Discussion re Pure Renko strategy

• 03/19/2020 at 11:28 PM #122648
  GraHal

  Francesco wrote:

  fx pair went crazy high on spreads

  At what time and to what value?

  Please can you optimise the Brick Size and  trailingstoptype over 200k and post the values on here.

  

  GraHal

  Participant

  Topics: 212

  Replies: 8789

  Been thanked: 2187 times
  03/20/2020 at 10:46 AM #122678
  Francesco

  Around 10:45 PM Italian Time, EURUSD went up to 5/6 spread pts.

  And as i said, also others fx pairs went up, but i don’t remember the exact spread.

  In the attachments the optimization you asked for. A1 is the trailingstoptype and A1 is the boxsize.

  1 user thanked author for this post.

  GraHal

  

  Francesco

  Participant

  Topics: 32

  Replies: 439

  Been thanked: 200 times
  03/20/2020 at 11:55 AM #122694
  GraHal

  GraHal wrote:

  Better still if you comment out the Flat before / after … images 3 and 4.

  This was a flawed judgement on my part, apologies!

  Yes the results were better but probably only because I had a fixed spread = 1 in the backtester!

  As pointed out by Francesco, the spread on eurusd went to > 5 at 9:45 ish (UK Time).

  I hadn’t researched overnight spreads as I’ve not used eurusd for quite a while … my enthusiasm blinded me! 🙂

   

  

  GraHal

  Participant

  Topics: 212

  Replies: 8789

  Been thanked: 2187 times
  03/21/2020 at 11:43 AM #122809
  Francesco

  Little 200k backtest of the previous days where it was not possible to test in demo due to the spreads. (Last Paul’s Version with renko=3 and boxsize=20)

  Positionsize=0.2

  

  Francesco

  Participant

  Topics: 32

  Replies: 439

  Been thanked: 200 times
  03/21/2020 at 2:16 PM #122827
  Paul

  @Francesco I haven’t looked yet. on eurusd, but for the dji it seems, although nice, a small stoploss doesn’t work.

  

  Paul

  Participant

  Topics: 40

  Replies: 812

  Been thanked: 660 times
  03/21/2020 at 2:28 PM #122829
  GraHal

  I agree Paul, until spread goes back to 2.4 / 1.6 on the DJI the Renko is a dead duck unless we can get average gain per trade well above current spread = 11.6 / 6.6.

  I note though that spreads overall are heading back down … maybe next week we will see less madness and more like normal trading?

  

  GraHal

  Participant

  Topics: 212

  Replies: 8789

  Been thanked: 2187 times
  03/21/2020 at 5:13 PM #122853
  Francesco

  @Grahal of course at these actual spreads is a madness work and run on dow 1 sec renko.

  I just did a 200k backtest for information with 2.4 spread since I suppose many of you have not run the strategy on live demo.

  

  Francesco

  Participant

  Topics: 32

  Replies: 439

  Been thanked: 200 times
  03/21/2020 at 5:32 PM #122858
  GraHal

  Francesco wrote:

  I just did a 200k backtest

  Yes it is very useful you doing 200k backtests for us all … thank you.

  Yeah I turned my 1 sec DJI off until better times.

  But with different settings or a tweaked strategy I think the 1 sec could live again … even on DJI?

  I think Paul may be working on a new version?

   

  

  GraHal

  Participant

  Topics: 212

  Replies: 8789

  Been thanked: 2187 times
  03/21/2020 at 5:39 PM #122859
  Francesco

  I think it’s not easy to work on these 1 sec strategies, because the backtest frame to work on is very short.

  For example: Paul’s last version in 200k last week made 4 consecutive profts days, while this week 4 consecutive losses.

  Putting myself in the shoes of an expert coder like paul, I wouldn’t know how to optimize the strategy not having a longer frame to observe. How can i know that my version in super profitable last 3 days but maybe 6 days ago was very bad?

  
  @nonetheless said to me that “Renko doesn’t have a time component to it, only price and direction”, but the problem is not that to better optimize the value of a strategy you must always know and work on more data as possible?

  

  Francesco

  Participant

  Topics: 32

  Replies: 439

  Been thanked: 200 times
  03/21/2020 at 6:40 PM #122864
  GraHal

  Francesco wrote:

  this week 4 consecutive losses
  to better optimize the value of a strategy you must always know and work on more data as possible?

  Yeah well my intention was / is to optimise every day! 🙂

  I guess it is a different way of thinking?

  Seconds in a day = 86,400 bars … so this is 86,400 green and red candles and mini up and down trends and flat periods etc?

  Scale that up to hourly bars and 86,400 bars = 14 years worth of trading days?? 🙂

  

  GraHal

  Participant

  Topics: 212

  Replies: 8789

  Been thanked: 2187 times
  03/21/2020 at 9:07 PM #122871
  Francesco

  With the results of the last week, one year would be enough 😀

  

  Francesco

  Participant

  Topics: 32

  Replies: 439

  Been thanked: 200 times
  03/22/2020 at 12:50 AM #122885
  Paul

  a try to calculate the spread (& custom slippage) based on the open of the market order bar timebased.

  // spread wallstreet if time > 090000 and time <= 153000 then spreaddji=2.4 elsif time > 153000 and time <= 220000 then spreaddji=1.6 elsif time > 221500 and time <= 223000 then spreaddji=9.8 elsif time > 230000 and time <= 235959 then spreaddji=9.8 elsif time = 000000 then spreaddji=9.8 else spreaddji=3.8 endif spreaddji=spreaddji/2 // orderhandling if orderhandling then once orderprice=close // prevent graph orderprice at 0 at start chart offsetvalue=((close/10000)*orderoffset)*pointsize // adjustment spread & slippage if longonmarket and not longonmarket[1] then orderprice=open+offsetvalue+spreaddji elsif shortonmarket and not shortonmarket[1] then orderprice=open-offsetvalue-spreaddji endif if longonmarket then pp=((close/orderprice)-1)*100 elsif shortonmarket then pp=((orderprice/close)-1)*100 endif else once orderprice=close // prevent graph orderprice at 0 at start chart orderprice=tradeprice(1) if longonmarket then pp=((close/orderprice)-1)*100 elsif shortonmarket then pp=((orderprice/close)-1)*100 endif 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 38 39 40 41 42 43 44

  // spread wallstreet if time    > 090000 and time <= 153000 then spreaddji=2.4 elsif time > 153000 and time <= 220000 then spreaddji=1.6 elsif time > 221500 and time <= 223000 then spreaddji=9.8 elsif time > 230000 and time <= 235959 then spreaddji=9.8 elsif time = 000000 then spreaddji=9.8 else spreaddji=3.8 endif   spreaddji=spreaddji/2   // orderhandling if orderhandling then   once orderprice=close // prevent graph orderprice at 0 at start chart   offsetvalue=((close/10000)*orderoffset)*pointsize // adjustment spread & slippage   if longonmarket and not longonmarket[1] then orderprice=open+offsetvalue+spreaddji elsif shortonmarket and not shortonmarket[1] then orderprice=open-offsetvalue-spreaddji endif   if longonmarket then pp=((close/orderprice)-1)*100 elsif shortonmarket then pp=((orderprice/close)-1)*100 endif else once orderprice=close // prevent graph orderprice at 0 at start chart orderprice=tradeprice(1) if longonmarket then pp=((close/orderprice)-1)*100 elsif shortonmarket then pp=((orderprice/close)-1)*100 endif endif

   

  

  Paul

  Participant

  Topics: 40

  Replies: 812

  Been thanked: 660 times
  03/22/2020 at 2:36 AM #122886
  Paul

  and again

  once orderhandling = 1 // [0] default; [1] based on open v/d market order once orderoffset = 1 // % average slippage & spread points if above=[1]

  1 2	once orderhandling    = 1    // [0] default; [1] based on open v/d market order once orderoffset      = 1    // % average slippage & spread points if above=[1]

  // spread wallstreet if time > 090000 and time <= 153000 then spreaddji=2.4 elsif time > 153000 and time <= 220000 then spreaddji=1.6 elsif time > 221500 and time <= 223000 then spreaddji=9.8 elsif time > 230000 and time <= 235959 then spreaddji=9.8 elsif time = 000000 then spreaddji=9.8 else spreaddji=3.8 endif spreaddji=spreaddji/2 // orderhandling if orderhandling then once orderprice=close // prevent graph orderprice at 0 at start chart offsetvalue=((close/10000)*orderoffset)*pointsize // adjustment spread & slippage if longonmarket and not longonmarket[1] then orderprice=open+offsetvalue+spreaddji elsif shortonmarket and not shortonmarket[1] then orderprice=open-offsetvalue-spreaddji endif if longonmarket then pp=((close/orderprice)-1)*100 elsif shortonmarket then pp=((orderprice/close)-1)*100 elsif not onmarket then pp=0 endif else once orderprice=close // prevent graph orderprice at 0 at start chart orderprice=tradeprice(1) if longonmarket then pp=((close/orderprice)-1)*100 elsif shortonmarket then pp=((orderprice/close)-1)*100 elsif not onmarket then pp=0 endif endif pp=pp

  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

  // spread wallstreet if time    > 090000 and time <= 153000 then spreaddji=2.4 elsif time > 153000 and time <= 220000 then spreaddji=1.6 elsif time > 221500 and time <= 223000 then spreaddji=9.8 elsif time > 230000 and time <= 235959 then spreaddji=9.8 elsif time = 000000 then spreaddji=9.8 else spreaddji=3.8 endif   spreaddji=spreaddji/2   // orderhandling if orderhandling then   once orderprice=close // prevent graph orderprice at 0 at start chart   offsetvalue=((close/10000)*orderoffset)*pointsize // adjustment spread & slippage   if longonmarket and not longonmarket[1] then orderprice=open+offsetvalue+spreaddji elsif shortonmarket and not shortonmarket[1] then orderprice=open-offsetvalue-spreaddji endif   if longonmarket then pp=((close/orderprice)-1)*100 elsif shortonmarket then pp=((orderprice/close)-1)*100 elsif not onmarket then pp=0 endif else once orderprice=close // prevent graph orderprice at 0 at start chart orderprice=tradeprice(1) if longonmarket then pp=((close/orderprice)-1)*100 elsif shortonmarket then pp=((orderprice/close)-1)*100 elsif not onmarket then pp=0 endif endif pp=pp

  if onmarket then once sl=0.5 once pt=undefined if longonmarket then slvalue=orderprice-(sl*(orderprice/100)) sell at slvalue stop endif if shortonmarket then slvalue=orderprice+(sl*(orderprice/100)) exitshort at slvalue stop endif if longonmarket then ptvalue=orderprice+(pt*(orderprice/100)) sell at ptvalue limit endif if shortonmarket then ptvalue=orderprice-(pt*(orderprice/100)) exitshort at ptvalue limit endif endif

  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

  if onmarket then   once sl=0.5 once pt=undefined   if longonmarket then slvalue=orderprice-(sl*(orderprice/100)) sell at slvalue stop endif if shortonmarket then slvalue=orderprice+(sl*(orderprice/100)) exitshort at slvalue stop endif   if longonmarket then ptvalue=orderprice+(pt*(orderprice/100)) sell at ptvalue limit endif if shortonmarket then ptvalue=orderprice-(pt*(orderprice/100)) exitshort at ptvalue limit endif endif

   

  1 user thanked author for this post.

  GraHal

  

  Paul

  Participant

  Topics: 40

  Replies: 812

  Been thanked: 660 times
  03/23/2020 at 11:06 AM #122987
  Francesco

  Did you tested this “spread finder”? It would be appliable also on fx pairs?

  

  Francesco

  Participant

  Topics: 32

  Replies: 439

  Been thanked: 200 times
  03/23/2020 at 1:59 PM #122996
  Paul

  yes i’am testing 10 sec with above & trendtrading.

  a pic of made yesterday and first one made just now.

  the spreadfix can’t be put on forex, even after changing spread values, just like the trailingstop isn’t suitable. It needs some changes.

   

   

  1 user thanked author for this post.

  Francesco

  

  Paul

  Participant

  Topics: 40

  Replies: 812

  Been thanked: 660 times
• Author
  Posts