1 Min Trading Strategy
Forums › ProRealTime English forum › ProOrder support › 1 Min Trading Strategy
- This topic has 219 replies, 29 voices, and was last updated 7 years ago by Leo.
-
-
09/27/2017 at 2:23 PM #4753209/27/2017 at 2:30 PM #4753409/27/2017 at 3:57 PM #4754409/27/2017 at 8:12 PM #4756909/27/2017 at 8:24 PM #47571
Check out these two sites, it has some statistics on all the different candlestick patterns:
09/27/2017 at 8:31 PM #47572To be checked.
That is the best I can get from my side…
Short version1234567891011121314151617181920212223242526272829303132333435363738394041424344454647484950515253545556575859606162636465666768697071727374757677DEFPARAM CumulateOrders = TrueDEFPARAM FLATBEFORE = 085500 // 85500DEFPARAM FLATAFTER = 200000noEntryBeforeTime = 085500timeEnterBefore = time >= noEntryBeforeTimenoEntryAfterTime = 170000timeEnterAfter = time < noEntryAfterTimedaysForbiddenEntry = OpenDayOfWeek = 6 OR OpenDayOfWeek = 0If longonmarket and close < positionprice and (barindex-tradeindex) > 650 ThenSell at marketElsIf shortonmarket and close > positionprice and (barindex-tradeindex) > 60 thenExitshort at marketEndIfbody=close-openabody=abs(body)if range>0 thenratio=abody/rangeelseratio=0endiflongcandle= (ratio>0.6)indicator1 = Average[1000](close)c1 = (close CROSSES UNDER indicator1)Bear = 0//Bearish SignalThreeInsideDown=(body[2]>0 and body[1]<0 and body<0 and close<close[1])if ThreeInsideDown thenBear = 1endifThreeBlackCrows=(body[2]<0 and body[1]<0 and body<0 and longcandle[2] and longcandle[1] and longcandle and low[1]<low[2] and low<low[1] and close[1]<close[2] and close<close[1] and open[1]<open[2] and open[1]>close[2] and open<open[1] and open>close[1])if ThreeBlackCrows thenBear = 1endif//Entry/Exit Criteriapossize = 1NearMA = Average[7,2](close)If countofposition = 0 and Bear = 1 and NearMA[1] < NearMA[3] and RSI[2](close) < 25 and NearMA < Average[150,2](close) and c1 and timeEnterBefore and timeEnterAfter and not daysForbiddenEntry Then//If shortonmarket Then//Exitshort at market//EndIfsellshort possize contract at marketEndIf//trailing stoptrailingstop = 10//resetting variables when no trades are on marketif not onmarket thenMINPRICE = closepriceexit = 0endif//case SHORT orderif shortonmarket thenMINPRICE = MIN(MINPRICE,close) //saving the MFE of the current tradeif tradeprice(1)-MINPRICE>=trailingstop*pointsize then //if the MFE is higher than the trailingstop thenpriceexit = MINPRICE+trailingstop*pointsize //set the exit price at the MFE + trailing stop price levelendifendif//exit on trailing stop price levelsif onmarket and priceexit>0 thenEXITSHORT AT priceexit STOPSELL AT priceexit STOPendifSet Target %Profit 0.8 // 0.42SET STOP %loss 0.34 // 0.4609/28/2017 at 4:19 PM #47695I played a little bit around with the short side and interestingly same as for the long side, thursdays are losing days. Mondays are much better short though with being second best (wednesdays are best).
No matter what, you don’t want to trade this strategy on thursdays, that’s for sure.
09/28/2017 at 8:15 PM #47728Okay folks, time for us to change gear. Let’s build a lean long and short scalper based on real research.
This guy has done some serious research on all known candlestick patterns. He has documented over 1000 candlestick patterns along with their stats.
Here is his Top 10 (I have already started coding some of them)
Bulkowski’s Top 10 Continuation Candlesticks:
http://thepatternsite.com/CandleContinue.htmlBulkowski’s Top 10 Reversal Candlesticks:
http://thepatternsite.com/CandleReverse.htmlBulkowski’s Top 10 Performing Candlesticks:
http://thepatternsite.com/CandlePerformers.html09/28/2017 at 8:57 PM #47731Here is something I coded to use as a platform for testing the different candlestick patterns.
12345678910111213141516171819202122232425262728293031323334Defparam cumulateorders = FalseIf hour > x and hour < x+y Thenpossize = 1If positionperf(1) < 0 and positionperf(2) < 0 Thenpossize = possize*2ElsIf positionperf(1) < 0 Thenpossize = possize*3ElsIf positionperf(1) > 0 and positionperf(2) > 0 and positionperf(3) > 0 Thenpossize = possize*3ElsIf positionperf(1) > 0 and positionperf(2) > 0 Thenpossize = possize*2EndIfElseIf longonmarket thensell at marketElsIf shortonmarket thenexitshort at marketEndIfpossize = 0EndIf//Three Line Strike Buillish Reversal//If countofposition = 0 and (close[3] < open[3] and low[2] < low[3] and close[2] < close[3] and low[1] < low[2] and close[1] < close[2] and low < low[1] and open < close[1] and close > high[3] and close < Average[20](close)) Then//Three Black Crows Bearish ContinuationIf countofposition = 0 and (close[2] < open[2] and ((high[2] - open[2]) <= (open[2]-close[2]/5)) and ((close[2] - low[2]) <= (open[2]-close[2]/5)) and open[1] > low[2] and close[1] < open[1] and ((high[1] - open[1]) <= (open[1]-close[1]/5)) and ((close[1] - low[1]) <= (open[1]-close[1]/5)) and close[1] < low[2] and open > low[1] and close < open and ((high - open) <= (open-close/5)) and ((close - low) <= (open-close/5)) and low < low[1] and close < Average[20](close)) ThenSellshort possize contract at marketEndIfIf positionperf(1) < 0 ThenSet Stop $Loss 90/possizeEndIfSet Target $Profit 30*possize09/28/2017 at 9:51 PM #47733@juanj The patternsite was “my dentist” at least for Dax, as he often already says patterns used to perform better in the past. Even when concentrating on the tops, I went better with breakout line patterns than with candle patterns. Just my experience.
09/30/2017 at 4:42 PM #47884As I mentioned before I have access to powerful computers with internet access that is online 24/7.
I did dome comparative tests. The conclusion is the more processing power does make some difference but the bottle neck is still data transfer from PRT.I can’t write code but I’m willing to help you with backtesting and WFA on these strategies.
Thank you.
09/30/2017 at 5:50 PM #47889@AVT I am not familiar with the expression ‘My Dentist’ but take it you meant you used to reference it regularly as an authoritative source.
I have found that much of his research still hold value today. So pattern reliability may have changed but I still use it on a daily basis as confirmation of my other signals.
09/30/2017 at 6:32 PM #47893I think there is still value in Bulkowskis work on both candlesticks and chart patterns but one has to keep in mind that he did all his research on stocks. Some of them might also apply to commodities and stock indices but some also might be plain wrong for other assets than stocks.
09/30/2017 at 6:49 PM #47896@Andre: I’m actually wondering where is done what when running a backtest. I was surprised what happened to me this week. I was running a backtest and then my internet failed so I was disconnected from the PRT server (happens since I live outside town little in the woods and still waiting for somebody to lay fiberoptic-cable to my house LOL). Anyway the backtest seemed unaffected and continued also while I was disconnected. I was amazed to see this since I thought all was done on the server side. So this would mean that backtest ARE run on ones computer and therefore should speed up with more computing power.
10/01/2017 at 1:06 PM #47920@juanj Sorry, it’s based on a phrase. Example: I have the idea, buy every 3rd day and sell 2 days later. Now you answer, due to statistics you are ruined within a month. “You pulled me this tooth” means you convinced me, that this idea does not work. If we play this several times, you might say, seems I am your dentist, next time I will invoice the treatment 😉
-
AuthorPosts
Find exclusive trading pro-tools on