Fully automated futures trading

Discussion in 'Journals' started by globalarbtrader, Feb 11, 2015.

  1. No I am saying what I do in Systematic trading is correct.

    Will explain in more detail on Monday

    GAT
     
    #2301     Oct 10, 2020
  2. sef88

    sef88

    Could I check if anyone has experience developing Rob's system in Saxo using its API? I was looking at Saxo and it seems to be the only broker where I live (Singapore) that offers a wide range of dated CFDs (non dated are crazy expensive). Checked it out as per some of Andreas Clenow's posts. Unfortunately, unlike IB (ib_insync, ibridgepy, tws api), I don't see people developing frameworks for Saxo. Would be great if someone has experience with Saxo and could shed some light on this.
     
    #2302     Oct 10, 2020
  3. Do you think the word “correct” applies to anything in the systematic investment space? I always feel that I am picking the least wrong option, whenever you’re making any choices. Hell, it’s even hard to say that something was a correct choice a posteriori IMHO.
     
    #2303     Oct 10, 2020
    globalarbtrader and sef88 like this.
  4. Good point. I meant the evidence in the blog post is consistent with the system presented in the book.

    GAT
     
    #2304     Oct 10, 2020
  5. In that case I must have misunderstood the first portion of your conclusion:
    Systematic Trading sets the risk target at a fixed value, and explains why half-Kelly (i.e. 25% annually) is a good choice.
     
    #2305     Oct 10, 2020
  6. tgibson11

    tgibson11

    The key part of that quote is "every single day".

    The system described in Systematic Trading has an average annualized volatility target that is fixed, but it does not attempt to hit that target every trading day, or every year for that matter, only over the long-run. Expected vol will be higher for some time periods and lower for others, depending on the combined forecasts for each instrument.

    Rob's recent blog post was investigating the effect of trying to hit that vol target every day.

    Apologies to Rob if I've misstated anything.
     
    #2306     Oct 10, 2020
    globalarbtrader likes this.
  7. That's pretty much it, with the added factor that differences in position and correlations will also cause expected risk to deviate from the long run target (what I call RCF in the post).

    Thanks I won't have to come back to this on Monday after all...

    GAT
     
    #2307     Oct 10, 2020
  8. Thanks for that clarification.
    What remains is that I still don't understand from the blog post how the system can be improved.
     
    #2308     Oct 11, 2020
  9. The blog post said this specific change doesn't lead to any improvement.

    GAT
     
    #2309     Oct 11, 2020
  10. Oh, I see. I must have had too high hopes while reading it.
     
    #2310     Oct 11, 2020