Does this system have an edge or is it just the last year?

victormork · Thu, 23 Mar 2017 13:18:37 +0000

I created this basic beginner code for the EURUSD 1H and added a trailing stop function which Nicolas had posted on the forum. The result looks okay but my concern is that it’s starting first after 2015 (1 pip spread). Any takes on what I could do differnet. And how is the performance before 2013? Short trades has been far more profitable than long trades during the backtest. Maybe not a surprise since the longer trend has been to the downside. // Definition of code parameters DEFPARAM CumulateOrders = False // Cumulating positions deactivated // Prevents the system from placing new orders on specified days of the week daysForbiddenEntry = OpenDayOfWeek = 6 OR OpenDayOfWeek = 0 // Conditions to enter long positions indicator1 = CCI[8] c1 = (indicator1 CROSSES OVER -240) indicator2 = BollingerBandWidth[20](close) c2 = (indicator2 > 0.003) indicator3 = TRIX[12](close) indicator4 = Average[9](indicator3) c3 = (indicator3 > indicator4) IF (c1 AND c2 AND c3) AND not daysForbiddenEntry THEN BUY 1 CONTRACT AT MARKET ENDIF // Conditions to enter short positions indicator7 = CCI[8] c5 = (indicator7 CROSSES UNDER 140) indicator8 = BollingerBandWidth[20](close) c6 = (indicator8 > 0.003) indicator9 = Average[9](TRIX[12](close)) indicator10 = TRIX[12](close) c7 = (indicator9 > indicator10) IF (c5 AND c6 AND c7) AND not daysForbiddenEntry THEN SELLSHORT 1 CONTRACT AT MARKET ENDIF // Stoploss SET STOP pLOSS 55 trailingstart = 50 //trailing will start @trailinstart points profit trailingstep = 5 //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 35 replies, 5 voices, and was last updated 7 years ago by victormork.

Viewing 15 posts - 16 through 30 (of 36 total)

← 1 2 3 →

03/30/2017 at 11:59 AM #30317

victormork

Thanks GraHal! I think that was a great way to stop the system from taking “stupid” trades where the market is looking to reverse.

Do you still have the system running? The second short trade was closed +25 pips, now long… think the first loss may come now.

victormork
Participant

Topics: 19
Replies: 165
Been thanked: 57 times

03/30/2017 at 12:22 PM #30341

GraHal

Ha yeah I stopped the ‘old version’ Short at 30 points gain (attached) as I saw the slightest of signs of a turn back up ( a few lows on the 5 Min Chart not making Lower Lows if I recall correctly).

Then I read you had done the new version so I set new going in place of old and within 3 points the new took a Long entry!

Seems your System and my brain work similar … so far anyway! 🙂

The Price is dithering around the Long entry point, so yeah could go either way at this stage I reckon?

Just checked and could be about to breakout UP at 10,744??

Regards
GraHal

GraHal
Participant

Topics: 204
Replies: 8403
Been thanked: 2043 times

03/30/2017 at 1:28 PM #30350

victormork

Yeah it’s anyone’s game at the moment but so far no signs of real buying. I’m planing to set the system on my real account since both entry and exit looks to be in order.

victormork
Participant

Topics: 19
Replies: 165
Been thanked: 57 times

03/30/2017 at 2:45 PM #30359

GraHal

My Long exited at 14:00 @ 10722 but on Demo it is still Long … weird? It may that odd thing where there is a ‘1 x TF delay’ in Demo before it shows, so it may look correct at 15:00?? Or is that ‘odd thing’ only on Daily TF??

Since 14:25 its gone up to 10749 on 4 or 5 Long 1 Min bars, but this wasn’t predictable so your system couldn’t have been expected to react that quick as it’s on a 1H TF … it may be back down again by 15:00.

All interesting market watching this innit! ha

GraHal

GraHal
Participant

Topics: 204
Replies: 8403
Been thanked: 2043 times

03/30/2017 at 5:18 PM #30381

GraHal

Ha I reckon you are still Long … as I am also on Demo.

It was my ‘Negative Equity Loss Limiter’ that exited my Long in Live at 14:00.

The first trade (on your latest version) started with zero Strategy Profit and so the Loss Limiter below triggered.

If STRATEGYPROFIT < 15 AND POSITIONPERF < - 0.0004 then SELL at Market EXITSHORT at Market Endif

1
2
3
4

If STRATEGYPROFIT < 15 AND POSITIONPERF < - 0.0004 then
SELL at Market
EXITSHORT at Market
Endif

Loss Limter wouldn’t trigger in Demo on the same Long Trade because there have been previous trades (which made profit) and so the Demo Account is not in negative equity at that point / 14:00.

GraHal

GraHal
Participant

Topics: 204
Replies: 8403
Been thanked: 2043 times

03/30/2017 at 5:40 PM #30383

manel

Hi victormork, many thanks for sharing this. I’ve attached the results over a longer timeframe. The positive trend only really starts at the beginning of 2013. Did you optimise each variable or only a set ? Maybe we can try and adjust some parameters so that it’s also more consistent over the previous years.

manel
Participant

Topics: 4
Replies: 259
Been thanked: 121 times

03/30/2017 at 6:09 PM #30392

GraHal

How come 200,000 x 1h bars go back 19 years (Apr 98) and yet 100,00 x 1h bars go back only 4 years 3 months to 27 Dec 12??

GraHal
Participant

Topics: 204
Replies: 8403
Been thanked: 2043 times

03/30/2017 at 6:38 PM #30394

manel

No idea but I’ve noticed since day 1 that it happens everywhere in PRT. The 100,000 bars history should be taken with a huge bag of salt as it doesn’t always actually give you 100,000 units but “whatever’s available” to your platform version. It’s common across all instruments, otherwise even in your case, 100,000 should produce (100,000/24)/365 = 11.4 yrs of data (actually even more given no trading on sat and most of sunday) which it obviously doesn’t. Mysteries….. 🙂

1 user thanked author for this post.

GraHal

manel
Participant

Topics: 4
Replies: 259
Been thanked: 121 times

03/30/2017 at 6:41 PM #30395

manel

Notwithstanding that, be aware that this is probably an over-optimised system so it’s ok running in live right now as it does work quite well in recent market conditions but just be prepared to get off the bandwagon if it starts underperforming (though knowing when that point comes will be even more difficult).

1 user thanked author for this post.

GraHal

manel
Participant

Topics: 4
Replies: 259
Been thanked: 121 times

03/30/2017 at 6:57 PM #30398

Eric

5 year 1 hour should go back to march 2012?

and 15 year march 2002?

https://www.prorealtime.com/en/technology_premium

Eric
Participant

Topics: 18
Replies: 441
Been thanked: 84 times

03/30/2017 at 7:04 PM #30400

GraHal

Manel, thanks yes I have been considering with Systems generally (not just this one) to make notes re a few performance stats and also under what conditions the System appears to work best (uptrend, down or ranging). Then weekly to do a quick check to satisfy myself that conditions and performance are not shifting etc. However, it’s all admin faffing which I’m not keen on! 🙂

I’ve yet to get PRT Walk Forward, but from what Nicolas says, the WF periods chosen could be used to indicate how often Systems need to be optimised to maintain performance etc.

Having stopped all Live Bots for a few weeks, I’m back running Live, but I also monitor market structure and stop a Bot if I judge it’s going the wrong way.

Cheers
GraHal

GraHal
Participant

Topics: 204
Replies: 8403
Been thanked: 2043 times

03/30/2017 at 7:10 PM #30401

manel

@ Eric, yes good spot. But unfortunately I find that table to be mostly untrue in reality. For eg for Dax there should be 8yrs of 5m data but in actual fact, there is around 3yrs that is available to Premium users.

I think sometimes there are limitations on the data the platform can provide (due to 200,000 bars limit) or that they simply don’t have the data history or that PRT have got it incorrect on the website. For eg. 5m 200,000 bars would equate to roughly 3yrs (excluding weekends) which is what is provided so the website statement of 100 mths of data simply cannot be achieved as the 200k limit would kick in first.

manel
Participant

Topics: 4
Replies: 259
Been thanked: 121 times

03/30/2017 at 7:21 PM #30402

victormork

That is weird! There was no sell signal. However I think the trade that is still running on demo is going to be a loser..

victormork
Participant

Topics: 19
Replies: 165
Been thanked: 57 times

03/30/2017 at 7:41 PM #30403

Nicolas

The link provided by Eric is not about CFDs contracts.

Nicolas
Keymaster

Topics: 81
Replies: 20828
Been thanked: 4742 times

03/30/2017 at 8:00 PM #30404

GraHal

@victormork yes my Demo Long is still running also, but my Live Long exit was due to my Loss Limiter (see my post #30381) … which got me out at £16 loss only. The Demo Long is currently – £50 ish.

Cheers
GraHal

GraHal
Participant

Topics: 204
Replies: 8403
Been thanked: 2043 times
Author

Posts

Viewing 15 posts - 16 through 30 (of 36 total)

← 1 2 3 →

Find exclusive trading pro-tools on

Does this system have an edge or is it just the last year?

1 user thanked author for this post.

1 user thanked author for this post.