EURUSD Morning Entry V1

• 06/13/2019 at 11:55 AM #100533
  O-jay8

  Hello Guys,
  Paul, thanks for posting this code. I really like it.
  Overall I did a little Analysis and the conditions C1, C2 and C5 are not improving the performance significantly. As Jebus89 already mentioned the strategy is already very profitable by just applying the time condition. In my opinion only rule c3 and C4 adding a substantial surplus which improves the Netprofit/DD ratio.
  See the chart attached.

  Overall I think it is important not to add too many rules and rise the level of complexity artificially to tweak it until the maximum.

  2 users thanked author for this post.

  Paul, GraHal

  

  O-jay8

  Participant

  Topics: 5

  Replies: 207

  Been thanked: 285 times
  06/13/2019 at 12:37 PM #100539
  stefou102

  C2 I agree with you.  C1 is important. C5 improves a bit.

  I tested also on EURJPY with a few changes in the parameters. Profit factor 1.55, so a little too low so far in order for me to test it live.

  

  stefou102

  Participant

  Topics: 22

  Replies: 171

  Been thanked: 71 times
  06/13/2019 at 12:49 PM #100542
  jebus89

  Using the same time in eurJPY as well? I thought i tried that without good results?

   

  Im running this live using 1 contract just to see what happens, the max DD is so low so its no worries for my account.

   

  Im using all the C rules, but as mentioned by me and other i think im going to review them a bit more and remove the not important ones as stated in earlier comment.

  

  jebus89

  Participant

  Topics: 118

  Replies: 586

  Been thanked: 238 times
  06/13/2019 at 1:24 PM #100545
  Magifina

  Condition C5 lowers the number of trades and drawdown. Without the condition drawdown is still acceptable. It could be interessant to test other filters that condition C5.

  

  Magifina

  Participant

  Topics: 2

  Replies: 41

  Been thanked: 12 times
  06/13/2019 at 1:31 PM #100547
  GraHal

  O-jay8 wrote:

  Overall I did a little Analysis

  Very interesting, thank you! Not seen this done before.

  1. Did it take long?
  2. Is it an analysis you do on other Systems or a first time one-off?
  3. If not a first time … are results generally similar to findings on the link above?

   

  

  GraHal

  Participant

  Topics: 204

  Replies: 8401

  Been thanked: 2042 times
  06/13/2019 at 1:35 PM #100550
  Magifina

  I’m trying to test the strategy with long trades. The results aren’t as good as short trades. Trading hours between 18h and 20h (UTC+2) improves a bit the strategy and maintain a smoothed equity curve. I’ve only made a test with 100k data. Do you still have an improvement with 200k data?

  EURUSD 15M Morning Entry LONG

  DEFPARAM CUMULATEORDERS = false CtimeAchat = time >= 180000 AND time < 200000 //LONGS if time=200000 then level=low[1] endif c1= close > dopen(0) c2= low > dlow(1) c3= level < dhigh(1) c4= level < dhigh(2) bu= BollingerUp[20](close) bd= BollingerDown[20](close) ba= (bu+bd)/2 c5= ba>ba[1] or low < bu if CtimeAchat and c1 and c2 and c3 and c4 and c5 THEN buy 1 contract at market endif if LongOnMarket and time=200000 THEN sell at market endif set stop %loss 0.5

  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

  DEFPARAM CUMULATEORDERS = false   CtimeAchat = time >= 180000 AND time < 200000   //LONGS if time=200000 then level=low[1] endif   c1= close > dopen(0) c2= low  > dlow(1) c3= level < dhigh(1) c4= level < dhigh(2)   bu= BollingerUp[20](close) bd= BollingerDown[20](close) ba= (bu+bd)/2   c5= ba>ba[1] or low < bu   if CtimeAchat and c1 and c2 and c3 and c4 and c5 THEN buy 1 contract at market endif   if LongOnMarket and time=200000 THEN sell at market endif   set stop %loss 0.5

   

  

  Magifina

  Participant

  Topics: 2

  Replies: 41

  Been thanked: 12 times
  06/13/2019 at 5:28 PM #100569
  Vonasi

  Sometimes I will have an optimized variable such as ‘c’ which is used purely to select a set of conditions to use in the back test. This way it is easy to test every possible combination side by side to see which ones have the most benefit. With a lot of conditions it can result in a lot of options but once you have the code then it is a simple thing just to change the c1, c2, c3 conditions to whatever your latest idea uses and paste it into your new strategy.

  Something like:

  //c = 0 //optimized variable c1 = open > close c2 = open > high[1] c3 = open < low[1] if c = 1 then condition = c1 and c2 and c3 endif if c = 2 then condition = c1 and c2 endif if c = 3 then condition = c1 and c3 endif if c = 4 then condition = c2 and c3 endif if condition then buy 1 contract 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

  //c = 0 //optimized variable   c1 = open > close c2 = open > high[1] c3 = open < low[1]   if c = 1 then condition = c1 and c2 and c3 endif   if c = 2 then condition = c1 and c2 endif   if c = 3 then condition = c1 and c3 endif   if c = 4 then condition = c2 and c3 endif   if condition then buy 1 contract at market endif

   

  2 users thanked author for this post.

  Paul, GraHal

  

  Vonasi

  Moderator

  Topics: 192

  Replies: 5088

  Been thanked: 1810 times
  06/13/2019 at 5:45 PM #100571
  Barney

  I started the algo live now but only 1 contract.

  More fun to run it live than on demo.

  DD with 1 contract is only 245 dollar..

  

  Barney

  Participant

  Topics: 9

  Replies: 47

  Been thanked: 7 times
  06/14/2019 at 10:13 AM #100633
  jebus89

  Im running this live as well with 1 contract barney. lets goo

  

  jebus89

  Participant

  Topics: 118

  Replies: 586

  Been thanked: 238 times
  06/14/2019 at 10:34 AM #100636
  GraHal

  Vonasi wrote:

  select a set of conditions to use in the back test.

  I added above code to here

  Snippet Link Library

  1 user thanked author for this post.

  Vonasi

  

  GraHal

  Participant

  Topics: 204

  Replies: 8401

  Been thanked: 2042 times
  06/14/2019 at 12:06 PM #100643
  stefou102

  Running it live too since yesterday. Watching it a bit, I think working on the 5min instead of the 15min (or MTF), and working with limit orders instead of at the market may improve the results. Will try

  

  stefou102

  Participant

  Topics: 22

  Replies: 171

  Been thanked: 71 times
  06/14/2019 at 1:18 PM #100649
  GraHal

  I’ve got 2 versions (Paul and Paul / Stefou) running on Demo Fwd Test. but not had a trade yesterday or today.

  Anybody had trades yesterday or today?

  

  GraHal

  Participant

  Topics: 204

  Replies: 8401

  Been thanked: 2042 times
  06/14/2019 at 1:19 PM #100650
  stefou102

  yes me. both days. a small profit and a small loss

  

  stefou102

  Participant

  Topics: 22

  Replies: 171

  Been thanked: 71 times
  06/14/2019 at 2:23 PM #100660
  Paul

  The breakeven is a pain to quickly test 200k bars.

  An alternative can be to close the position when the Bollingerbands distance are getting smaller.

  bu= BollingerUp[20](close) bd= BollingerDown[20](close) ba= (bu+bd)/2 br= (bu-bd)

  1 2 3 4

  bu= BollingerUp[20](close) bd= BollingerDown[20](close) ba= (bu+bd)/2 br= (bu-bd)

  if shortonmarket and time=141500 and positionperf(0)*100<0.3 then exitshort at market else if time> 141500 and br<br[1] and br[1]<br[2] and br[2]<br[3] and br[3]<br[4] then exitshort at market endif endif

  1 2 3 4 5 6 7

  if shortonmarket and time=141500 and positionperf(0)*100<0.3 then exitshort at market else if time> 141500 and br<br[1] and br[1]<br[2] and br[2]<br[3] and br[3]<br[4] then exitshort at market endif endif

  Found this quick, perhaps there’s room to improve without a breakeven!

  Small one,  lower the entrycriteria for the bollingerbands, and use previous bu, instead of the current

  c5= ba<ba[1] or high > bu[1]

  1	c5= ba<ba[1] or high > bu[1]

   

  

  Paul

  Participant

  Topics: 40

  Replies: 812

  Been thanked: 660 times
  06/14/2019 at 4:37 PM #100686
  O-jay8

  Hi Grahal,

  1. It can be quick if one prepared everything on Excel already. There you just need the final result from each added condition
  2. I just read it in the book (trading systems a new approach to system optimisattion from  Jaekle/Tomasini) Was the first time I did it and will continue do to so for future optimisations

  Overall I can only recommend to read that book for systematic trading, especially how to optimise

  Same is true for the book “Kevin Davey – Building Algorithmic Trading Systems”

  1 user thanked author for this post.

  GraHal

  

  O-jay8

  Participant

  Topics: 5

  Replies: 207

  Been thanked: 285 times
• Author
  Posts