New Strategy Idea In Development

• 06/18/2019 at 9:42 PM #100980
  juanj

  Hi Guys

  I love the idea of combining statistics and probability with my trading strategies so came up with the following idea;

  Considering a single timeframe and looking at how each individual candle is formed we will see that each candle starts with the opening price (previous closing price) then make both a high and a low and eventually settle with the closing price.

  This means that the majority of candles have some leg room between the opening price and it’s high and low. These wicks are usually referred to as the candle ‘shadow’

  So the basic idea here is to try and leverage this phenomenon by entering a trade at the opening price of a candle and scalping some points for a profit.

  To do this we need to first establish the statistical advantage or probability of the trade to succeed.

  This means we need to evaluate the current market condition by looking at some of the previous candles to see which direction to trade as well as how many points to target.

  We also need to take into consideration the spread and whether statistically speaking there are enough points to make the trade worthwhile.

  So here is what I have so far:

  Defparam cumulateorders = False //Timeframe: I am thinking that timeframes in the range of 5min to 1Day would work best //Market Hour Filter If (hour < 9 or (hour = 9 and minute <= 0)) or (hour > 17 or (hour = 17 and minute >= 0)) Then possize = 1 //Use 0 to enable filter If longonmarket Then Sell at market ElsIf shortonmarket Then Exitshort at market EndIf Else possize = 1 EndIf //variables Spread = 1 //Enter spread here (For pips use format 0.0001) SpreadVar = 4 // Spread variance factor is used to establish whether there is enough volatility taking into account the market spread (optimize for values between 3-7) ShadowVar = 0.9 // Shadow variance factor is used to account for error in shadow length calculation (optimize for values between 0.8-0.9) RiskReward = 3 // Risk to reward ratio used to calculate stop loss (optimize for values between 1-3) StdDev = 1.2 // Standard deviation used to filter outliers (optimize for values between 1-2) EvalPeriod = 6 // Evaluation period that will be used to make estimations (optimize for values between 3-9) //Initialize Variables FollowHighGreenA = 0 FollowLowGreenA = 0 FollowHighRedA = 0 FollowLowRedA = 0 FGA = 0 FRA = 0 //Establish the length of upper and lower shadows for candles following both red and green candles in the evaluation period For i = 1 to EvalPeriod Do If close[i+1] > open[i+1] Then FollowHighGreenA = high[i] - open[i] FollowLowGreenA = open[i] - low[i] FGA = FGA + 1 ElsIf close[i+1] < open[i+1] Then FollowHighRedA = high[i] - open[i] FollowLowRedA = open[i] - low[i] FRA = FRA + 1 EndIf Next //Calculate the raw average values of the shadows FollowHighGreenB = FollowHighGreenA/FGA FollowLowGreenB = FollowLowGreenA/FGA FollowHighRedB = FollowHighRedA/FRA FollowLowRedB = FollowLowRedA/FRA //Test each shadow again similar to above but filter(cut) outliers against STD devaition For ii = 1 to EvalPeriod Do If close[ii+1] > open[ii+1] Then If high[ii] - open[ii] <= (StdDev*STD(FollowHighGreenB)) Then FollowHighGreenC = high[ii] - open[ii] Else FollowHighGreenC = (StdDev*STD(FollowHighGreenB)) EndIf If open[ii] - low[ii] <= (StdDev*STD(FollowLowGreenB)) Then FollowLowGreenC = open[ii] - low[ii] Else FollowLowGreenC = (StdDev*STD(FollowLowGreenB)) EndIf ElsIf close[ii+1] < open[ii+1] Then If high[ii] - open[ii] <= (StdDev*STD(FollowHighRedB)) Then FollowHighRedC = high[ii] - open[ii] Else FollowHighRedC = (StdDev*STD(FollowHighRedB)) EndIf If open[ii] - low[ii] <= (StdDev*STD(FollowLowRedB)) Then FollowLowRedC = open[ii] - low[ii] Else FollowLowRedC = (StdDev*STD(FollowLowRedB)) EndIf EndIf Next //Re-calculate the average values of the shadows FollowHighGreen = FollowHighGreenC/FGA FollowLowGreen = FollowLowGreenC/FGA FollowHighRed = FollowHighRedC/FRA FollowLowRed = FollowLowRedC/FRA //Validate entry conditions and set targets and stops If close > open and FollowHighGreen > FollowLowGreen and ((high - open)*ShadowVar) > (spread*SpreadVar) and ((FollowHighGreen*ShadowVar)/RiskReward) > (spread * SpreadVar) Then Buy possize contract at market Set target pprofit FollowHighGreen*ShadowVar Set stop ploss (FollowHighGreen*ShadowVar)/RiskReward ElsIf close > open and FollowHighGreen < FollowLowGreen and ((open - low)*ShadowVar) > (spread*SpreadVar) and ((FollowLowGreen*ShadowVar)/RiskReward) > (spread * SpreadVar) Then Sellshort possize contract at market Set target pprofit FollowLowGreen*ShadowVar Set stop ploss (FollowLowGreen*ShadowVar)/RiskReward ElsIf close < open and FollowHighRed > FollowLowRed and ((high - open)*ShadowVar) > (spread*SpreadVar) and ((FollowHighRed*ShadowVar)/RiskReward) > (spread * Spreadvar) Then Buy possize contract at market Set target pprofit FollowHighRed*ShadowVar Set stop ploss (FollowHighRed*ShadowVar)/RiskReward ElsIf close < open and FollowHighRed < FollowLowRed and ((open - low)*ShadowVar) > (spread*SpreadVar) and ((FollowLowRed*ShadowVar)/RiskReward) > (spread * spreadvar) Then Sellshort possize contract at market Set target pprofit FollowLowRed*ShadowVar Set stop ploss (FollowLowRed*ShadowVar)/RiskReward 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 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109

  Defparam cumulateorders = False   //Timeframe: I am thinking that timeframes in the range of 5min to 1Day would work best //Market Hour Filter   If (hour < 9 or (hour = 9 and minute <= 0)) or (hour > 17 or (hour = 17 and minute >= 0)) Then possize = 1 //Use 0 to enable filter If longonmarket Then Sell at market ElsIf shortonmarket Then Exitshort at market EndIf Else possize = 1 EndIf   //variables   Spread = 1 //Enter spread here (For pips use format 0.0001) SpreadVar = 4 // Spread variance factor is used to establish whether there is enough volatility taking into account the market spread (optimize for values between 3-7) ShadowVar = 0.9 // Shadow variance factor is used to account for error in shadow length calculation (optimize for values between 0.8-0.9) RiskReward = 3 // Risk to reward ratio used to calculate stop loss (optimize for values between 1-3) StdDev = 1.2 // Standard deviation used to filter outliers (optimize for values between 1-2) EvalPeriod = 6 // Evaluation period that will be used to make estimations (optimize for values between 3-9)   //Initialize Variables   FollowHighGreenA = 0 FollowLowGreenA = 0 FollowHighRedA = 0 FollowLowRedA = 0 FGA = 0 FRA = 0   //Establish the length of upper and lower shadows for candles following both red and green candles in the evaluation period   For i = 1 to EvalPeriod Do If close[i+1] > open[i+1] Then FollowHighGreenA = high[i] - open[i] FollowLowGreenA = open[i] - low[i] FGA = FGA + 1 ElsIf close[i+1] < open[i+1] Then FollowHighRedA = high[i] - open[i] FollowLowRedA = open[i] - low[i] FRA = FRA + 1 EndIf Next   //Calculate the raw average values of the shadows   FollowHighGreenB = FollowHighGreenA/FGA FollowLowGreenB = FollowLowGreenA/FGA FollowHighRedB = FollowHighRedA/FRA FollowLowRedB = FollowLowRedA/FRA   //Test each shadow again similar to above but filter(cut) outliers against STD devaition   For ii = 1 to EvalPeriod Do If close[ii+1] > open[ii+1] Then If high[ii] - open[ii] <= (StdDev*STD(FollowHighGreenB)) Then FollowHighGreenC = high[ii] - open[ii] Else FollowHighGreenC = (StdDev*STD(FollowHighGreenB)) EndIf If open[ii] - low[ii] <= (StdDev*STD(FollowLowGreenB)) Then FollowLowGreenC = open[ii] - low[ii] Else FollowLowGreenC = (StdDev*STD(FollowLowGreenB)) EndIf ElsIf close[ii+1] < open[ii+1] Then If high[ii] - open[ii] <= (StdDev*STD(FollowHighRedB)) Then FollowHighRedC = high[ii] - open[ii] Else FollowHighRedC = (StdDev*STD(FollowHighRedB)) EndIf If open[ii] - low[ii] <= (StdDev*STD(FollowLowRedB)) Then FollowLowRedC = open[ii] - low[ii] Else FollowLowRedC = (StdDev*STD(FollowLowRedB)) EndIf EndIf Next   //Re-calculate the average values of the shadows   FollowHighGreen = FollowHighGreenC/FGA FollowLowGreen = FollowLowGreenC/FGA FollowHighRed = FollowHighRedC/FRA FollowLowRed = FollowLowRedC/FRA   //Validate entry conditions and set targets and stops   If close > open and FollowHighGreen > FollowLowGreen and ((high - open)*ShadowVar) > (spread*SpreadVar) and ((FollowHighGreen*ShadowVar)/RiskReward) > (spread * SpreadVar) Then Buy possize contract at market Set target pprofit FollowHighGreen*ShadowVar Set stop ploss (FollowHighGreen*ShadowVar)/RiskReward ElsIf close > open and FollowHighGreen < FollowLowGreen and ((open - low)*ShadowVar) > (spread*SpreadVar) and ((FollowLowGreen*ShadowVar)/RiskReward) > (spread * SpreadVar) Then Sellshort possize contract at market Set target pprofit FollowLowGreen*ShadowVar Set stop ploss (FollowLowGreen*ShadowVar)/RiskReward ElsIf close < open and FollowHighRed > FollowLowRed and ((high - open)*ShadowVar) > (spread*SpreadVar) and ((FollowHighRed*ShadowVar)/RiskReward) > (spread * Spreadvar) Then Buy possize contract at market Set target pprofit FollowHighRed*ShadowVar Set stop ploss (FollowHighRed*ShadowVar)/RiskReward ElsIf close < open and FollowHighRed < FollowLowRed and ((open - low)*ShadowVar) > (spread*SpreadVar) and ((FollowLowRed*ShadowVar)/RiskReward) > (spread * spreadvar) Then Sellshort possize contract at market Set target pprofit FollowLowRed*ShadowVar Set stop ploss (FollowLowRed*ShadowVar)/RiskReward EndIf

  Improvements would include:

  Better ways to sample past candles and arrive at a more accurate ‘average’

  Additional filters to enhance probability i.e. volatility filter

  2 users thanked author for this post.

  winnie37, GraHal

  

  juanj

  Participant

  Topics: 51

  Replies: 505

  Been thanked: 170 times
  06/18/2019 at 10:46 PM #100983
  GraHal

  Wow! Great ideas there juanj !

  I hope some of the  coding wizards help you out with this!

  I’m already looking forward to checking it out tomorrow!

  Thank You for sharing thus far

  

  GraHal

  Participant

  Topics: 222

  Replies: 9297

  Been thanked: 2462 times
  06/19/2019 at 11:52 AM #101040
  juanj

  Thanks, Grahal, I also hope some of the old timers can jump onboard. I know @Vonasi have done some good work on probability analysis as well.

  One of the presumptions I made with the above is that it is useful to consider the color of the candle preceding the candle being analyzed. But thinking about it again last night I am not so sure this is valid. Will have to do a couple of tests.

  

  juanj

  Participant

  Topics: 51

  Replies: 505

  Been thanked: 170 times
  06/19/2019 at 12:20 PM #101045
  GraHal

  Maybe colour + body size is the factor to look at and wick sizes and which side wick is the biggest. (You may already have done this as I’ve not had time to go over the code). 

  When manual trading one’s eyes take all these factors into account without even thinking. If one is asked why did you go Long at that point, often it’s like somebody asking … so why and how did you take a breath (breathe in)  just then! 🙂

  

  GraHal

  Participant

  Topics: 222

  Replies: 9297

  Been thanked: 2462 times
  06/19/2019 at 1:22 PM #101051
  juanj

  Just found a couple of mistakes in calculating the average as well as validating them against the spread. Here is the fix:

  Defparam cumulateorders = False //Timeframe: I am thinking that timeframes in the range of 5min to 1Day would work best //Market Hour Filter If (hour < 9 or (hour = 9 and minute <= 0)) or (hour > 17 or (hour = 17 and minute >= 0)) Then possize = 1 //Use 0 to enable filter If longonmarket Then Sell at market ElsIf shortonmarket Then Exitshort at market EndIf Else possize = 1 EndIf //variables Spread = 1 //Enter spread here (For pips use format 0.0001) SpreadVar = 4 // Spread variance factor is used to establish whether there is enough volatility taking into account the market spread (optimize for values between 3-7) ShadowVar = 0.9 // Shadow variance factor is used to account for error in shadow length calculation (optimize for values between 0.8-0.9) RiskReward = 3 // Risk to reward ratio used to calculate stop loss (optimize for values between 1-3) StdDev = 1.2 // Standard deviation used to filter outliers (optimize for values between 1-2) EvalPeriod = 6 // Evaluation period that will be used to make estimations (optimize for values between 3-9) //Initialize Variables FollowHighGreenA = 0 FollowLowGreenA = 0 FollowHighRedA = 0 FollowLowRedA = 0 FGA = 0 FRA = 0 //Establish the length of upper and lower shadows for candles following both red and green candles in the evaluation period For i = 1 to EvalPeriod Do If close[i+1] > open[i+1] Then FollowHighGreenA = FollowHighGreenA + (high[i] - open[i]) FollowLowGreenA = FollowLowGreenA + (open[i] - low[i]) FGA = FGA + 1 ElsIf close[i+1] < open[i+1] Then FollowHighRedA = FollowHighRedA + (high[i] - open[i]) FollowLowRedA = FollowLowRedA + (open[i] - low[i]) FRA = FRA + 1 EndIf Next //Calculate the raw average values of the shadows FollowHighGreenB = FollowHighGreenA/FGA FollowLowGreenB = FollowLowGreenA/FGA FollowHighRedB = FollowHighRedA/FRA FollowLowRedB = FollowLowRedA/FRA //Initialize Variables FollowHighGreenC = 0 FollowLowGreenC = 0 FollowHighRedC = 0 FollowLowRedC = 0 //Test each shadow again similar to above but filter(cut) outliers against STD devaition For ii = 1 to EvalPeriod Do If close[ii+1] > open[ii+1] Then If high[ii] - open[ii] <= (StdDev*STD(FollowHighGreenB)) Then FollowHighGreenC = FollowHighGreenC + (high[ii] - open[ii]) Else FollowHighGreenC = FollowHighGreenC + (StdDev*STD(FollowHighGreenB)) EndIf If open[ii] - low[ii] <= (StdDev*STD(FollowLowGreenB)) Then FollowLowGreenC = FollowLowGreenC + (open[ii] - low[ii]) Else FollowLowGreenC = FollowLowGreenC + (StdDev*STD(FollowLowGreenB)) EndIf ElsIf close[ii+1] < open[ii+1] Then If high[ii] - open[ii] <= (StdDev*STD(FollowHighRedB)) Then FollowHighRedC = FollowHighRedC + (high[ii] - open[ii]) Else FollowHighRedC = FollowHighRedC + (StdDev*STD(FollowHighRedB)) EndIf If open[ii] - low[ii] <= (StdDev*STD(FollowLowRedB)) Then FollowLowRedC = FollowLowRedC + (open[ii] - low[ii]) Else FollowLowRedC = FollowLowRedC + (StdDev*STD(FollowLowRedB)) EndIf EndIf Next //Re-calculate the average values of the shadows FollowHighGreen = FollowHighGreenC/FGA FollowLowGreen = FollowLowGreenC/FGA FollowHighRed = FollowHighRedC/FRA FollowLowRed = FollowLowRedC/FRA //Validate entry conditions and set targets and stops If close > open and FollowHighGreen > FollowLowGreen and (FollowHighGreen *ShadowVar) > (spread*SpreadVar) and ((FollowHighGreen*ShadowVar)/RiskReward) > (spread * SpreadVar) Then Buy possize contract at market Set target pprofit FollowHighGreen*ShadowVar Set stop ploss (FollowHighGreen*ShadowVar)/RiskReward ElsIf close > open and FollowHighGreen < FollowLowGreen and (FollowLowGreen *ShadowVar) > (spread*SpreadVar) and ((FollowLowGreen*ShadowVar)/RiskReward) > (spread * SpreadVar) Then Sellshort possize contract at market Set target pprofit FollowLowGreen*ShadowVar Set stop ploss (FollowLowGreen*ShadowVar)/RiskReward ElsIf close < open and FollowHighRed > FollowLowRed and (FollowHighRed *ShadowVar) > (spread*SpreadVar) and ((FollowHighRed*ShadowVar)/RiskReward) > (spread * Spreadvar) Then Buy possize contract at market Set target pprofit FollowHighRed*ShadowVar Set stop ploss (FollowHighRed*ShadowVar)/RiskReward ElsIf close < open and FollowHighRed < FollowLowRed and (FollowLowRed *ShadowVar) > (spread*SpreadVar) and ((FollowLowRed*ShadowVar)/RiskReward) > (spread * spreadvar) Then Sellshort possize contract at market Set target pprofit FollowLowRed*ShadowVar Set stop ploss (FollowLowRed*ShadowVar)/RiskReward 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 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116

  Defparam cumulateorders = False //Timeframe: I am thinking that timeframes in the range of 5min to 1Day would work best //Market Hour Filter If (hour < 9 or (hour = 9 and minute <= 0)) or (hour > 17 or (hour = 17 and minute >= 0)) Then possize = 1 //Use 0 to enable filter If longonmarket Then Sell at market ElsIf shortonmarket Then Exitshort at market EndIf Else possize = 1 EndIf //variables Spread = 1 //Enter spread here (For pips use format 0.0001) SpreadVar = 4 // Spread variance factor is used to establish whether there is enough volatility taking into account the market spread (optimize for values between 3-7) ShadowVar = 0.9 // Shadow variance factor is used to account for error in shadow length calculation (optimize for values between 0.8-0.9) RiskReward = 3 // Risk to reward ratio used to calculate stop loss (optimize for values between 1-3) StdDev = 1.2 // Standard deviation used to filter outliers (optimize for values between 1-2) EvalPeriod = 6 // Evaluation period that will be used to make estimations (optimize for values between 3-9) //Initialize Variables FollowHighGreenA = 0 FollowLowGreenA = 0 FollowHighRedA = 0 FollowLowRedA = 0 FGA = 0 FRA = 0 //Establish the length of upper and lower shadows for candles following both red and green candles in the evaluation period For i = 1 to EvalPeriod Do If close[i+1] > open[i+1] Then FollowHighGreenA = FollowHighGreenA + (high[i] - open[i]) FollowLowGreenA = FollowLowGreenA  + (open[i] - low[i]) FGA = FGA + 1 ElsIf close[i+1] < open[i+1] Then FollowHighRedA = FollowHighRedA  + (high[i] - open[i]) FollowLowRedA = FollowLowRedA  + (open[i] - low[i]) FRA = FRA + 1 EndIf Next //Calculate the raw average values of the shadows FollowHighGreenB = FollowHighGreenA/FGA FollowLowGreenB = FollowLowGreenA/FGA FollowHighRedB = FollowHighRedA/FRA FollowLowRedB = FollowLowRedA/FRA   //Initialize Variables FollowHighGreenC = 0 FollowLowGreenC = 0 FollowHighRedC = 0 FollowLowRedC = 0 //Test each shadow again similar to above but filter(cut) outliers against STD devaition For ii = 1 to EvalPeriod Do If close[ii+1] > open[ii+1] Then If high[ii] - open[ii] <= (StdDev*STD(FollowHighGreenB)) Then FollowHighGreenC = FollowHighGreenC  + (high[ii] - open[ii]) Else FollowHighGreenC = FollowHighGreenC + (StdDev*STD(FollowHighGreenB)) EndIf If open[ii] - low[ii] <= (StdDev*STD(FollowLowGreenB)) Then FollowLowGreenC = FollowLowGreenC  + (open[ii] - low[ii]) Else FollowLowGreenC = FollowLowGreenC + (StdDev*STD(FollowLowGreenB)) EndIf ElsIf close[ii+1] < open[ii+1] Then If high[ii] - open[ii] <= (StdDev*STD(FollowHighRedB)) Then FollowHighRedC = FollowHighRedC + (high[ii] - open[ii]) Else FollowHighRedC = FollowHighRedC + (StdDev*STD(FollowHighRedB)) EndIf If open[ii] - low[ii] <= (StdDev*STD(FollowLowRedB)) Then FollowLowRedC = FollowLowRedC  + (open[ii] - low[ii]) Else FollowLowRedC = FollowLowRedC  + (StdDev*STD(FollowLowRedB)) EndIf EndIf Next //Re-calculate the average values of the shadows FollowHighGreen = FollowHighGreenC/FGA FollowLowGreen = FollowLowGreenC/FGA FollowHighRed = FollowHighRedC/FRA FollowLowRed = FollowLowRedC/FRA //Validate entry conditions and set targets and stops If close > open and FollowHighGreen > FollowLowGreen and (FollowHighGreen *ShadowVar) > (spread*SpreadVar) and ((FollowHighGreen*ShadowVar)/RiskReward) > (spread * SpreadVar) Then Buy possize contract at market Set target pprofit FollowHighGreen*ShadowVar Set stop ploss (FollowHighGreen*ShadowVar)/RiskReward ElsIf close > open and FollowHighGreen < FollowLowGreen and (FollowLowGreen *ShadowVar) > (spread*SpreadVar) and ((FollowLowGreen*ShadowVar)/RiskReward) > (spread * SpreadVar) Then Sellshort possize contract at market Set target pprofit FollowLowGreen*ShadowVar Set stop ploss (FollowLowGreen*ShadowVar)/RiskReward ElsIf close < open and FollowHighRed > FollowLowRed and (FollowHighRed *ShadowVar) > (spread*SpreadVar) and ((FollowHighRed*ShadowVar)/RiskReward) > (spread * Spreadvar) Then Buy possize contract at market Set target pprofit FollowHighRed*ShadowVar Set stop ploss (FollowHighRed*ShadowVar)/RiskReward ElsIf close < open and FollowHighRed < FollowLowRed and (FollowLowRed *ShadowVar) > (spread*SpreadVar) and ((FollowLowRed*ShadowVar)/RiskReward) > (spread * spreadvar) Then Sellshort possize contract at market Set target pprofit FollowLowRed*ShadowVar Set stop ploss (FollowLowRed*ShadowVar)/RiskReward EndIf

   

  

  juanj

  Participant

  Topics: 51

  Replies: 505

  Been thanked: 170 times
  06/19/2019 at 2:31 PM #101056
  juanj

  So preliminary tests seem to show more robust performance when not considering candle colors and only looking at the average shadow lengths. I would appreciate any ideas on better sampling data to determine trade direction and profit targets.

  Defparam cumulateorders = False //Timeframe: I am thinking that timeframes in the range of 5min to 1Day would work best //Market Hour Filter If (hour < 9 or (hour = 9 and minute <= 0)) or (hour > 17 or (hour = 17 and minute >= 0)) Then possize = 1 //Use 0 to enable filter If longonmarket Then Sell at market ElsIf shortonmarket Then Exitshort at market EndIf Else possize = 1 EndIf //variables Spread = 1 //Enter spread here (For pips use format 0.0001) SpreadVar = 4 // Spread variance factor is used to establish whether there is enough volatility taking into account the market spread (optimize for values between 3-7) ShadowVar = 0.9 // Shadow variance factor is used to account for error in shadow length calculation (optimize for values between 0.8-0.9) RiskReward = 3 // Risk to reward ratio used to calculate stop loss (optimize for values between 1-3) StdDev = 1.2 // Standard deviation used to filter outliers (optimize for values between 1-2) EvalPeriod = 6 // Evaluation period that will be used to make estimations (optimize for values between 3-9) //Initialize Variables HighA = 0 LowA = 0 //Establish the length of upper and lower shadows for candles following both red and green candles in the evaluation period For i = 1 to EvalPeriod Do HighA = HighA + (high[i] - open[i]) LowA = LowA + (open[i] - low[i]) Next //Calculate the raw average values of the shadows HighB = HighA/EvalPeriod LowB = LowA/EvalPeriod //Initialize Variables HighC = 0 LowC = 0 //Test each shadow again similar to above but filter(cut) outliers against STD devaition For ii = 1 to EvalPeriod Do If high[ii]-open[ii] <= (StdDev*STD(HighB)) Then HighC = HighC + (high[ii]-open[ii]) Else HighC = HighC + (StdDev*STD(HighB)) EndIf If open[ii] - low[ii] <= (StdDev*STD(LowB)) Then LowC = LowC + (open[ii] - low[ii]) Else LowC = LowC + (StdDev*STD(LowB)) EndIf Next //Re-calculate the average values of the shadows HighShadow = HighC/EvalPeriod LowShadow = LowC/EvalPeriod //Validate entry conditions and set targets and stops If HighShadow > LowShadow and (HighShadow*ShadowVar) > (spread*SpreadVar) and ((HighShadow*ShadowVar)/RiskReward) > (spread * SpreadVar) Then Buy possize contract at market Set target pprofit HighShadow*ShadowVar Set stop ploss (HighShadow*ShadowVar)/RiskReward ElsIf HighShadow < LowShadow and (LowShadow*ShadowVar) > (spread*SpreadVar) and ((LowShadow*ShadowVar)/RiskReward) > (spread * SpreadVar) Then Sellshort possize contract at market Set target pprofit LowShadow*ShadowVar Set stop ploss (LowShadow*ShadowVar)/RiskReward 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 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78

  Defparam cumulateorders = False //Timeframe: I am thinking that timeframes in the range of 5min to 1Day would work best //Market Hour Filter If (hour < 9 or (hour = 9 and minute <= 0)) or (hour > 17 or (hour = 17 and minute >= 0)) Then possize = 1 //Use 0 to enable filter If longonmarket Then Sell at market ElsIf shortonmarket Then Exitshort at market EndIf Else possize = 1 EndIf //variables Spread = 1 //Enter spread here (For pips use format 0.0001) SpreadVar = 4 // Spread variance factor is used to establish whether there is enough volatility taking into account the market spread (optimize for values between 3-7) ShadowVar = 0.9 // Shadow variance factor is used to account for error in shadow length calculation (optimize for values between 0.8-0.9) RiskReward = 3 // Risk to reward ratio used to calculate stop loss (optimize for values between 1-3) StdDev = 1.2 // Standard deviation used to filter outliers (optimize for values between 1-2) EvalPeriod = 6 // Evaluation period that will be used to make estimations (optimize for values between 3-9) //Initialize Variables HighA = 0 LowA = 0 //Establish the length of upper and lower shadows for candles following both red and green candles in the evaluation period For i = 1 to EvalPeriod Do HighA = HighA + (high[i] - open[i]) LowA = LowA + (open[i] - low[i]) Next //Calculate the raw average values of the shadows HighB = HighA/EvalPeriod LowB = LowA/EvalPeriod //Initialize Variables HighC = 0 LowC = 0 //Test each shadow again similar to above but filter(cut) outliers against STD devaition For ii = 1 to EvalPeriod Do If high[ii]-open[ii] <= (StdDev*STD(HighB)) Then HighC = HighC + (high[ii]-open[ii]) Else HighC = HighC + (StdDev*STD(HighB)) EndIf If open[ii] - low[ii] <= (StdDev*STD(LowB)) Then LowC = LowC + (open[ii] - low[ii]) Else LowC = LowC + (StdDev*STD(LowB)) EndIf Next //Re-calculate the average values of the shadows HighShadow = HighC/EvalPeriod LowShadow = LowC/EvalPeriod //Validate entry conditions and set targets and stops If HighShadow > LowShadow and (HighShadow*ShadowVar) > (spread*SpreadVar) and ((HighShadow*ShadowVar)/RiskReward) > (spread * SpreadVar) Then Buy possize contract at market Set target pprofit HighShadow*ShadowVar Set stop ploss (HighShadow*ShadowVar)/RiskReward ElsIf HighShadow < LowShadow and (LowShadow*ShadowVar) > (spread*SpreadVar) and ((LowShadow*ShadowVar)/RiskReward) > (spread * SpreadVar) Then Sellshort possize contract at market Set target pprofit LowShadow*ShadowVar Set stop ploss (LowShadow*ShadowVar)/RiskReward EndIf

   

  

  juanj

  Participant

  Topics: 51

  Replies: 505

  Been thanked: 170 times
• Author
  Posts