re-examine the old strategies from the library

Viewing 14 posts - 16 through 29 (of 29 total)
  • Author
    Posts
  • #107685 quote
    solar
    Participant
    Senior

    I also found Francesco’s oil hammer negated strategy works consistently well on 15min TF. Here’s the link to it: https://www.prorealcode.com/prorealtime-trading-strategies/oil-10min-hammernegated-pattern-strategy/

    #107711 quote
    Stefanb
    Participant
    Senior

    @fatlung

    This has to be the best algo in the library?

    Francesco’s algo runs until July 2017.

    I only have data from March 2018, but from then to today it has gone as smoothly as in his BT.

    The difference is just having used it for 10 min.

    In 15 minutes it has gone really well.

    Spread=3

    oil.jpg oil.jpg
    #107713 quote
    Stefanb
    Participant
    Senior

    Think about starting it at the smallest possible size.

    Manually you can trade 0.25 contracts.

    What is the smallest size auto on US Crude?

    #107724 quote
    Vonasi
    Moderator
    Master

    I get the results shown in the first two images for Francesco’s oil hammer negated strategy on 10 minute and 15 minute time frames with a spread of 3.

    I tried playing with the bull and bear filters to give them a no trading band by changing them to:

    bull = low > av1 and low > av2 and low > av3
    bear = high < av1 and high < av2 and high < av3

    This change massively improved the 10 minute chart results but not the 15 minute. See second two images.

    I also ran a robustness test on my modified 10 minute version. The average gain per trade score is not great but the trade numbers are also very low so it is not a very conclusive test.

    My opinion would be that the OOS 10 minute results are a fail for the original strategy. It would be nice to see the 15 minute results prior to what we can currently see on the 100k chart and also for a 10 minute strategy with my bear and bull filter changes. Perhaps someone with 200k might be able to help?

    Screenshot_15-1.png Screenshot_15-1.png Screenshot_16-2.png Screenshot_16-2.png Screenshot_18-3.png Screenshot_18-3.png Screenshot_17-3.png Screenshot_17-3.png Screenshot_19-3.png Screenshot_19-3.png
    #107739 quote
    solar
    Participant
    Senior

    thanks Vonasi, everytime I see your analysis I learn something from you.

    The equity curve of the 15min version is really appealing, there are 2 problems on this strategy as i have run it for almost 2 months on demo.

    1. there is no hard stoploss, which means sometimes it can be hard to survive huge drawdown
    2. the ‘hammer negated’ pattern is not a really good entry point as average position duration is over 8 hours (but idk why it yields that appealing curve)
    #107772 quote
    Stefanb
    Participant
    Senior

    Am I crazy

    Don’t understand what you mean?

    Isn’t the 10 minute ok to start according to the picture I attached?

    Doesn’t look like you need to change anything.

    It has shown good results for 2 years after it was made.

    #107777 quote
    Vonasi
    Moderator
    Master

    Isn’t the 10 minute ok to start according to the picture I attached?

    The image you attached is on the 15 minute chart and not the 10 minute chart.

    #107801 quote
    Stefanb
    Participant
    Senior

    @Vonasi

    Sorry, i mean the 15 min attached foto of course.

    The code was made  a long time a go and the results is still ok!

    They’ve been fine for the last two years so why wouldn’t it continue for a while?

    #107825 quote
    Vonasi
    Moderator
    Master

    They’ve been fine for the last two years so why wouldn’t it continue for a while?

    So you took a strategy that had been developed and tested on a 10 minute time frame and decided to run it on a 15 minute time frame. I’m guessing that you back tested it on the different time frame first and the results were good or better than the 10 minute which is why you chose to go forward with the 15 instead of 10 minute time frame? Do you have any equity curves for the 15 minute strategy from prior to the data we can use today as they would be interesting to see?

    So why would it not work in the future? Just look at the OOS equity curve for the 10 minute strategy and it clearly didn’t. Perhaps you got lucky with the switch to 15 minutes or was the decision based on any sound reasoning that 15 minutes was better than 10? One benefit I guess would be the increased period of time available for the back test so the 15 minute strategy has a longer in sample test. The strategy is based on price action and if market structure changes then the strategy could just simply stop working overnight. It would also be nice if it worked on a lot more than just one instrument.

    Fingers crossed that you have found something that works better than Francesco originally discovered.

    #108314 quote
    PastaPesto
    Participant
    Veteran

    The Dayopen Straddle strategy https://www.prorealcode.com/prorealtime-trading-strategies/dayopen-straddle-for-dax-3-minute-timeframe/

    This strategy has so many different timeframes and settings in the forum so it is a bit confusing, but I put the 10 minutes settings in demo more than 6 mounths ago and it keeps on making money. I have no idea how the different versions are doing and maybe Im just lucky with this one. Does anyone else have tested this one or do you have any opinion about it?

    //-------------------------------------------------------------------------
    // Main code : Straddle DayOpen
    //-------------------------------------------------------------------------
    
    // common rules
    DEFPARAM CUMULATEORDERS = false
    DEFPARAM PRELOADBARS = 10000
    
    // optional
    ExtraTradeCriteria=1
    
    // positionsize and stops
    positionsize = 1
    sl = 1 // % Stoploss
    pt = 1.5 // % Profit Target
    ts = 0.40 // % MFETrailing
    
    // indicator settigns
    NOP=25 //number of points
    TimeOpen=090000
    
    // day & time rules
    ONCE entertime = TimeOpen
    ONCE lasttime = 100000
    ONCE closetime = 240000 // greater then 23.59 means it continues position overnight
    ONCE closetimeFriday=173000
    
    tt1 = time >= entertime
    tt2 = time <= lasttime
    
    tradetime = tt1 and tt2
    
    DayForbidden = 0 // 0=sunday
    df = dayofweek <> dayforbidden
    
    // setup number of trades intraday
    if IntradayBarIndex = 0 then
    longtradecounter = 0
    Shorttradecounter = 0
    Tradecounter=0
    endif
    
    // general criteria
    GeneralCriteria = tradetime and df
    
    // trade criteria
    tcLong = countoflongshares < 1 and longtradecounter < 1 and tradecounter <1
    tcShort = countofshortshares < 1 and shorttradecounter < 1 and tradecounter <1
    
    // indicator criteria
    If time = TimeOpen then
    DayOpen=open
    endif
    
    if IntradayBarIndex = 0 then
    lx=0
    sx=0
    endif
    
    if high > DayOpen+NOP then
    lx=1
    else
    lx=0
    endif
    
    if low < DayOpen-NOP then
    sx=1
    else
    sx=0
    endif
    
    // trade criteria extra
    min1 = MIN(dhigh(0),dhigh(1))
    min2 = MIN(dhigh(1),dhigh(2))
    
    max1 = MAX(dlow(0),dlow(1))
    max2 = MAX(dlow(1),dlow(2))
    
    If ExtraTradeCriteria then
    tcxLong = high < MIN(min1,min2)
    tcxShort = low > MAX(max1,max2)
    else
    tcxLong = high
    tcxShort = low
    endif
    
    // long entry
    If GeneralCriteria then
    if lx and tcLong and tcxLong then
    buy positionsize contract at market
    longtradecounter=longtradecounter + 1
    tradecounter=tradecounter+1
    endif
    endif
    
    // short entry
    If GeneralCriteria then
    if sx and tcShort and tcxShort then
    sellshort positionsize contract at market
    shorttradecounter=shorttradecounter + 1
    tradecounter=tradecounter+1
    endif
    endif
    
    // MFETrailing
    trailingstop = (tradeprice/100)*ts
    if not onmarket then
    MAXPRICE = 0
    MINPRICE = close
    priceexit = 0
    endif
    if longonmarket then
    MAXPRICE = MAX(MAXPRICE,close)
    if MAXPRICE-tradeprice(1)>=trailingstop*pipsize then
    priceexit = MAXPRICE-trailingstop*pipsize
    endif
    endif
    if shortonmarket then
    MINPRICE = MIN(MINPRICE,close)
    if tradeprice(1)-MINPRICE>=trailingstop*pipsize then
    priceexit = MINPRICE+trailingstop*pipsize
    endif
    endif
    If onmarket and priceexit>0 then
    sell at market
    exitshort at market
    endif
    
    // exit at closetime
    If onmarket then
    if time >= closetime then
    sell at market
    exitshort at market
    endif
    endif
    
    // exit friday at set closetime
    if onmarket then
    if (CurrentDayOfWeek=5 and time>=closetimefriday) then
    sell at market
    exitshort at market
    endif
    endif
    
    // build-in exit
    SET TARGET %PROFIT pt
    SET STOP %LOSS sl
    
    GRAPH 0 coloured(300,0,0) AS "zeroline"
    GRAPH (positionperf*100)coloured(0,0,0,255) AS "PositionPerformance"
    DayOpen-straddle-10min.jpg DayOpen-straddle-10min.jpg Straddle-10min.jpg Straddle-10min.jpg
    #108317 quote
    solar
    Participant
    Senior

    Thank you PastaPesto.

    Would you mind posting the report of this strategy on your demo account?

    #108318 quote
    PastaPesto
    Participant
    Veteran

    Hi Fatlung,

    I attached pictures of the report from demo. I hope this is what you want to see. It started in February and then I accidently stopped it in August, so the report is no longer than Aug 14th.

    solar and Edmond thanked this post
    Straddle-Demo.jpg Straddle-Demo.jpg Straddle-Demo-2.jpg Straddle-Demo-2.jpg
    #108328 quote
    GraHal
    Participant
    Master

    I added PastaPesto    results to here

    Forward Test – Sys Perf

    Are those results for Lot size = 1  @PastaPesto please??

    If not Lot size = 1 then what is the Lot size you used please? Only asking so the Log has a common baseline re performance.

    #108339 quote
    PastaPesto
    Participant
    Veteran

    @GraHal

    Lot size = 1.4

    GraHal thanked this post
Viewing 14 posts - 16 through 29 (of 29 total)
  • You must be logged in to reply to this topic.

re-examine the old strategies from the library


ProOrder: Automated Strategies & Backtesting

New Reply
Author
author-avatar
bluetime6 @bluetime6 Participant
Summary

This topic contains 28 replies,
has 8 voices, and was last updated by PastaPesto
6 years, 5 months ago.

Topic Details
Forum: ProOrder: Automated Strategies & Backtesting
Language: English
Started: 08/27/2019
Status: Active
Attachments: 19 files
Logo Logo
Loading...