Asian session high/low trading strategy, improvement ideas welcome!

• 03/17/2019 at 8:42 AM #93880
  mr blue

  Dear all

  I came up with the idea to create a short trading strategies that will use the asian high/low of the day as trigger for the rest of the forex trading day.

  At the first glance the code looks good, but I don’t like the sharp decline you see in the  end of the chart.

  Any ideas how to improve the code is very welcome.

  Thanks

  

  mr blue

  Participant

  Topics: 10

  Replies: 77

  Been thanked: 14 times
  03/17/2019 at 9:30 AM #93887
  Vonasi

  Could you post the actual text of the code (using the ‘Insert PRT Code’ button) to save people from having to import the ITF file? You might get more help if it is easier for people to see the code.

  

  Vonasi

  Moderator

  Topics: 192

  Replies: 5088

  Been thanked: 1811 times
  03/17/2019 at 11:14 AM #93889
  mr blue

  version 1

  defparam cumulateorders = false once vHoch = 000000 once vTief = 999999 kStart = 000000 kEnde = 070000 if time > kStart and time < kEnde then if high > vHoch then vHoch = high endif if low < vTief then vTief = low endif endif if time > kEnde and time < 220000 then if close crosses over vHoch then buy 1 contracts at market endif if close crosses under vTief then sellshort 1 contracts at market endif endif if time = 220000 then vHoch = 000000 vTief = 999999 endif

  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

  defparam cumulateorders = false   once vHoch = 000000 once vTief = 999999   kStart = 000000 kEnde = 070000   if time > kStart and time < kEnde then   if high > vHoch then vHoch = high endif if low < vTief then vTief = low endif   endif   if time > kEnde and time < 220000 then   if close crosses over vHoch then buy 1 contracts at market endif   if close crosses under vTief then sellshort 1 contracts at market endif   endif   if time = 220000 then   vHoch = 000000 vTief = 999999   endif

   

  

  mr blue

  Participant

  Topics: 10

  Replies: 77

  Been thanked: 14 times
  03/17/2019 at 11:15 AM #93890
  mr blue

  I am aware that the same functionality could be written smarter – but regarding readabiliy I kept this version.

  

  mr blue

  Participant

  Topics: 10

  Replies: 77

  Been thanked: 14 times
  03/17/2019 at 11:22 AM #93891
  Nicolas

  Don’t worry about your coding style 😉 and thanks for sharing the idea.

  So the strategy is allowed to loose multiple times the same day? I think that there might be something to do with orders quantity in a single day?

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20863

  Been thanked: 4774 times
  03/17/2019 at 11:32 AM #93892
  mr blue

  I did some analysis on the charts and also regarding fake moves. so it’s better – much better to add the final trigger mark 3 pips above high , lower low.

  the revisted code looks now like this:

   

  version 2

  defparam cumulateorders = false once vHoch = 000000 once vTief = 999999 kStart = 000000 kEnde = 080000 if time > kStart and time < kEnde then if high > vHoch then vHoch = high endif if low < vTief then vTief = low endif endif if time > kEnde and time < 220000 then if close crosses over vHoch+0.003 then buy 1 contracts at market endif if close crosses under vTief-0.003 then sellshort 1 contracts at market endif set target pprofit 20 endif if time = 220000 then vHoch = 000000 vTief = 999999 endif

  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

  defparam cumulateorders = false   once vHoch = 000000 once vTief = 999999   kStart = 000000 kEnde = 080000   if time > kStart and time < kEnde then   if high > vHoch then vHoch = high endif if low < vTief then vTief = low endif   endif   if time > kEnde and time < 220000 then   if close crosses over vHoch+0.003 then buy 1 contracts at market endif   if close crosses under vTief-0.003 then sellshort 1 contracts at market endif set target pprofit 20 endif   if time = 220000 then   vHoch = 000000 vTief = 999999   endif

   

  

  mr blue

  Participant

  Topics: 10

  Replies: 77

  Been thanked: 14 times
  03/17/2019 at 11:36 AM #93894
  mr blue

  Nicolas good point. To be honest I haven’t thought about it. But as you see in v2 I added a TP of 20pips now.

  Yes, maybe additional profit could be generated by placing a “smart” stop loss fx here as well.

  Any suggestions folks?

  

  mr blue

  Participant

  Topics: 10

  Replies: 77

  Been thanked: 14 times
  03/17/2019 at 4:17 PM #93900
  Vonasi

  Basically what you have coded is a break out strategy that only trades during the European and US sessions on breakouts of the Asian session high and low. This will always work better in trending markets than in sideways markets so adding a market sentiment filter of some sort might help. Also separating the strategy into two strategies – a long only and a short only might also be beneficial as when the Asian session ends with very close high and low then it is very likely that you will open a trade and quickly get stopped out in a reversal – and possibly many times! Having separate strategies allows you to play a little with different stop ideas other than relying on a breakout in the opposite direction for a stop which could be very close or very far away depending on Asian session volatility.

  1 user thanked author for this post.

  mr blue

  

  Vonasi

  Moderator

  Topics: 192

  Replies: 5088

  Been thanked: 1811 times
  03/17/2019 at 6:27 PM #93901
  Nicolas

  You could also make some data mining (stats) about how orders are performing depending of the High/Low range. Partial closure at predetermined level (e.g. 20 pips profit, next major round level, pivot points, ..) and let the rest run in profit could argue in favor of improvements?

  1 user thanked author for this post.

  mr blue

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20863

  Been thanked: 4774 times
  03/17/2019 at 11:17 PM #93910
  Vonasi

  Nicolas wrote:

  You could also make some data mining (stats) about how orders are performing depending of the High/Low range. Partial closure at predetermined level (e.g. 20 pips profit, next major round level, pivot points, ..) and let the rest run in profit could argue in favor of improvements?

  Are you suggesting simulating trades to match TP and SL to current market volatility? Sounds right up my street although so far I have found that it as badly lagging as a very lagging indicator from my tests so far.

  Your second suggestion of partial closure would be interesting to test but PRT currently don’t allow us to run this live so it might be wasted effort if you want to run the idea automatically.

  

  Vonasi

  Moderator

  Topics: 192

  Replies: 5088

  Been thanked: 1811 times
  03/18/2019 at 8:59 AM #93920
  mr blue

  Vonasi wrote:

  Basically what you have coded is a break out strategy that only trades during the European and US sessions on breakouts of the Asian session high and low.
  > YES

  This will always work better in trending markets than in sideways markets so adding a market sentiment filter of some sort might help.
  -> Could you give me a hint which sentiment filter you think?

  Also separating the strategy into two strategies – a long only and a short only might also be beneficial as when the Asian session ends with very close high and low then it is very likely that you will open a trade and quickly get stopped out in a reversal – and possibly many times! Having separate strategies allows you to play a little with different stop ideas other than relying on a breakout in the opposite direction for a stop which could be very close or very far away depending on Asian session volatility.
  -> Totally agree. But how to accomplish this in PRT? I cannot open long euro and at the same time short euro. Yes, I could go 10x euro mini long and hedge by 1x euro short or play with options – as I do manually. In general – automatic hedging is very tricky or almost impossible with PRT I guess – as we are not able to monitor/trade more than 1 financial instrument withing one algo code.

   

  

  mr blue

  Participant

  Topics: 10

  Replies: 77

  Been thanked: 14 times
  03/18/2019 at 12:17 PM #93940
  Vonasi

  I’m afraid I can’t tell you which sentiment filter but it could be something as simple as a longer term moving average or anyone of the multitude of filters in the library. Maybe jebus89’s idea being discussed here might be a starting point for testing:

  https://www.prorealcode.com/topic/higher-timeframes-bollinger-filter/

  As far as I am aware it is possible to be both long in one strategy and short in another on an instrument. However I believe that if your account is a limited risk account then you can’t do this.

   

  

  Vonasi

  Moderator

  Topics: 192

  Replies: 5088

  Been thanked: 1811 times
• Author
  Posts