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 - 91 through 105 (of 344 total)

← 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 … 19 20 21 22 23 →

11/16/2020 at 8:52 PM #150699

nonetheless

thanks for that, looks like it could be worth breaking into 2 separate algos.

nonetheless

Participant

Topics: 83

Replies: 1237

Been thanked: 978 times

11/16/2020 at 10:49 PM #150713

josef1604

Hello, could you tell me how it can be only long or only short? I would like to learn, please

josef1604

Participant

Topics: 1

Replies: 50

Been thanked: 2 times

11/16/2020 at 11:48 PM #150720

nonetheless

Simple but inelegant: duplicate the code, then delete or // anything related to short on one and anything related to long on the other. Then they can both be re-optimized.

Otherwise you can add a long/short selector code, such as the one Paul uses:

once tradetype       = 1 // [1] long/short [2]long [3]short

1	once tradetype = 1 // [1] long/short [2]long [3]short

then you have to put your entry conditions inside an If statement:

if tradetype=1 or tradetype=2 then
(long entry conditions)
endif

if tradetype=1 or tradetype=3 then
 (short entry conditions)
 endif

if tradetype=1 or tradetype=2 then

(long entry conditions)

endif

if tradetype=1 or tradetype=3 then

(short entry conditions)

endif

nonetheless

Participant

Topics: 83

Replies: 1237

Been thanked: 978 times

12/17/2020 at 11:37 PM #154220

nonetheless

Here’s a Long only update for this one – 6 year backtest, 75/25 opt. The 18 months OOS is much better than it should be … a good problem to have but still a bit weird ???

I feel it could be improved, esp with some sort of exit conditions, but nothing I’ve tried seems to help. All ideas welcome.

This has max positions = 3, above that makes more money but less stable.

6 users thanked author for this post.

winnie37,

josef1604,

swedshare,

Roger,

MAKSIDE

nonetheless

Participant

Topics: 83

Replies: 1237

Been thanked: 978 times

12/18/2020 at 12:01 PM #154255

Tanou

Hello @Nonetheless!

Thanks for this sharing! I don’t have quite the same results as you however. In the screenshot you’ve provided I can see that it’s written “opt” at the end of the backtest name. Do you have an optimised version?

This one looks pretty good, I see if I spot something to improve 😉

Tanou

Participant

Topics: 18

Replies: 153

Been thanked: 29 times

12/18/2020 at 12:21 PM #154259

nonetheless

the ‘opt’ version is just my working algo with the variables in the opt box. The one I posted has the same values inserted in the code.

How do your results differ?

nonetheless

Participant

Topics: 83

Replies: 1237

Been thanked: 978 times

12/18/2020 at 3:56 PM #154311

Tanou

Ah ok! Do you have this algo in real?

I have 2.80 of gain/loss ratio. The curve looks quite close anyway 😉

Tanou

Participant

Topics: 18

Replies: 153

Been thanked: 29 times

12/18/2020 at 4:24 PM #154318

nonetheless

Tanou wrote:

I have 2.80 of gain/loss ratio

On a 6 year backtest?

I’m not running it live yet, hoping for some inspired suggestions on how to improve it…

nonetheless

Participant

Topics: 83

Replies: 1237

Been thanked: 978 times

12/19/2020 at 8:11 PM #154464

OboeOpt

Hi @nonetheless

I don´t get the same result as you either, any ideas?

OboeOpt

Participant

Topics: 11

Replies: 101

Been thanked: 7 times

12/19/2020 at 10:41 PM #154483

nonetheless

I have no idea why you would get a different result, although I have had some v strange experiences with backtests in v11.

I ran it again just now and I got the attached print.

nonetheless

Participant

Topics: 83

Replies: 1237

Been thanked: 978 times

12/19/2020 at 11:11 PM #154489

OboeOpt

Thank you! Oh, I have had many strange experiences with PRT overall…

Did you run it again from your installed algo, or did you try to install your published itf again?

OboeOpt

Participant

Topics: 11

Replies: 101

Been thanked: 7 times

12/19/2020 at 11:52 PM #154500

nonetheless

it’s the exact same that I published.

nonetheless

Participant

Topics: 83

Replies: 1237

Been thanked: 978 times

12/20/2020 at 12:04 AM #154503

OboeOpt

Ok strange, thanks!

OboeOpt

Participant

Topics: 11

Replies: 101

Been thanked: 7 times

12/30/2020 at 2:44 PM #155688

Munshun

I have been using this algo for only a week, so I don’t have a ton of data for this observation. The trades the algo has takes so far has always been 3 at a time and it has always happened almost at the same time, just a few moments in between. I am aware of the Cumulative order setting and I see in this thread that max number of positions are 3.

Is it intentional that the cumulative orders are picked up at the “same” time however? Shouldn’t they be paced a bit more, like if the algo takes a trade the conditions for that trade is “nullified” and it will not take another trade until the conditions have happened again based on the moment of the first trade, which would likeley be some time in between.

I am total novice on interpreting the algos but I am trying to learn, and maybe the behaviour I suggest is just not possible?

Munshun

Participant

Topics: 0

Replies: 8

Been thanked: 1 time

12/30/2020 at 3:14 PM #155689

Roger

It sounds like a good idea, maybe you can add a variable (“space”) and try to add something like : (not tested)

Cspace = (barindex-tradeindex) > space

IF not longonmarket and Ctime and c1 and c1a AND C3a and c3b and c3c AND C5a and c5b THEN
BUY positionsize CONTRACT AT MARKET
elsif longonmarket and Ctime and c1 and c1a and c3 and c3a and c5 AND C5a and COUNTOFLONGSHARES < MaxPositionsAllowed and Cspace then
BUY positionsize CONTRACT AT MARKET
ENDIF

Cspace = (barindex-tradeindex) > space

IF not longonmarket and Ctime and c1 and c1a AND C3a and c3b and c3c AND C5a and c5b THEN

BUY positionsize CONTRACT AT MARKET

elsif longonmarket and Ctime and c1 and c1a and c3 and c3a and c5 AND C5a and COUNTOFLONGSHARES < MaxPositionsAllowed and Cspace then

BUY positionsize CONTRACT AT MARKET

ENDIF

And then optimize “space”. But Im not sure of the Tradeindex behaviour when accumulating positions.

Roger

Participant

Topics: 9

Replies: 93

Been thanked: 30 times

Author

Posts

Viewing 15 posts - 91 through 105 (of 344 total)

← 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 … 19 20 21 22 23 →

Find exclusive trading pro-tools on