5 Min EUR/USD Strategy

• 03/03/2018 at 7:56 PM #64311
  Gertrade

  @ Grahal,

  Indeed, v2.4.2 is the only strategy I developed on TF18s. Apparently, the TF18s and especially the TF28s are promising since we have several strategies that prove profitable over almost a month soon.

  When to drawdowns, I have no information about this in ProOrder statistics, but from memory, I just know that these 2 strategies did not know like the drawdown v2 important and they never had a negative gain .

  Thank you again for your interest in my work.

  Cordially.
  Gertrade

  

  Gertrade

  Participant

  Topics: 4

  Replies: 170

  Been thanked: 60 times
  03/03/2018 at 8:11 PM #64313
  GraHal

  Gertrade the drawdown figure is on the Tab – Statistics of Closed Trades – see red arrows on attached

  Thanks
  GraHal

  

  GraHal

  Participant

  Topics: 204

  Replies: 8399

  Been thanked: 2041 times
  03/03/2018 at 9:13 PM #64318
  Gertrade

  GraHal wrote:

  the drawdown figure is on the Tab – Statistics of Closed Trades

  Unfortunately the value of drawdowns is 0 (below the stats)

  Gertrade

  1 user thanked author for this post.

  GraHal

  

  Gertrade

  Participant

  Topics: 4

  Replies: 170

  Been thanked: 60 times
  03/03/2018 at 9:36 PM #64320
  Gertrade

  below the equity curves of the last week (no more history)

  Gertrade

  1 user thanked author for this post.

  GraHal

  

  Gertrade

  Participant

  Topics: 4

  Replies: 170

  Been thanked: 60 times
  03/04/2018 at 12:24 PM #64352
  Aloysius

  @Gertrade,

  If you change the timeframe in this window (for 1 mn for ex), you can see (and show us) the equity curve since the begining of the strategy, even if the strategy is in a different TF.

  2 users thanked author for this post.

  manel, Gertrade

  

  Aloysius

  Participant

  Topics: 1

  Replies: 44

  Been thanked: 16 times
  03/04/2018 at 6:12 PM #64369
  Gertrade

  Aloysius wrote:

  If you change the timeframe in this window (for 1 mn for ex), you can see (and show us) the equity curve since the begining of the strategy

  Great! Awesome! Bravo! it’s a real pleasure this trick to see the equity curve from the beginning.

  below the equity curve graphs.

  2 users thanked author for this post.

  Aloysius, GraHal

  

  Gertrade

  Participant

  Topics: 4

  Replies: 170

  Been thanked: 60 times
  03/06/2018 at 7:29 PM #64614
  manel

  Hi guys – Just an update from my end on testing of the original 5m versions in Live. I did test a few of the 28s versions initially but they were not very successful at all and the backtests and live were completely different so I gave up on that TF as I feel it’s a bit of a lottery and as a general rule I’m always a bit wary of running anything where there is such limited historic data. But I see that they have been successful for some of you so well done !

  Anyway, with regards to the 5m versions I tested EUR/USD 1.1, 1.2.2 and 1.2.3 and GBP/USD 1.1. Following are key highlights :

  • Live vs Backtest match perfectly and results have shown a net gain over last 3 weeks.
  • As an observation, the strategies only have a high success rate in bullish price patterns – they tend to be much less successful in sideways or bearish price trends. This may create an opportunity to filter some of these losing trades out via a stochastic momentum, RSI indicator etc etc – For next stage of development.
  • I changed the trailing stop to a fixed one as we are aware that the inbuilt PRT trailing code is not always accurate.
  • I amended the stop losses to a much lower level than the original 90 odd pips as some of them I felt were too wide and caused too big a draw. I used 25-35 in most cases and that provides a more acceptable risk reward given the high 70%+ win rate. Each person can adjust to suit their risk profile.
  • Short trades had a higher win rate than long trades historically in backtesting so I doubled the position size for shorts. Given that there are far fewer short trades than long this has not affected the results in a meaningful negative manner if there are losses on these types of trades.
  • I have assumed an increased spread in my backtest of 0.8 whilst testing to allow for the occasional widening of market spreads.
  • Attached are the backtest results of the original EUR/USD v1.1 vs 1.11 (amended) over 200k units with the slight adjustments mentioned above to provide an example of the tweaks made in each version.

  Overall, I think this shows some good promise but there is more work that can be done to improve it, especially the risk reward and the average gain/trade. Thanks again to all those above who have been working on it. I will provide more updates in the future if anything interesting comes up.

  3 users thanked author for this post.

  ltrade, Aloysius, GraHal

  

  manel

  Participant

  Topics: 4

  Replies: 259

  Been thanked: 121 times
  03/08/2018 at 10:25 AM #64754
  Aloysius

  Thank you very much Manel, to improve this code in 5mn, which seems reliable. I will try to work on it too. The backtest works on my platform, But when I try live trading in demo, it is impossible and PRT tells me: “combined stops can not be used with Proorder”. But I see nothing like this in the code. Is someone know where is the problem? Thanks.

  

  Aloysius

  Participant

  Topics: 1

  Replies: 44

  Been thanked: 16 times
  03/11/2018 at 6:42 PM #65011
  Cyril 21

  Hi,

  Here are some ProOrder results for several versions of the algo.

  These versions run for 2 or 3 weeks now.

  Unfortunately, all the 28-second versions have big losses. The best 30-second version that I kept (v5.12) is a little bit negative too.

  But the 5-minute @Gertrade version (v 7.0 for EUR-USD) seems to have great performance on these 2 weeks. On the right of the picture are the details about it.

  It seems that the 5-minute TF isn’t bad… this week I will try the @Gertrade v1.1, and the @manel v1.11 version, and will let you know after some days…
  Note that I stopped the DAX30 v7.01 that had more than €1100 losses in 1 week.

  Have a good week
  Cyril

  3 users thanked author for this post.

  ltrade, GraHal, Gertrade

  

  Cyril 21

  Participant

  Topics: 0

  Replies: 10

  Been thanked: 12 times
  03/11/2018 at 7:12 PM #65013
  Cyril 21

  @manel : I have the same problem than Aloysius with your v1.11 version : “combined stops can not be used with Proorder”. Is somebody able to explain that ? Thanks

  

  Cyril 21

  Participant

  Topics: 0

  Replies: 10

  Been thanked: 12 times
  03/11/2018 at 8:48 PM #65016
  manel

  Hi – Let me check again and see if it happens with me. It’s strange.

  

  manel

  Participant

  Topics: 4

  Replies: 259

  Been thanked: 121 times
  03/11/2018 at 9:10 PM #65017
  manel

  Well, this is strange, I’ve tried starting the code again in my platform and I do indeed get the same error message. Don’t see how this can be as the code I am running live is the same (bar my position size is different) and is working fine. And this code as you can see only has one coded stop function so I don’t understand why ProOrder seems to think that there is another conflicting stop in there. I’m perplexed as to why this is happening, hopefully someone with a bit more experience can help debug. I’ve attached the code below.

  Boll/WR 5 min EUR/USD v1.11

  //100218 //Changes from original code - amended trailing stop to fixed stop with 28 optimised. Doubled short position size. Increased assumed market spread to 0.8 // 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 // Conditions to enter long positions indicator1 = BollingerDown[20](close) c1 = (close <= indicator1) indicator2 = ADX[14] c2 = (indicator2 <= 17) indicator3 = Williams[14](close) c3 = (indicator3 <= -90) IF (c1 AND c2 AND c3) AND not daysForbiddenEntry THEN BUY 1 PERPOINT AT MARKET ENDIF // Conditions to exit long positions indicator4 = Average[20](close) c4 = (close CROSSES OVER indicator4) c5 = (close CROSSES UNDER close[1]) IF c4 AND c5 THEN SELL AT MARKET ENDIF // Conditions to enter short positions indicator5 = BollingerUp[20](close) c6 = (close CROSSES OVER indicator5) indicator6 = ADX[14] c7 = (indicator6 >= 49) IF (c6 AND c7) AND not daysForbiddenEntry THEN SELLSHORT 2 PERPOINT AT MARKET//1 ENDIF // Conditions to exit short positions indicator7 = Average[20](close) c8 = (close CROSSES UNDER indicator7) c9 = (close CROSSES OVER close[1]) IF c8 AND c9 THEN EXITSHORT AT MARKET ENDIF // Stops and targets SET STOP ploss 28//pTRAILING 30 SET TARGET pPROFIT 13

  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52

  //100218 //Changes from original code - amended trailing stop to fixed stop with 28 optimised. Doubled short position size. Increased assumed market spread to 0.8   // 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   // Conditions to enter long positions indicator1 = BollingerDown[20](close) c1 = (close <= indicator1) indicator2 = ADX[14] c2 = (indicator2 <= 17) indicator3 = Williams[14](close) c3 = (indicator3 <= -90)   IF (c1 AND c2 AND c3) AND not daysForbiddenEntry THEN BUY 1 PERPOINT AT MARKET ENDIF   // Conditions to exit long positions indicator4 = Average[20](close) c4 = (close CROSSES OVER indicator4) c5 = (close CROSSES UNDER close[1])   IF c4 AND c5 THEN SELL AT MARKET ENDIF   // Conditions to enter short positions indicator5 = BollingerUp[20](close) c6 = (close CROSSES OVER indicator5) indicator6 = ADX[14] c7 = (indicator6 >= 49)   IF (c6 AND c7) AND not daysForbiddenEntry THEN SELLSHORT 2 PERPOINT AT MARKET//1 ENDIF   // Conditions to exit short positions indicator7 = Average[20](close) c8 = (close CROSSES UNDER indicator7) c9 = (close CROSSES OVER close[1])   IF c8 AND c9 THEN EXITSHORT AT MARKET ENDIF   // Stops and targets SET STOP ploss 28//pTRAILING 30 SET TARGET pPROFIT 13

   

   

  

  manel

  Participant

  Topics: 4

  Replies: 259

  Been thanked: 121 times
  03/11/2018 at 9:28 PM #65018
  manel

  Solved it ! I deleted the stop after //”pTRAILING 30″ in line 51 above and its now fine ! Makes absolutely no sense as PRT is supposed to ignore anything that comes after //….. Maybe it’s a bug, will need to get them to look into it. In the meantime, I’ve reattached the file without this so it should be fine to run now.

  Be aware, like I mentioned above, this needs a bit more work as I would ideally like the risk/reward to be much higher. Am working on some adjustments and one in particular looks promising and will post here after live testing if it’s looking good.

  2 users thanked author for this post.

  GraHal, Cyril 21

  

  manel

  Participant

  Topics: 4

  Replies: 259

  Been thanked: 121 times
  03/12/2018 at 8:54 AM #65032
  Nicolas

  It is not possible to have at the same time SET STOP LOSS and a TRAILING instruction in a strategy running live through ProOrder.

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20828

  Been thanked: 4741 times
  03/12/2018 at 9:22 AM #65036
  Vonasi

  Nicolas wrote:

  It is not possible to have at the same time SET STOP LOSS and a TRAILING instruction in a strategy running live through ProOrder.

  But surely if the pTRAILING stop is remarked out using // then it is not technically in the strategy?

  

  Vonasi

  Moderator

  Topics: 192

  Replies: 5088

  Been thanked: 1810 times
• Author
  Posts