Grid orders with one combined stop loss and limit, can it be done?

• 04/14/2016 at 8:38 PM #5409
  cfta

  Hi Traders,

  This is my first post in the forum so let me introduce myself. I discovered the user friendly and creative PRT platform a couple of months ago and made the switch from the more frustrating MT4 platform, I have been trading FX along with commodity and index CFDs for over 7 years and picked up a few good strategies along the way, one which I am going to share and ask for help coding.

  The strategy is based on scaling into positions while maintaining the same risk by placing stop orders in the direction of the trend and momentum of the selected instrument, for instance earlier this week I was anticipating that EUR/USD would drop and at the time the market price was 1.1385. I would start the system placing an order to sell 1 contract at market price along with several sell stop orders to sell additional contracts at predetermined intervals by 20 pips, 1.1365, 1.1345 and 1.1325 etc.

  The key for the strategy is to maintain the same risk as more contracts are added to the position, starting with a 1% risk on the initial market order taken at 1.1385 with 1% risk equaling 100 pips with a stop loss at 1.1485, once the second order is filled at 1.1365 the 1% risk is maintained by the stop loss being moved automatically to 1.1425 (50 pips from the average price at 1.1375), once the third order is filled at 1.1345 the 1% risk is maintaned by the stop loss being moved automatically to 1.1398 (33 pips from the new average price at 1.1365) and on it goes until reaching the desired number of contracts, in my case most often 5-10 contracts. In a perfect world I would like to include a function to set the stop loss to break even once reaching a certain percentage of profit but I have not figured that part out yet…

  If going for 5 contracts at 20 pip intervalls the first 100 pips yield a 3% return and evry additional 20 pips yield 1 additional percent, for instance given that pullbacks and whipsaws are moderate and do not cause an exit by the stop loss a 240 pip move yield a 10% profit with only 1% risk, I say that is pretty sweet R:R and MM 🙂

  I already have the code to do this on the frustrating and not so reliable MT4 platform but can it be coded for PRT?

  A few weeks ago I thought of course it can be done! Now I am a bit less convinced since I have made two requests to the inhouse ProRealTime programmers and been given two different sets of code but neither of them works as intended for my IG account so now I am asking all of you and in particular Nicolas if it es even possible to code this?

  This is the first code provided by the PRT programmers;

  stop1=round(close*1000)/1000 if stop1 <= close then stop1=stop1+0.001 endif tradesum=0 stop2=stop1+0.002 stop3=stop2+0.002 stop4=stop3+0.002 stop5=stop4+0.002 buy 1 share at stop1 stop buy 1 share at stop2 stop buy 1 share at stop3 stop buy 1 share at stop4 stop buy 1 share at stop5 stop if longonmarket then for i=0 to countoflongshares-1 do tradesum=tradesum+tradeprice(i) next averagetrade=tradesum/countoflongshares sell at averagetrade-((100/countoflongshares)*pointsize) stop 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

  stop1=round(close*1000)/1000   if stop1 <= close then   stop1=stop1+0.001   endif   tradesum=0   stop2=stop1+0.002   stop3=stop2+0.002   stop4=stop3+0.002   stop5=stop4+0.002   buy 1 share at stop1 stop   buy 1 share at stop2 stop   buy 1 share at stop3 stop   buy 1 share at stop4 stop   buy 1 share at stop5 stop   if longonmarket then   for i=0 to countoflongshares-1 do   tradesum=tradesum+tradeprice(i)   next   averagetrade=tradesum/countoflongshares   sell at averagetrade-((100/countoflongshares)*pointsize) stop   endif

  And here is the second code;

  DefParam CumulateOrders = True Once firstPrice = undefined // Change these settings accordingly Once lotSize = 10 Once orderLimit = 5 Once pipInterval = 10 Once profitPercentage = 0.2 Once lossPercentage = 0.5 If CountofLongShares[1]=0 and CountofLongShares=lotSize then firstPrice = TradePrice Elsif CountOfLongShares>lotSize then Sell At firstPrice Stop Endif If Not LongOnMarket then Buy lotSize Lots At Market Elsif CountOfLongShares<orderLimit*lotSize then Buy lotSize Lots At TradePrice+pipInterval*PointValue Stop Endif Set Target %Profit profitPercentage Set Stop %Loss lossPercentage

  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

  DefParam CumulateOrders = True   Once firstPrice = undefined   // Change these settings accordingly Once lotSize = 10 Once orderLimit = 5 Once pipInterval = 10 Once profitPercentage = 0.2 Once lossPercentage = 0.5     If CountofLongShares[1]=0 and CountofLongShares=lotSize then    firstPrice = TradePrice Elsif CountOfLongShares>lotSize then    Sell At firstPrice Stop Endif   If Not LongOnMarket then    Buy lotSize Lots At Market Elsif CountOfLongShares<orderLimit*lotSize then    Buy lotSize Lots At TradePrice+pipInterval*PointValue Stop Endif   Set Target %Profit profitPercentage Set Stop %Loss lossPercentage

  In the first code the SL function seemed to be working fine but not the order placement which tended to place all orders virtually instantly at the same price, in second code the orders were sometimes place at the proper intervalls and sometimes at the same price as a previous order.

  Please help me figure this out, I look forward to any replies and assistance!

   

  Regards, Filip

  1 user thanked author for this post.

  Adolfo

  

  cfta

  Participant

  Topics: 2

  Replies: 110

  Been thanked: 62 times
  04/14/2016 at 9:46 PM #5414
  Henrik

  Hi !

  Im trying to code a simular strategy and and having trouble to. So i would be very happy if someone can solve this problem.

  Regards, Henrik

  

  Henrik

  Participant

  Topics: 9

  Replies: 97

  Been thanked: 19 times
  04/15/2016 at 3:16 PM #5440
  Nicolas

  Hello,

  Welcome here and thanks for sharing your strategy with everyone.

  About your tests and bugs finding, have you spotted them in real market conditions or only in backtest please?

  So the 1% stoploss would be dynamically compute at each bar for the whole basket orders, isn’t it?

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20828

  Been thanked: 4739 times
  04/16/2016 at 10:55 AM #5482
  cfta

  Hi fellows,

  I appreciate your interest and fast replies.

  I have only been forward testing in real market conditions so far.

  Yes, I suppose so regarding the stoploss or alternatively move automatically to a new fixed level once a new order is triggerd.

  As mentioned above it seems that the first code is working better for the stoploss but the entry orders have been filled at the same or random levels…

   

   

  

  cfta

  Participant

  Topics: 2

  Replies: 110

  Been thanked: 62 times
  04/16/2016 at 1:05 PM #5487
  Nicolas

  Hi,

  I rewrote the whole thing for the sell side only. I think it should make it, but test it with minimal lot size under real market conditions please.

  I add 2 GRAPH instructions for displaying the new stoploss and the average price position of the whole basket for debugging purpose (see picture attached). Please have a look if it’s ok with your strategy.

  // first trade whatever condition if NOT ONMARKET AND close<close[1] then SELLSHORT 1 LOT AT MARKET endif // case SELL - add orders on the same trend if tradeprice(1)-close>20*pipsize then SELLSHORT 1 LOT AT MARKET endif // stoploss at 1% from the average price of the whole positions on market if shortonmarket then newSL = positionprice + (positionprice*0.01) if close>newSL then EXITSHORT AT MARKET endif GRAPH newsl coloured(200,0,0) as "stop loss 1%" GRAPH positionprice coloured(100,100,100) as "position price" endif

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

  // first trade whatever condition if NOT ONMARKET AND close<close[1] then SELLSHORT 1 LOT AT MARKET endif   // case SELL - add orders on the same trend if tradeprice(1)-close>20*pipsize then SELLSHORT 1 LOT AT MARKET endif   // stoploss at 1% from the average price of the whole positions on market if shortonmarket then newSL = positionprice + (positionprice*0.01)   if close>newSL then    EXITSHORT AT MARKET   endif GRAPH newsl coloured(200,0,0) as "stop loss 1%" GRAPH positionprice coloured(100,100,100) as "position price" endif

   

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20828

  Been thanked: 4739 times
  04/16/2016 at 2:14 PM #5489
  cfta

  Wow, you are really fast Nicolas, I’m impressed!

  Sure thing, I will test it in real market conditions first thing on Monday morning and come back with an update.

  Thanks a lot for your help and have a nice weekend.

   

  

  cfta

  Participant

  Topics: 2

  Replies: 110

  Been thanked: 62 times
  04/16/2016 at 5:37 PM #5499
  Henrik

  HI!

  This would work in small timeframe like 1 sec but not in 1h, 1d or 1w. The problem is that its hard to find god buy/sell conditions in small timeframes. How do you solve that? Im using IG version by the way. Is it possible to build indicators to simulate 15min, 1d, 1w price and then connect other inducaters to that  simulated price timeframe?

  Regards, Henrik

  

  Henrik

  Participant

  Topics: 9

  Replies: 97

  Been thanked: 19 times
  04/18/2016 at 6:39 PM #5597
  cfta

  Hello Nicolas,

  Unfortunately I lost my remote connection from work today so I have only been able to test for a couple of hours tonight. I learned that graphs are not allowed when testing in real market conditions so I did a slight modification to the code removing the graphs;

  // first trade whatever condition if NOT ONMARKET AND close<close[1] then SELLSHORT 1 LOT AT MARKET endif // case SELL - add orders on the same trend if tradeprice(1)-close>20*pipsize then SELLSHORT 1 LOT AT MARKET endif // stoploss at 1% from the average price of the whole positions on market if shortonmarket then newSL = positionprice + (positionprice*0.01) if close>newSL then EXITSHORT AT MARKET endif endif

  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

  // first trade whatever condition if NOT ONMARKET AND close<close[1] then SELLSHORT 1 LOT AT MARKET endif   // case SELL - add orders on the same trend if tradeprice(1)-close>20*pipsize then SELLSHORT 1 LOT AT MARKET endif   // stoploss at 1% from the average price of the whole positions on market if shortonmarket then newSL = positionprice + (positionprice*0.01) if close>newSL then EXITSHORT AT MARKET endif endif

  However I also did some backtesting and it seems like the code cause the positions to close at -1 % each instead of -1 % in total, see the attached screenshot, for the first set of trades with 10 sell orders the average price once all orders had been filled was 1.4162 and the-1 % in total SL should have kicked in at about 10 pips higher around 1.4172.

  Nicolas do you have an idea how to fix this please?

   

  

  cfta

  Participant

  Topics: 2

  Replies: 110

  Been thanked: 62 times
  04/18/2016 at 7:01 PM #5603
  Nicolas

  Sorry for graph instruction, i always forget to remove it when i send code to people … 🙂 It’s only for debugging purpose and not allowed by ProOrder.

  Well, I never used averaging down in real market conditions with IG. So I believe the problem is in the use of POSITIONPRICE which would return the average open price of the whole trades… Apparently not! I will ask PRT guys why.

  So I believe the solution is in the first code you post (from ProRealTime services), so remove the lines 11 to 19 from my code and add this instead:

  if shortonmarket then for i=0 to countofshortshares-1 do tradesum=tradesum+tradeprice(i) next averagetrade=tradesum/countofshortshares sell at averagetrade-((100/countofshortshares)*pointsize) stop endif

  1 2 3 4 5 6 7 8 9 10 11

  if shortonmarket then   for i=0 to countofshortshares-1 do   tradesum=tradesum+tradeprice(i) next   averagetrade=tradesum/countofshortshares   sell at averagetrade-((100/countofshortshares)*pointsize) stop   endif

   

   

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20828

  Been thanked: 4739 times
  04/18/2016 at 9:21 PM #5607
  cfta

  No worries I would probably forget it all the time too 🙂

  I updated the code with some minor modifications replacing “lots” with “shares” to match your code with the first one I got;

  // first trade whatever condition if NOT ONMARKET AND close<close[1] then SELLSHORT 1 SHARE AT MARKET endif // case SELL - add orders on the same trend if tradeprice(1)-close>20*pipsize then SELLSHORT 1 SHARE AT MARKET endif if shortonmarket then for i=0 to countofshortshares-1 do tradesum=tradesum+tradeprice(i) next averagetrade=tradesum/countofshortshares sell at averagetrade-((100/countofshortshares)*pointsize) stop endif

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

  // first trade whatever condition if NOT ONMARKET AND close<close[1] then SELLSHORT 1 SHARE AT MARKET endif   // case SELL - add orders on the same trend if tradeprice(1)-close>20*pipsize then SELLSHORT 1 SHARE AT MARKET endif   if shortonmarket then   for i=0 to countofshortshares-1 do tradesum=tradesum+tradeprice(i) next   averagetrade=tradesum/countofshortshares   sell at averagetrade-((100/countofshortshares)*pointsize) stop   endif

  In backtesting the problem remains, five orders triggered but not exited until way past the total -1 %. Could it be due to the first code being a long only and yours being short only? Or is necessary for line 1 – 9 in the first code to be included for the SL to work?

  

  cfta

  Participant

  Topics: 2

  Replies: 110

  Been thanked: 62 times
  04/18/2016 at 9:55 PM #5609
  Nicolas

  Maybe averagetade variable is not calculated correctly? Do the whole orders closed at the same time anyway? I’ll have a look by tomorrow.

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20828

  Been thanked: 4739 times
  04/19/2016 at 9:39 AM #5623
  Nicolas

  I have just tested once more, this time with long only orders, and set the exit at 1% and things go well under tester:

  // first trade whatever condition if NOT ONMARKET AND close>close[1] then BUY 1 LOT AT MARKET endif // case BUY - add orders on the same trend if close-tradeprice(1)>20*pipsize then BUY 1 LOT AT MARKET endif // stoploss at 1% from the average price of the whole positions on market if longonmarket then newSL = positionprice - (positionprice*0.01) if close<=newSL then SELL AT MARKET endif GRAPH newSL endif

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

  // first trade whatever condition if NOT ONMARKET AND close>close[1] then BUY 1 LOT AT MARKET endif   // case BUY - add orders on the same trend if close-tradeprice(1)>20*pipsize then BUY 1 LOT AT MARKET endif   // stoploss at 1% from the average price of the whole positions on market if longonmarket then newSL = positionprice - (positionprice*0.01) if close<=newSL then   SELL AT MARKET endif GRAPH newSL endif

   

  POSITIONPRICE is really the average open price of the whole opened orders.

   

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20828

  Been thanked: 4739 times
  04/19/2016 at 6:47 PM #5678
  cfta

  Thanks Nicolas, it looks like we are making a lot of progress.

  I only removed a zero from (positionprice*0.001), the entry orders seem to be working good except for a minor issue that sometimes when the system is started it buys two contracts instead of one, in addition occasionally the intervals are a bit larger around 30 instead of 20 pips I guess this is due to dependency to close of the current candle?

  If you have time it would be great if it is possible to a limit function to exit all positions for an instrument.

  I have begun testing your latest code in real time market conditions so we will see if it works and I will give an update tomorrow!

   

   

  

  cfta

  Participant

  Topics: 2

  Replies: 110

  Been thanked: 62 times
  04/19/2016 at 6:58 PM #5680
  Nicolas

  About the 30 pips instead of 20, it is obviously about the one time test only on each bar.

  And for the 2 contracts instead of 1 only at system start up, I have no clue right now… sorry. Maybe the entry position has to be better defined than a close versus close[1] ?

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20828

  Been thanked: 4739 times
  04/20/2016 at 7:47 PM #5716
  cfta

  Hey guys,

  I bought a new aparment today so I haven’t had as much time for testing as I wanted but I set up the system on NZD/JPY and Dax which have been running on real market conditions for about 24 hours now.

  The good news is that the entries seem to work very good with orders being filled at every 20 pips with small differencies. The less good news is that the SL seem to kick in too early with (positionprice*0.001) at only -0.1 % but kick in too late with (positionprice*0.01) with floating loss of CAD 200 on a USD 10 000 account. I may have missed something in the calculations but so far I can’t figure it out.

  NOTE! Due to my lack of experience of PRT and lack of coding knowledge I have not mentioned one of the most important features, since this strategy is meant to manage R:R and MM for manual trades, though taken automatically, it is crucial that no new entry orders are placed after a SL kicked in. The system must wait to be started again manually, how can this feature be added?

  

  cfta

  Participant

  Topics: 2

  Replies: 110

  Been thanked: 62 times
• Author
  Posts