, I could not explain where the problem is coming from, if you have any idea…Thanks

Glums

Participant

Topics: 1

Replies: 9

Been thanked: 0 times

11/04/2020 at 9:19 AM #149432

nonetheless

I expect that you have MM switched on for the long version but not for the short.

1 user thanked author for this post.

Glums

nonetheless

Participant

Topics: 83

Replies: 1237

Been thanked: 978 times

11/04/2020 at 2:54 PM #149464

Trader2323

Thanks again to all the contributors on this amazing topic.

Anyone running the NAS 5m MoD v4.1 S here ? I launched it without MM (MM=0 & positionsize=1), and it took a position on Nov 3rd, at 06:00:00 UTC, which hit stoploss at 15:15:27 UTC. However, this doesn’t appear in the backtest.

I reached PRT out, and I am now waiting for their answer.

I read that @BobOgden had a similar issue a few days ago with the DJ version.

Trader2323

Participant

Topics: 0

Replies: 4

Been thanked: 3 times

11/04/2020 at 6:45 PM #149483

swetraders

Hi @nonetheless,

Thank you for sharing this code. I have looked at and I’m trying to re-optimize and adapt the code on many different markets; DAX, US500, NDQ, etc.
The problem I’m experiencing is that after optimizing the H2 and M15/M30 timeframes, the backtest actually looks good enough to start, and if I try to add / optimize the conditions on the 5 minute timeframe, it only makes the result worse.

When optimizing the higher timeframe I look for many trades and high WF. I can have around 400 trades after High and Middle TF optimization, with a 1.6pf when starting to optimize the M5-timeframe.
Whichever values I choose for M5 it only makes the results worse.
If I’m optimizing more boldly I could have a backtest with around 250 trades with a 3pf (still without the m5-timeframe).

I’m thinking that my process is wrong? Maybe one should not optimize the H2 tf without having any M5-conditions at all?
I really don’t see how your process goes to produce your results. How do you find the M5-values?

swetraders

Participant

Topics: 0

Replies: 6

Been thanked: 1 time

11/04/2020 at 9:30 PM #149488

nonetheless

When optimizing for a new instrument I use 80/20 WF x 1. This gives about 6 months OOS.

I start at the top and do 1 indicator at a time, leaving the TS, SL and TP for last. Then I do it over again at least twice. It’s v time consuming.

I tried to do the SP500 recently but the oos was so bad that I gave up. DAX might be worth it.

2 users thanked author for this post.

swetraders,

josef1604

nonetheless

Participant

Topics: 83

Replies: 1237

Been thanked: 978 times

11/05/2020 at 3:41 AM #149496

April O’Neil

I reoptimised the DJ v5.1 for the Dax, see attached

Not a big amelioration, but, better on the lastest data and less drawdown on the total period. The v2.2a version looks still good. Some variables don’t change this much. Sometimes it can be a good thing to run the same strategy with different parameters like in thise case. We got pretty much the same strategy, but with different parameters and one with the lastest data optimisation, no wf and 200k. Supposed to perform better for this period.

I gave a little eye for a S&P version but looks like this strategy is not supposed to perform on, take care with overfitting for this one. At the moment i’m looking at FTSE, maybe something is possible… Gonna give tests on others but it is very long, more than 2 days.

The best way to do it it’s to slash your variables range and to execute many optimisation windows of a limited number of possibility, like 1000. You can have an eye one more than 1000 results and select variables who give a sense, not just max gain. But i optimise like this only on MoD, because she already proven she win. Don’t optimise like this on strategy you are not sure, or do it on 100k and confirm on 200k.