Zilliq Challenge : Give me a Donkey Indicator to transform in a RaceHorse ;-)

• 09/26/2020 at 8:20 PM #145475
  zilliq

  Hi Guys,

  Next game and next trials…

  Perhaps as you see on the CCI, STO, SAR, Repulse file we can create a winning algo with a dumb indicator …

  I really think we can create a winning algo with whatever Donkey indicator because the most important things in an algo is Market structure and money management

  So the new challenge of this week (hope to have some time for) is to create an algo with an indicator you choose  and my goal is to create an algo who can win after one week

  Always on EUR/USD (Less bias) 1 mn and so on …

  Let’s play Guys !

   

  

  zilliq

  Participant

  Topics: 58

  Replies: 683

  Been thanked: 152 times
  09/26/2020 at 8:37 PM #145478
  MAKSIDE

  Hi Zilliq

  In this case, i suggest only an algo based on candle and pivot.  😉

   

   

  

  MAKSIDE

  Participant

  Topics: 37

  Replies: 343

  Been thanked: 98 times
  09/26/2020 at 8:37 PM #145479
  Scooby

  Hi Zilliq,

  I don’t know if it should be a good challenge, but I worked with the bollinger bands with some timeframes and some markets.

  Then I try to purpose this algo :

   

  Bollinger bands

  defparam cumulateorders = true defparam preloadbars = 5000 defparam flatbefore = 080000 defparam flatafter = 180000 period = 20 dev = 2 data = customclose MA = average[period](data) BolUp = MA+STD[period]*dev BolDn = MA-STD[period]*dev //VENTE c1v = high[1] > BolUp[1] or high[2] > BolUp[2] c2v = high < BolUp if c1v and c2v then sellshort 1 contract at market set stop ploss 20 set target pprofit 50 endif //SORTIE VENTE cc1v = low[5] < BolDn[5] or low[4] < BolDn[4] or low[3] < BolDn[3] or low[2] < BolDn[2] or low[1] < BolDn[1] cc2v = close > open if cc1v and cc2v then exitshort at market endif //ACHAT c1a = low[1] < BolDn[1] or low[2] < BolDn[2] c2a = low > BolDn if c1a and c2a then buy 1 contract at market set stop ploss 20 set target pprofit 50 endif //SORTIE VENTE cc1a = high[5] > BolUp[5] or high[4] > BolUp[4] or high[3] > BolUp[3] or high[2] > BolUp[2] or high[1] > BolUp[1] cc2a = close < open if cc1a and cc2a then sell at market endif if time > 222500 then sell at market exitshort at market 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 45 46 47 48 49 50 51 52 53 54 55 56

  defparam cumulateorders = true defparam preloadbars = 5000 defparam flatbefore = 080000 defparam flatafter = 180000   period = 20 dev = 2 data = customclose MA = average[period](data) BolUp = MA+STD[period]*dev BolDn = MA-STD[period]*dev   //VENTE   c1v = high[1] > BolUp[1] or high[2] > BolUp[2] c2v = high < BolUp   if c1v and c2v then sellshort 1 contract at market set stop ploss 20 set target pprofit 50 endif   //SORTIE VENTE   cc1v = low[5] < BolDn[5] or low[4] < BolDn[4] or low[3] < BolDn[3] or low[2] < BolDn[2] or low[1] < BolDn[1] cc2v = close > open   if cc1v and cc2v then exitshort at market endif   //ACHAT   c1a = low[1] < BolDn[1] or low[2] < BolDn[2] c2a = low > BolDn   if c1a and c2a then buy 1 contract at market set stop ploss 20 set target pprofit 50 endif   //SORTIE VENTE   cc1a = high[5] > BolUp[5] or high[4] > BolUp[4] or high[3] > BolUp[3] or high[2] > BolUp[2] or high[1] > BolUp[1] cc2a = close < open   if cc1a and cc2a then sell at market endif   if time > 222500 then sell at market exitshort at market endif

   

  

  Scooby

  Participant

  Topics: 9

  Replies: 52

  Been thanked: 26 times
  09/26/2020 at 10:45 PM #145484
  Vonasi

  I suggest using the indicator called price.

  defparam cumulateorders = false if not onmarket and close < open then buy 1 contract at market endif if not onmarket and close > open then sellshort 1 contract at market endif

  1 2 3 4 5 6 7 8 9

  defparam cumulateorders = false   if not onmarket and close < open then buy 1 contract at market endif   if not onmarket and close > open then sellshort 1 contract at market endif

  If you can make that profitable across a very large historic data sample and also in a large forward testing sample then you will have proven a point!

  1 user thanked author for this post.

  snucke

  

  Vonasi

  Moderator

  Topics: 192

  Replies: 5088

  Been thanked: 1811 times
  09/27/2020 at 11:12 AM #145502
  zilliq

  Guys, I ask for an indicator for this challenge, not a setup on Price 😉

  But whatever, I test with

  c1 = (close > open) c3 = (close < open)

  1 2 3

  c1 = (close > open)   c3 = (close < open)

  And you can find the results on 100 000 bars (Always with the same code for analyze structure and money management)

  +436 %, 82.14 % winners, on 56 trades Win/Loss 3.94 Sharpe ratio 1.14 (Little bit small)

  EUR/USD 1 mn Capital 10 000

  Bye

  

  zilliq

  Participant

  Topics: 58

  Replies: 683

  Been thanked: 152 times
  09/27/2020 at 11:19 AM #145505
  Vonasi

  Price movements are an indicator. A candle is just an indicator of price and what it has just done in the last x period of time.

  

  Vonasi

  Moderator

  Topics: 192

  Replies: 5088

  Been thanked: 1811 times
  09/27/2020 at 1:10 PM #145524
  zilliq

  Strange, my last message disappear …

  Whatever, I test with the Bollinger @Scooby but only as a signal not with all the strategy you propose because I couldn’t add the code for market structure and money management

  The results are not bad

  +479 %, 93.62 % winners win/loss of 10.5 and sharpe ratio of 0.95

  I will run this second algo and see if it win the challenge to be a winner after one week

  Bye

  indicator1=BollingerUp[a](close) indicator2=BollingerDown[a](close) c1 = (close crosses over indicator1) c3 = (close crosses under indicator2)

  1 2 3 4 5 6

  indicator1=BollingerUp[a](close) indicator2=BollingerDown[a](close)   c1 = (close crosses over indicator1)   c3 = (close crosses under indicator2)

   

  

  zilliq

  Participant

  Topics: 58

  Replies: 683

  Been thanked: 152 times
  09/27/2020 at 1:14 PM #145527
  GraHal

  zilliq wrote:

  And you can find the results on 100 000 bars

  Re the disappeared message (very strange?) … Please forgive me if I am wrong, but I can’t stop myself thinking that ‘smoke and mirrors’ might be used here? 🙂

  How come you choose not to show us the dates of the equity curve and also the duration of the backtest on the performance screen shot?

  I sincerely want to believe what you are promoting as maybe then we all could have yachts like Vonasi and sail when the weather is sunny! 🙂

   

  

  GraHal

  Participant

  Topics: 211

  Replies: 8778

  Been thanked: 2182 times
  09/27/2020 at 1:20 PM #145528
  zilliq

  Hi @Grahal

  Not sure what picture you want to with dates. This one ?

  If you want tto another I can post another, no problem

  Bye

   

  1 user thanked author for this post.

  GraHal

  

  zilliq

  Participant

  Topics: 58

  Replies: 683

  Been thanked: 152 times
  09/27/2020 at 1:23 PM #145530
  zilliq

  Another graph with dates with the Boll setup

  2 users thanked author for this post.

  Paul, GraHal

  

  zilliq

  Participant

  Topics: 58

  Replies: 683

  Been thanked: 152 times
  09/27/2020 at 1:48 PM #145535
  Vonasi

  GraHal wrote:

  How come you choose not to show us the dates of the equity curve and also the duration of the backtest on the performance screen shot?

  I’m with you on this one. It is so much better if everyone posts screenshots with all the data included – dates, time frame, instrument and also position size. My guess is that what you call smoke and mirrors is in fact just some form of averaging down – or catching a falling knife.

  

  Vonasi

  Moderator

  Topics: 192

  Replies: 5088

  Been thanked: 1811 times
  09/27/2020 at 1:59 PM #145540
  zilliq

  Sorry Guys, I don’t understand what you call “form of averaging down – or catching a falling knife.”

  No problem to post dates and so on …

  Bye

  

  zilliq

  Participant

  Topics: 58

  Replies: 683

  Been thanked: 152 times
  09/27/2020 at 2:30 PM #145553
  Vonasi

  Averaging down is buying more positions as price falls (if long) or shorting more positions as price rises (if short). You average your entry price down and reduce the distance price has to recover to return your position to profit. However sometimes price continues dropping and you run out of money before you can buy enough to save yourself. That’s why it is called ‘catching a falling knife’ – you try to catch it but the knife just keeps goeing through your hand and just keeps falling!

  

  Vonasi

  Moderator

  Topics: 192

  Replies: 5088

  Been thanked: 1811 times
  09/27/2020 at 2:36 PM #145555
  zilliq

  OK thanks @vonasi

  I know what it was, but don’t understand the relationship with these algos ?

  Whatever

  

  zilliq

  Participant

  Topics: 58

  Replies: 683

  Been thanked: 152 times
  09/27/2020 at 2:36 PM #145556
  GraHal

  Vonasi wrote:

  averaging down – or catching a falling knife.

  How would averaging down result in  the (mostly) staircase rise in equity curve that zilliq shows?

  The results would be more conclusive if the Price Curve were shown and also Positions under the equity curve.

  Even allowing for  a large position size (to give overall gain size) the results zilliq post look impressive.

  We should try and reverse engineer his market structure set-ups and money management?? 🙂

  

  GraHal

  Participant

  Topics: 211

  Replies: 8778

  Been thanked: 2182 times
• Author
  Posts