NAS 2m HULL-SAR trading system

nonetheless · Fri, 09 Oct 2020 07:49:05 +0000

This is a new 2m strategy I’ve been working on, uses Hull MA and Parabolic SAR in 3 time frames with cumulative orders. Short back test so may be a curve fit but the VRT is good. I think it can still be improved but I’ve run out of ideas so any suggestions are welcome, maybe a limit to the number of positions? Cumulative orders are not to everyone’s taste, and the drawdown is huge – you’d need a lot of capital to run it – but as a % of the runup it’s not that bad. Could be worth playing with… Ctime = time >=143000 and time <210000 TIMEFRAME(18 minutes) Period= 155 inner = 2*weightedaverage[round( Period/2)](typicalprice)-weightedaverage[Period](typicalprice) HULLa = weightedaverage[round(sqrt(Period))](inner) c1 = HULLa > HULLa[1] c2 = HULLa < HULLa[1] ST1 = SAR[.005,.005,.02] c1a = (close > ST1) c2a = (close < ST1) TIMEFRAME(6 minutes) Periodb= 29 innerb = 2*weightedaverage[round( Periodb/2)](typicalprice)-weightedaverage[Periodb](typicalprice) HULLb = weightedaverage[round(sqrt(Periodb))](innerb) c3 = HULLb > HULLb[1] c4 = HULLb < HULLb[1] ST2 = SAR[.005,.005,.02] c3a = (close > ST2) c4a = (close < ST2) TIMEFRAME(default) Periodc= 4 innerc = 2*weightedaverage[round( Periodc/2)](typicalprice)-weightedaverage[Periodc](typicalprice) HULLc = weightedaverage[round(sqrt(Periodc))](innerc) c5 = HULLc > HULLc[1] and HULLc[1] < HULLc[2] c6 = HULLc < HULLc[1] and HULLc[1] > HULLc[2] ST3 = SAR[.01,.01,.005] c5a = (close > ST3) c6a = (close < ST3) // Conditions to enter long positions IF Ctime and c1 and c1a AND C3 and c3a AND C5 and c5a THEN BUY positionsize CONTRACT AT MARKET SET STOP %LOSS .7 SET TARGET %PROFIT 1.7 ENDIF // Conditions to enter short positions IF Ctime and c2 and c2a AND C4 and c4a AND C6 and c6a THEN SELLSHORT positionsize CONTRACT AT MARKET SET STOP %LOSS 1.1 SET TARGET %PROFIT 1.3 ENDIF //%trailing stop function trailingstop = 1 if trailingstop =1 then once trailingPercent = 0.38 once stepPercent = 0.006 if onmarket then trailingstart = tradeprice(1)*(trailingpercent/100) //trailing will start @trailingstart points profit trailingstep = tradeprice(1)*(stepPercent/100) //% step to move the stoploss endif //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 THEN newSL = tradeprice(1)+trailingstep ENDIF //next moves IF newSL>0 AND close-newSL>trailingstep THEN newSL = newSL+trailingstep ENDIF ENDIF //manage short positions IF SHORTONMARKET THEN //first move (breakeven) IF newSL=0 AND tradeprice(1)-close>=trailingstart THEN newSL = tradeprice(1)-trailingstep ENDIF //next moves IF newSL>0 AND newSL-close>trailingstep THEN newSL = newSL-trailingstep ENDIF ENDIF //stop order to exit the positions IF newSL>0 THEN SELL AT newSL STOP EXITSHORT AT newSL STOP ENDIF endif

This topic has 343 replies, 42 voices, and was last updated 1 year ago by bege.

Viewing 15 posts - 76 through 90 (of 344 total)

← 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 … 19 20 21 22 23 →

10/25/2020 at 11:59 AM #148394

nonetheless

Not sure if I understand what you mean. If you want it to run with lower max loss you could reduce the max positions allowed to 2 … still profitable but gain/loss ration is worse.

1 user thanked author for this post.

Fab28

nonetheless
Participant

Topics: 83
Replies: 1237
Been thanked: 978 times

10/27/2020 at 8:54 AM #148623

Fab28

The DJ-1M-HULL-SAR strategy is active from 2:30 p.m. to 9:00 p.m. Yesterday, 2 orders were not triggered at 4.30 p.m. I think there must be a problem with the US and French times. What should I put in the strategy so that there is no problem?

Fab28
Participant

Topics: 9
Replies: 29
Been thanked: 1 time

10/27/2020 at 9:21 AM #148626

nonetheless

It runs on Wall St opening hours. Check what they are for France and use that. Normally it would be 153000 – 220000 but countries change their clocks on different dates so best to double check.

2 users thanked author for this post.

winnie37, Fab28

nonetheless
Participant

Topics: 83
Replies: 1237
Been thanked: 978 times

10/27/2020 at 6:36 PM #148707

Fab28

When I launch the Dj_Hull strategy in real automatic mode, it stops for the following reasoning :

” for the controlled risk order, neither the contingent stop nor the trailing is a stop “

What should I do

Fab28
Participant

Topics: 9
Replies: 29
Been thanked: 1 time

10/27/2020 at 7:12 PM #148714

nonetheless

Are you using guaranteed stops? Do you have a limited risk account?

1 user thanked author for this post.

Fab28

nonetheless
Participant

Topics: 83
Replies: 1237
Been thanked: 978 times

10/27/2020 at 7:51 PM #148719

Fab28

Yes. I have a limited risk account.

When I test the DJ_HULL simulation, the maximum loss of losing positions is 60 €, over 6 months.

If I remove the limited risk option from IG, can I lose more than € 60?

Fab28
Participant

Topics: 9
Replies: 29
Been thanked: 1 time

10/27/2020 at 9:00 PM #148725

nonetheless

That is something you need to discuss with IG, I’ve never had a limited risk account so not sure how they work.

But I do have to urge caution about running this live – it’s a very short backtest and may be completely unreliable. I didn’t do any WF (out of sample) testing as I would do that in forward testing for at least a couple of months. If you do want to try it, my advice is run it with minimum position and MM turned off, just to see what happens. If it’s a winner you can always turn it on later.

1 user thanked author for this post.

nonetheless
Participant

Topics: 83
Replies: 1237
Been thanked: 978 times

10/28/2020 at 8:15 AM #148747

Fab28

I will follow your advices

Thank you

Fab28
Participant

Topics: 9
Replies: 29
Been thanked: 1 time

11/10/2020 at 11:37 AM #150006

nonetheless

I’ve had several variations of this on forward testing for the past month and this is the clear winner.

I reduced the max positions to 3 and re-optimized @ 75/25 so 3 months OOS. VRT looks good.

Seems to work well with max positions up to 8 … if you have sufficient intestinal fortitude (v3b.8)

Goes without saying that this is a very short backtest and the possibility of a curve fit is still very high.

6 users thanked author for this post.

Roger, LaurentBZH35, josef1604, April O'Neil, swedshare

nonetheless
Participant

Topics: 83
Replies: 1237
Been thanked: 978 times

11/11/2020 at 1:49 AM #150083

josef1604

Hello, thanks again, I am testing and I see that I get negative months and I do not get the same number of operations … I get different. Could there be a different TS configuration?

josef1604
Participant

Topics: 1
Replies: 50
Been thanked: 2 times

11/11/2020 at 8:35 AM #150091

Mestor

josef1604 wrote:

Hello, thanks again, I'm testing and I see that I get negative months and I don't get the same amount of operations … I get different. Could there be a different TS configuration?

check lines 35 and 36 …

Mestor
Participant

Topics: 9
Replies: 15
Been thanked: 7 times

11/11/2020 at 8:47 AM #150094

robertogozzi

@Mestor

@josef1604

Only post in the language of the forumthat you are posting in. For example English only in the English speaking forums and French only in the French speaking forums.

Respect the other people, you don’t want to take some time to translate your post and expect people to waste their time to translate what you have written.

Thank you 🙂

2 users thanked author for this post.

josef1604, Dow Jones

robertogozzi
Moderator

Topics: 69
Replies: 16455
Been thanked: 3636 times

11/11/2020 at 9:31 AM #150095

nonetheless

josef1604 wrote:

I get negative months and I do not get the same number of operations

@josef1604, all testing is done with MM switched off which is what I have posted. Change the Ctime setting to Euro time and MM = 0 and you should get something close to what I have shown.

As different brokers offer different leverage, you will probably have to change the factor at line 11 if you want to use MM – try with 10 or 12

1 user thanked author for this post.

josef1604

nonetheless
Participant

Topics: 83
Replies: 1237
Been thanked: 978 times

11/11/2020 at 2:34 PM #150127

josef1604

You were right, making the changes with ctime euro is very similar to yours. Thank you for your great work

josef1604
Participant

Topics: 1
Replies: 50
Been thanked: 2 times

11/15/2020 at 12:43 PM #150549

LaurentBZH35

Hello guys,

I backtested NAS 2min HULL SAR V3b.3 and noticed results were very better in Long Only. You can find below the results.

1 user thanked author for this post.

nonetheless

LaurentBZH35
Participant

Topics: 4
Replies: 75
Been thanked: 35 times
Author

Posts