Baby Steps – First Live System – DAX 5 Min

Forums ProRealTime English forum ProOrder support Baby Steps – First Live System – DAX 5 Min

Viewing 14 posts - 1 through 14 (of 14 total)
  • #134068

    Hello – I hope this finds everybody well.

    As I’ve just made my first foray into coding, after hanging around on the site for sometime over a year now, I thought I’d share and of course would welcome any feedback. By way on context, I’ve been fiddling around with trading (via a spread betting account with IG) for a couple of years. I’ve been up a bit, down a bit and a couple of months ago found myself pretty much back where I started, and struggling to find the time to concentrate properly what with a full time job and a young family. I’ve had several false starts with ProRealCode and never got as far as putting a system live – until about 3 weeks ago.

    The very simple system runs on DAX 5 minute chart and uses a 200 period EMA and SuperTrend as filters, and an ADX/ADXR crossover as entry trigger. Stops and profit target is fixed (about a 1 to 1.6 R/R). Back-testing showed it to be profitable based on a 100K bars test (starting in February 2019), but with a horrible curve – a fairly consistent losing streak early on, but consistently profitable since about the end of last July. Now, I know this is far from satisfactory for the purist, but recent performance was enough to persuade me to make it live, with low position size and guaranteed stops.

    I absolutely don’t expect this to run for ever – it’s going to stop working based on back test, but it’s given me the confidence that there’s something to build on for the future. Happily after 3 and a bit weeks of running, account is up about 18% and win rate is just the right side of 50%.

    Back test view attached along with account growth chart based on the 46 trades placed to date.

    Comments and advice welcome. I’m very new to this, and not even scratched the surface.

    Cheers

    WT

    3 users thanked author for this post.
    #145500

    Thanks for sharing your code.

    I think it will be useful to many and a source for new ideas 🙂

     

    #145628

    Any news from the recent results of your strategy? Thanks for sharing! 🙂

    #145658

    So, I ran this live for 2 months – 11th May to around 11th July. As per my original post, it started very well then began to tail off as market conditions changed. I lucked out really starting it at a good time. Win rate for the period was about 42% overall and it was a couple of hundred quid in profit when I stopped it. My intention has always been to go back and look at how I can refine it, probably initially looking at some multi-timeframe filters. Also keen to explore alternative exit (e.g. trailing stop) rather than fixed points target.

    Results from running in back-test from 11th July to current are very bad, but the market’s basically been going sideways. So it definitely needs some better filtering to look for trending periods.

    1 user thanked author for this post.
    #149062

    Hi Welshtrader, thank you for sharing your code. I’m a beginner, so please bear with me. I’ve added a few conditions (VWAP, KST and VWMA) to increase the WinRate to around 64% (Sept. 16th 2019 to date) with a R/R of x1.9 (Capital of 1500€, TF 5 minutes and Size = 1 DAX 1€ Contract). I’ve also copied the ML code from juanj (https://www.prorealcode.com/topic/machine-learning-in-proorder/) and the MFE from Nicolas (https://www.prorealcode.com/blog/learning/trailing-stop-max-favorable-excursion-mfe/).

    Walk Forward shows a Win Rate of 62.5% and R/R of x1.6.

    Not sure of the Tick by Tick mode in the backtest. 

    Please test and let me know…

    A big thank you to robertogozzi and Nicolas who guided me through my learning process.

     

    2 users thanked author for this post.
    #149076

    i get this error using the code…

    #149085

    Below the code I’ve used for VWAP courtesy of one of the members here. Let me know if you spot an error in the code.

     

    #149093

    Hello Haled

    Could you put you .ITF ?

    Thank you

    #149094

    V WAP … hmm, brings to mind a certain song title. And everything else. And now you’re all going to think of that every time you see that indicator, hahaha.

    1 user thanked author for this post.
    #149102

    Cardi. B? Can you please post the lycrics here? ))))))

    #149103

    No … you will be banned! 🙂

    I’d never heard of Cardi B or the song, but being an inquisitive chap I had to check it out! 🙂

    #149111

    Just checked, it’s Megan T. S. Anyway, everyone here undertands what VWAP stands for 🙂

    #149127

    What is Heuristics used for in your program.
    don’t use valueX

    1 user thanked author for this post.
    #149228

    fifi743

    Thank you for spotting this gross mistake. My bad!

    I’ve “tried” to correct it and run again backtest and WF.

    So, the initial version (ver 1 ) I posted (working witout ML) did generate in backtest  6574€ WinRate 64.23% and Profit Rate 1.9x. When running the WF, the results are 5509€ WinRate 60.96% and Profit Rate 1.5x.  (files below)

    The version with ML embedded (ver 3) did generate in backtest  8073€ WinRate 60.43% and Profit Rate 2.04x. When running the WF, the results are 6249€ WinRate 55.79% and Profit Rate 1.56x.  (files below + ITF)

    So, the results with ML appear to be better in normal backtest and in WF.

    Please review and let me know how we can further improve the result.

    1 user thanked author for this post.
Viewing 14 posts - 1 through 14 (of 14 total)

Create your free account now and post your request to benefit from the help of the community
Register or Login