strategy BarHunter DAX v1p

• 01/16/2020 at 7:24 PM #117013
  Francesco

  Last update with your adjustments @paul on nikkei. Tested on 200K, spread on 8.

  

  Francesco

  Participant

  Topics: 32

  Replies: 439

  Been thanked: 200 times
  01/16/2020 at 8:46 PM #117016
  Paul

  Every year positive on the Nikkei! I’ve seen a lot worse tests. Can you post your changed settings so I can have a look?

  On the south african 40.

  I increased the spread to 20. As with the nikkei I want to take a different approach to handle that, but that’s for later.

  So it’s basically the file you got, activated trailing stop, stoploss 2, profittarget 2

  break=20 & only one bar is analysed for long & short in range to 23

  once steps=0.1 once minatrdist=4 once atrtrailingperiod = 14 // atr parameter once minstop = 10 // minimum distance if barindex=tradeindex then trailingstoplong = 8 // trailing stop atr distance trailingstopshort = 8 // trailing stop atr distance

  1 2 3 4 5 6 7 8 9

  once steps=0.1 once minatrdist=4   once atrtrailingperiod    = 14   // atr parameter once minstop              = 10   // minimum  distance   if barindex=tradeindex then trailingstoplong     = 8   // trailing stop atr distance trailingstopshort    = 8   // trailing stop atr distance

  It looks good without WF and it still looks good with WF. (often it breaks down completely) A lot of time the 3 bar is the one, with very few exceptions.

  Ofcourse could be better, but for 1 variable in range to 23 and with this simplicity its great base!

  edit;WF

  1e bar 11

  2e bar 3

  3e bar 3

  4e bar 1

  all others; barnumber 3

  So it’s pretty consistent

  1 user thanked author for this post.

  Francesco

  

  Paul

  Participant

  Topics: 40

  Replies: 812

  Been thanked: 660 times
  01/16/2020 at 8:58 PM #117020
  Paul

  couldn’t resist, the difference to WF other method.

  

  Paul

  Participant

  Topics: 40

  Replies: 812

  Been thanked: 660 times
  01/16/2020 at 9:59 PM #117023
  Francesco

  I’ll take a look later to the last update of South Africa, the results looks amazing!
  Here you have the .its of NIKKEI 1H, i’m not an expert in coding, not even remotely; so probably my attempts to diversify the strategy are futile, but I’m just trying to do something productive to stimulate development 😀

  1 user thanked author for this post.

  Paul

  

  Francesco

  Participant

  Topics: 32

  Replies: 439

  Been thanked: 200 times
  01/17/2020 at 10:27 AM #117043
  GraHal

  Any of you guys got attached Reject at 4:00 am this morning on the V1p version on the DAX?

  Or at any time, any version, any Market?

  What is the fix please? How to code it into the Strategy?

  Would 2 x different min distances be best … 1 x for Market Hours and 1 x for Out of Market Hours?

  Thank You

   

  

  GraHal

  Participant

  Topics: 204

  Replies: 8401

  Been thanked: 2042 times
  01/17/2020 at 11:43 AM #117048
  MAKSIDE

  Hi Grahal,

  I will test also this strategy asap

  but i think the ddmax is too important..

  

  MAKSIDE

  Participant

  Topics: 37

  Replies: 343

  Been thanked: 98 times
  01/17/2020 at 11:52 AM #117050
  GraHal

  I changed below to 12 (from 10) to see if it makes any difference?

  For info: I set below temporarily to 100 (to test the boundaries) but the Long trade was still executed at 4:00 am on backtest.

  once minstop = 12 //10 // minimum distance

  1	once minstop              = 12   //10 // minimum  distance

   

   

  

  GraHal

  Participant

  Topics: 204

  Replies: 8401

  Been thanked: 2042 times
  01/17/2020 at 11:58 AM #117052
  GraHal

  MAKSIDE wrote:

  i think the ddmax is too important..

  Are you referring to Maximum Drawdown?

  Too much / large at around £1k (for lot size =1) over 100k bars over around 4 Trades?

  

  GraHal

  Participant

  Topics: 204

  Replies: 8401

  Been thanked: 2042 times
  01/17/2020 at 12:35 PM #117063
  MAKSIDE

  Yes,.. for me, everyone manages their risk. In general, I try to have the smallest drawdown even if it is to have less performance. PRT is not good for backtesting by drawdown result..  maybe one day..

  

  MAKSIDE

  Participant

  Topics: 37

  Replies: 343

  Been thanked: 98 times
  01/17/2020 at 12:43 PM #117068
  MAKSIDE

  and.. for me,  the perfect system is to have a drawdown as little as possible and to be very very little on the market. But nothing is perfect 😉

  

  MAKSIDE

  Participant

  Topics: 37

  Replies: 343

  Been thanked: 98 times
  01/17/2020 at 12:45 PM #117070
  Francesco

  @makside Gotta risk it for the biscuit 🙂

  

  Francesco

  Participant

  Topics: 32

  Replies: 439

  Been thanked: 200 times
  01/17/2020 at 2:47 PM #117083
  Paul

  @GraHal Your are quick in testing! It’s a good find but not so good in terms of results at first glance.

  As I understand it, the error occurs when the stop is getting placed but is too near to the current high/low.

  you can try this. If it still gives an error, increase the minstopdistance slightly.

   

  if longtrading or (longtrading and shorttrading) then minstopdistance=10 //minimum stop distance between stop and current price if intradaybarindex=barnumberlong then breakvaluelong=high //experiment! if high-close<minstopdistance then breakvaluelong=close+minstopdistance else breakvaluelong=breakvaluelong endif endif endif if shorttrading or (longtrading and shorttrading) then if intradaybarindex=barnumbershort then breakvalueshort=low //experiment! if close-low<minstopdistance then breakvalueshort=close-minstopdistance else breakvalueshort=breakvalueshort endif endif endif

  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

  if longtrading or (longtrading and shorttrading) then minstopdistance=10 //minimum stop distance between stop and current price if intradaybarindex=barnumberlong then breakvaluelong=high //experiment! if high-close<minstopdistance then breakvaluelong=close+minstopdistance else breakvaluelong=breakvaluelong endif endif endif   if shorttrading or (longtrading and shorttrading) then if intradaybarindex=barnumbershort then breakvalueshort=low //experiment! if close-low<minstopdistance then breakvalueshort=close-minstopdistance else breakvalueshort=breakvalueshort endif endif endif

   

  

  Paul

  Participant

  Topics: 40

  Replies: 812

  Been thanked: 660 times
  01/17/2020 at 3:30 PM #117088
  grimweasel47

  Hello – I’m just trying to test this but I keep getting error ‘backtest exceeds limit of repetitions for Walk Forward optimization’ – but I haven’t got WF selected??

  Thanks

  

  grimweasel47

  Participant

  Topics: 21

  Replies: 107

  Been thanked: 29 times
  01/17/2020 at 3:37 PM #117089
  Paul

  @grimweasel47  Hi, Adjust the repetitions to 3, even if you don’t use them. That should fixes it.

  

  Paul

  Participant

  Topics: 40

  Replies: 812

  Been thanked: 660 times
  01/17/2020 at 6:59 PM #117106
  GraHal

  Paul wrote:

  As I understand it, the error occurs when the stop is getting placed but is too near to the current high/low.

  Yes above is the reason for the Reject error.

  I have added the code-fix you kindly and quickly provided and I will run a version without the fix and with the fix to validate the settings etc.

  Many Thanks

  

  GraHal

  Participant

  Topics: 204

  Replies: 8401

  Been thanked: 2042 times
• Author
  Posts