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 - 61 through 75 (of 344 total)

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

10/21/2020 at 11:46 AM #147986

VinzentVega

swedshare wrote:

Nasdaq has a minimum position of 1.

No, 0.5 ist also possible. Maybe 0.5 is the smallest unit

1 user thanked author for this post.

josef1604

VinzentVega

Participant

Topics: 6

Replies: 115

Been thanked: 27 times

10/21/2020 at 1:24 PM #147993

nonetheless

Paul wrote:

It used positionprice[1] and when there’s no market position and opened a new one it went off the rails.

I ran some quick tests and using positionprice[0] definitely works better with the sensitivity settings and is more consistent with Roberto’s mod.

Strangely though, with ts2sensitivity = 0, I’m seeing better performance using positionprice[1] … which doesn’t really make sense. I’ll do some more tests with other algos.

1 user thanked author for this post.

josef1604

nonetheless

Participant

Topics: 83

Replies: 1237

Been thanked: 978 times

10/21/2020 at 1:34 PM #147994

nonetheless

paisantrader wrote:

If I set custom trading hours in PRT that matches the exchange, for Nasdaq UTC-4. I just set the regular trading hours (9:30-16:00) in the script?

Not sure how custom trading hours works tbh. I always run PRT on my computer time and then adjust the algos accordingly as different indices are optimal at different times.

nonetheless

Participant

Topics: 83

Replies: 1237

Been thanked: 978 times

10/21/2020 at 1:38 PM #147995

robertogozzi

Bear in mind that positionprice[1] is NOT the average price before the last position accumulated, but the value that positionprice retained the previous bar. They may, or may not, be the same. They will different anytime a new position is added, otherwise they will be the same.

It’s not equivalent to Dclose and Dclose(1) which will always differ.

1 user thanked author for this post.

nonetheless

robertogozzi

Moderator

Topics: 69

Replies: 17188

Been thanked: 4064 times

10/21/2020 at 1:42 PM #147996

nonetheless

josef1604 wrote:

I keep seeing that he makes many entries in a row without closing the first ones. For that you have to have a large capital. I have also tried reducing the 0.4lotti, but it keeps coming in with 1lotti. Can you help me please? is there something that escapes me?

This was designed to run with cumulativeorders=true so yes it will make more entries. If you don’t want that, you can change cumulativeorders=false or you can set MaxPositionsAllowed = 1*positionsize

In demo mode the min position on US Tech 100 is 1, when running live min position is 0.5

1 user thanked author for this post.

josef1604

nonetheless

Participant

Topics: 83

Replies: 1237

Been thanked: 978 times

10/21/2020 at 1:53 PM #147997

Paul

thnx for looking into it.

If using positionprice[1] and sensitivity 0 or 1, you find in the detailed rapport a lot of very short trades of mainly 3 bars with a no cumulative orders algo.

mmm it seems with your algo, using positionprice[0], you still have a lot off trades of 0 bars, but that must be a bug in the detailed rapport because of cumulative orders. Was mentioned somewhere?

positionprice[1] works not (always) as it is supposed to, visible when using graphonprice newsl with sensitivity 0 or 1

Paul

Participant

Topics: 40

Replies: 812

Been thanked: 660 times

10/22/2020 at 9:09 AM #148054

josef1604

Thank you first. I knew about FLASE Y TRUE and it works well. But I have observed that if we put MaxPositionsAllowed = 1 * positionize it makes 2, if we put MaxPositionsAllowed = 2 * positionize it makes 3 at most … that is, it always makes one more as I see in the tests.

Thanks for the work you do nonetheless

josef1604

Participant

Topics: 1

Replies: 50

Been thanked: 2 times

10/22/2020 at 12:43 PM #148082

nonetheless

You can change

COUNTOFLONGSHARES <= MaxPositionsAllowed
COUNTOFSHORTSHARES <= MaxPositionsAllowed

1 2	COUNTOFLONGSHARES <= MaxPositionsAllowed COUNTOFSHORTSHARES <= MaxPositionsAllowed

COUNTOFLONGSHARES < MaxPositionsAllowed
COUNTOFSHORTSHARES < MaxPositionsAllowed

1 2	COUNTOFLONGSHARES < MaxPositionsAllowed COUNTOFSHORTSHARES < MaxPositionsAllowed

1 user thanked author for this post.

josef1604

nonetheless

Participant

Topics: 83

Replies: 1237

Been thanked: 978 times

10/22/2020 at 4:43 PM #148103

nonetheless

variation on the 1m version for forward testing.

9 users thanked author for this post.

Balmora74,

LaurentBZH35,

paisantrader,

Kovit,

ichimoku18,

josef1604,

swedshare,

boonet

nonetheless

Participant

Topics: 83

Replies: 1237

Been thanked: 978 times

10/22/2020 at 10:27 PM #148132

josef1604

Thanks @nonetheless I have learned something new. Thank you very much

Now to try in demo the last version that you have uploaded today. Do you have it on real?

josef1604

Participant

Topics: 1

Replies: 50

Been thanked: 2 times

10/22/2020 at 11:04 PM #148137

swedshare

Nonetheless, take a look at the entry conditions in v7.5. There should be SL and TP for the first position as well, not only for the additional, right?

// Conditions to enter long positions
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 COUNTOFLONGSHARES < MaxPositionsAllowed then
BUY positionsize CONTRACT AT MARKET
SET STOP %LOSS 1.4
SET TARGET %PROFIT 1.5
ENDIF

// Conditions to enter long positions

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 COUNTOFLONGSHARES < MaxPositionsAllowed then

BUY positionsize CONTRACT AT MARKET

SET STOP %LOSS 1.4

SET TARGET %PROFIT 1.5

ENDIF

swedshare

Participant

Topics: 2

Replies: 66

Been thanked: 30 times

10/23/2020 at 10:45 AM #148194

nonetheless

Hmm, I had assumed that the stop and target would apply to everything within that IF … ENDIF, but adding it again at line 4 does give a different result

// Conditions to enter long positions
IF not longonmarket and Ctime and c1 and c1a and c1b AND C3a and c3b and c3c AND C5a and c5b THEN
BUY positionsize CONTRACT AT MARKET
SET STOP %LOSS sl
SET TARGET %PROFIT tp
elsif longonmarket and Ctime and c1 and c1a and c3 and c3a and c5 and COUNTOFLONGSHARES < MaxPositionsAllowed then
BUY positionsize CONTRACT AT MARKET
SET STOP %LOSS sl
SET TARGET %PROFIT tp
ENDIF

// Conditions to enter long positions

IF not longonmarket and Ctime and c1 and c1a and c1b AND C3a and c3b and c3c AND C5a and c5b THEN

BUY positionsize CONTRACT AT MARKET

SET STOP %LOSS sl

SET TARGET %PROFIT tp

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

BUY positionsize CONTRACT AT MARKET

SET STOP %LOSS sl

SET TARGET %PROFIT tp

ENDIF

unfortunately though the result was slightly worse. I’m not sure which is technically correct but I think I’ll keep it the way it is for now. It seems to behave as expected in trading.

1 user thanked author for this post.

josef1604

nonetheless

Participant

Topics: 83

Replies: 1237

Been thanked: 978 times

10/24/2020 at 2:35 PM #148324

Monochrome

I added a default macd with no optimisation as an additional confirmation. Less max profit but less drawdown and higher Avg profit/loss ratio.

Only tested for 100k, i just cant get it to work on premium 200k. What instrument is it? dow jones industrial average? not working. If someone can advice me that’d be helpful.

Monochrome

Participant

Topics: 17

Replies: 116

Been thanked: 56 times

10/24/2020 at 3:45 PM #148335

Paul

200k, the mod uses an addition that for each cumulative position, the trigger to activate the trailingstop get’s lower. Not sure If I leave it in or if this strategy is suited for that addition.

if countoflongshares=0 or countofshortshares=0 then
trailingpercentlong  = 0.35 // %
trailingpercentshort = 0.38 // %
endif
if countoflongshares>0.2 then
trailingpercentlong  = max(0.1,trailingpercentlong-(countoflongshares*0.02))
endif
if countofshortshares>0.2 then
trailingpercentshort = max(0.1,trailingpercentshort-(countofshortshares*0.02))
endif

once accelerator     = 0.03 // 1 = default; always > 0 (i.e. 0.5-3)
once accelerator2     = 0.1 // 1 = default; always > 0 (i.e. 0.5-3)
once ts2sensitivity  = 0 // [0]close;[1]high/low;[2]low;high
//====================
once steppercentlong  = (trailingpercentlong/10)*accelerator
once steppercentshort = (trailingpercentshort/10)*accelerator2
if onmarket then
trailingstartlong = positionprice[0]*(trailingpercentlong/100)
trailingstartshort = positionprice[0]*(trailingpercentshort/100)
 
trailingsteplong = positionprice[0]*(steppercentlong/100)
trailingstepshort = positionprice[0]*(steppercentshort/100)
endif

if countoflongshares=0 or countofshortshares=0 then

trailingpercentlong = 0.35 // %

trailingpercentshort = 0.38 // %

endif

if countoflongshares>0.2 then

trailingpercentlong = max(0.1,trailingpercentlong-(countoflongshares*0.02))

endif

if countofshortshares>0.2 then

trailingpercentshort = max(0.1,trailingpercentshort-(countofshortshares*0.02))

endif

once accelerator = 0.03 // 1 = default; always > 0 (i.e. 0.5-3)

once accelerator2 = 0.1 // 1 = default; always > 0 (i.e. 0.5-3)

once ts2sensitivity = 0 // [0]close;[1]high/low;[2]low;high

//====================

once steppercentlong = (trailingpercentlong/10)*accelerator

once steppercentshort = (trailingpercentshort/10)*accelerator2

if onmarket then

trailingstartlong = positionprice[0]*(trailingpercentlong/100)

trailingstartshort = positionprice[0]*(trailingpercentshort/100)

trailingsteplong = positionprice[0]*(steppercentlong/100)

trailingstepshort = positionprice[0]*(steppercentshort/100)

endif

Paul

Participant

Topics: 40

Replies: 812

Been thanked: 660 times

10/25/2020 at 11:30 AM #148392

Fab28

Which variable must be modified to divide by 2 or by 3 the maximum losses of the NAS-1M-HULL-SAR-v7.5 strategy ?

Fab28

Participant

Topics: 9

Replies: 29

Been thanked: 1 time

Author

Posts

Viewing 15 posts - 61 through 75 (of 344 total)

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

Find exclusive trading pro-tools on