Hang seng 200K bars

• 04/05/2018 at 8:31 PM #67349
  Francesco78

  Guys, I have an apparently nice strategy on Hang seng that is purely driven by time bias.

  Since I am not very confident with the instrument itself, I would be extremely grateful if someone could backtest it with 200K bars and see how the OOS behave.

  The code is the following

  Defparam cumulateorders = false golong =1 goshort = 1 timegolong = 110000 timeexitlong = 60000 filterdaylong = dayofweek = 1 or dayofweek = 2 or dayofweek = 3 filterpricelong = close < (Dhigh(1)+dlow(2))/2 timegoshort = 30000 timeexitshort = 80000 filterdayshort = dayofweek = 2 or dayofweek = 4 filterpriceshort = close> (Dhigh(1)+dlow(2))/2 if goshort then if time = timegoshort and filterpriceshort and filterdayshort then sellshort 1 contract at market endif if time = timeexitshort then exitshort at market endif endif if golong then if time= timegolong and filterdaylong and filterpricelong then buy 1 contract at market endif if time = timeexitlong then sell at market endif endif set target pprofit 600 set stop ploss 700

  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

  Defparam cumulateorders = false   golong =1 goshort = 1   timegolong  = 110000 timeexitlong  = 60000 filterdaylong = dayofweek = 1 or dayofweek = 2 or dayofweek = 3 filterpricelong = close < (Dhigh(1)+dlow(2))/2   timegoshort = 30000 timeexitshort = 80000 filterdayshort = dayofweek = 2  or dayofweek = 4 filterpriceshort = close> (Dhigh(1)+dlow(2))/2     if goshort then   if time = timegoshort and filterpriceshort and filterdayshort   then sellshort 1 contract at market endif   if time = timeexitshort then exitshort  at market endif endif   if golong then   if time= timegolong  and filterdaylong  and filterpricelong  then buy 1 contract at market endif if time = timeexitlong then sell  at market endif endif   set target pprofit 600 set stop ploss 700

   

  I attach the equity curve.

   

  Thanks a lot in advance.

  Francesco

  1 user thanked author for this post.

  Nicolas

  

  Francesco78

  Participant

  Topics: 43

  Replies: 242

  Been thanked: 77 times
  04/05/2018 at 8:54 PM #67352
  Henrik

  Hi!

  Here is 200k bars.

  Regards Henrik

  1 user thanked author for this post.

  Francesco78

  

  Henrik

  Participant

  Topics: 9

  Replies: 97

  Been thanked: 19 times
  04/05/2018 at 9:01 PM #67356
  Vonasi

  Well ALE it looks like you have written a strategy that is a very good advert for getting PRT Premium with 200K bars!

  Time based strategies are not worth pursuing in my experience but I have to admit that you’d have been very happy to have traded this one over the last 100k bars. The equity curve is beautiful!

  1 user thanked author for this post.

  Francesco78

  

  Vonasi

  Moderator

  Topics: 192

  Replies: 5088

  Been thanked: 1812 times
  04/05/2018 at 9:38 PM #67358
  Francesco78

  Ok thank you. How about this one on eurusd?

  Defparam cumulateorders = false golong =1 goshort = 1 timegolong = 140000 timeexitlong = 70000 filterdaylong = dayofweek = 1 or dayofweek = 2 or dayofweek = 3 filterpricelong = close < Dhigh(1)//(Dhigh(1)+dlow(2))/2 timegoshort = 70000 timeexitshort = 140000 filterdayshort = dayofweek <>0 filterpriceshort = 1//close>Dlow(1) if goshort then if time = timegoshort and filterpriceshort and filterdayshort then sellshort 1 contract at market endif if time = timeexitshort then exitshort at market endif endif if golong then if time= timegolong and filterdaylong and filterpricelong then buy 1 contract at market endif if time = timeexitlong then sell at market 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

  Defparam cumulateorders = false   golong =1 goshort = 1   timegolong  = 140000 timeexitlong  = 70000 filterdaylong =  dayofweek = 1 or dayofweek = 2 or dayofweek = 3 filterpricelong = close < Dhigh(1)//(Dhigh(1)+dlow(2))/2   timegoshort = 70000 timeexitshort = 140000 filterdayshort = dayofweek <>0 filterpriceshort = 1//close>Dlow(1)     if goshort then   if time = timegoshort and filterpriceshort and filterdayshort   then sellshort 1 contract at market endif   if time = timeexitshort then exitshort  at market endif endif   if golong then   if time= timegolong  and filterdaylong  and filterpricelong  then buy 1 contract at market endif if time = timeexitlong then sell  at market endif endif

   

  

  Francesco78

  Participant

  Topics: 43

  Replies: 242

  Been thanked: 77 times
  04/05/2018 at 10:00 PM #67364
  Vonasi

  Vonasi wrote:

  Well ALE

  Oooops – sorry Francesco78. I shouldn’t reply after a couple of  glasses of wine!

  

  Vonasi

  Moderator

  Topics: 192

  Replies: 5088

  Been thanked: 1812 times
  04/05/2018 at 10:10 PM #67365
  Vonasi

  Here is the 200K with no spread.

  

  Vonasi

  Moderator

  Topics: 192

  Replies: 5088

  Been thanked: 1812 times
  04/05/2018 at 10:18 PM #67367
  Vonasi

  I can’t see a reason why you would need tick by tick on for that strategy so here is the result without tick by tick. First image is with no spread and second is with spread of just 1 (I had to change the start capital to 100k to get through the drawdown).

  The interesting thing is that around 2005 it suddenly becomes a winning strategy. Around 2004 automated trading and HFT started to come online so that may be a link to the sudden improvement in performance?

  

  Vonasi

  Moderator

  Topics: 192

  Replies: 5088

  Been thanked: 1812 times
  04/05/2018 at 10:30 PM #67372
  Vonasi

  Your strategy if only used since beginning of 2005. Walk Forward 70/50 x 5.

  Personally I think that the drawdown is a bit big for the gain/loss ratio but at least the right side of the equity curve is much higher than the left!

  

  Vonasi

  Moderator

  Topics: 192

  Replies: 5088

  Been thanked: 1812 times
  04/06/2018 at 6:50 AM #67382
  Nicolas

  Vonasi wrote:

  I shouldn’t reply after a couple of  glasses of wine!

  What is the alcohol level allowed to drive a boat? 😆

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20872

  Been thanked: 4799 times
  04/06/2018 at 7:07 AM #67383
  Vonasi

  Nicolas wrote:

  What is the alcohol level allowed to drive a boat?

  We don’t drive – we sail!

  The level is unlimited – it is called the freedom of the sea – the only restriction is that you have to wait until the sun is under the yardarm. Most sailors have height adjustable yardarms to deal with this issue. Anyway at the moment I am house sitting for friends in the hills overlooking Kalamata bay so there is definitely no restrictions on intake!

  

  Vonasi

  Moderator

  Topics: 192

  Replies: 5088

  Been thanked: 1812 times
  04/06/2018 at 1:41 PM #67444
  Francesco78

  Vonasi wrote:

  I can’t see a reason why you would need tick by tick on for that strategy so here is the result without tick by tick. First image is with no spread and second is with spread of just 1 (I had to change the start capital to 100k to get through the drawdown). The interesting thing is that around 2005 it suddenly becomes a winning strategy. Around 2004 automated trading and HFT started to come online so that may be a link to the sudden improv

  without tick to tick the results are much worst? That is puzzling. I guess the more realistic one are the 1 with tick to tick.

   

  

  Francesco78

  Participant

  Topics: 43

  Replies: 242

  Been thanked: 77 times
  04/06/2018 at 2:34 PM #67448
  Vonasi

  Francesco78 wrote:

  without tick to tick the results are much worst? That is puzzling

  That is very puzzling as you only trade at the close of a candle/open of the next candle so tick by tick should make no difference at all.

  

  Vonasi

  Moderator

  Topics: 192

  Replies: 5088

  Been thanked: 1812 times
  04/06/2018 at 3:53 PM #67459
  Vonasi

  It just occurred to me that I may be on a different time zone to you so my tests may not be the same as yours. I am trading on UTC+1 British Summertime.

  

  Vonasi

  Moderator

  Topics: 192

  Replies: 5088

  Been thanked: 1812 times
  04/06/2018 at 4:37 PM #67463
  Francesco78

  Vonasi wrote:

  It just occurred to me that I may be on a different time zone to you so my tests may not be the same as yours. I am trading on UTC+1 British Summertime.

  Yep I am in UTC + 2 any way this strategy perform quite well, 10 year OOS. Should be taken in consideration, what do you think?

  

  Francesco78

  Participant

  Topics: 43

  Replies: 242

  Been thanked: 77 times
  04/07/2018 at 7:14 AM #67478
  Vonasi

  Francesco78 wrote:

  what do you think?

  My main concern is that I don’t understand why it should work. You are basically shorting the EURUSD every day during the European session and going long during the US and Asian sessions on pull backs but only on Monday, Tuesday and Wednesday. I cannot think why that should always work. Also you have to factor in that you are holding overnight for up to three days a week so there will be overnight fees to cut into the already quite small gain/loss ratio.

  I have not tested it on your time zone yet but will try to later today as I have rather a lot on this morning.

  1 user thanked author for this post.

  Francesco78

  

  Vonasi

  Moderator

  Topics: 192

  Replies: 5088

  Been thanked: 1812 times
• Author
  Posts