Discussing the strategy VECTORIAL DAX (M5)

Forums ProRealTime English forum ProOrder support Discussing the strategy VECTORIAL DAX (M5)

Viewing 15 posts - 1,006 through 1,020 (of 1,239 total)
  • #148153

    @volpiemanuele

    Thanks you for the file🙏.

    I will try it this afternoon.👍

    #148231

    which parameters would you then optimise?

    There are many parameters, but with walk forward I focussed on the rsi stochastic, only 2 parameters, 10-50k bars, unlinked with 3 repeats. Whatever you try, likely the last week will be neutral to very bad.

    @ichimoku18 yes your right that was the intend, changing it would make it worse, so I leave it at current & previous bar and removed the last part. It could be better written too.

    2 users thanked author for this post.
    #148233

    Great, in particular which are these two parameters to optimize?

    Thank you

    #148235

    Hi,

    only the first 2. Having said that i’am just a bit experimenting.

    dax 10s today. I know it’s not realistic with the stops etc and a lot would needs to be changed, but still.

     

    #148242

    @Paul, juste for fun i launch the v5b on Dax in 10s timeframe, but il crashed.

    #148244

    I use Stoch/RSI a lot, it’s one of my favourite indicators, but I find that you really have to optimise all the variables as each one reacts to other changes. With most indicators you usually have a reasonable expectation that if a value of 14 is good, then 12 or 16 will be a bit better or a bit worse. With Stoch/RSI values of 12, 14, 8, 2 could look good one minute then suddenly something like 8,6,8,4 is miles better – impossible to predict the optimal result. Default values are 14,14,10,3 as a theoretical starting point.

    I usually run it with variations of 4 to 16 by 2 for the first 3, and 2 to 5 by 1 for smoothd  (must be >1). Unfortunately that makes +1000 permutations so it’s time consuming, but almost always worth it – can massively improve almost any algo if you get it right.

    1 user thanked author for this post.
    #148245

    @Volpiemanuele, I tried to implement it on DJ 3 minutes and no orders showed up… Any ideas why?

    #148247

    @tanou: The last entry was this morning at 5am and still in progress … my live account

    #148248

    You simply implement it / Backtest it on Dow Jones 3 minutes or Dax ? Which UT?

     

    Here what appears when I backtest it in 200k candles / DJ / 3min, nothing…

    #148253

    Restart the platform because it works, I just made a BT

    #148269

    It’s working now, it was coming from the positionsize

    #148270

    Do not hesitate to give me your feedback. I’ve worked on this, I haven’t fully optimize it for now, just had a look at the mistakes it was doing and try to eliminate them with a simple Ichimoku.

     


    @Nonetheless
    & @Paul, hope you see something interesting, I think it is

     

    going live for me

    4 users thanked author for this post.
    #148272

    Nice. But the profit is in direct comparison to the V5 very low.

    #148274

    nice work Tanou and to see another approach!

     

    #148278

    I’ve worked on the entry on p5 version, it was copy & past to see what cumulative orders would do in your strategy. What’s strange is that for the short no cumulative orders happen, only for long.  I haven’t put in a limit on them. (ts 0.5% for both long & short)

    edit; to make it work, I removed the breakeven &  atr trailingstop and put in the % trailingstop for cumulative positions.

     

Viewing 15 posts - 1,006 through 1,020 (of 1,239 total)

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