Fully automated futures trading

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

  1. It's good to be terrified when you make too much money. Are you happy that 25% in a month is in line with historic expectations / your risk target? If not you might need to have a rethink about your leverage - unless you'd be just as comfortable with a 25% loss.

    #731     May 9, 2017
  2. isotope1


    That's reassuring. It's a 25% annual target, and the daily volatility seems to be being realised bang on. I know from looking at the backtested equity curve that you do get some 'steps' up (particularly on smaller account sizes), so I've increased my account size, which should smooth returns.

    Out of curiosity, do you still maintain ISAs? Even with a 28% clip, it's better off in futures. I'm considering drawing them out.
    #732     May 9, 2017
  3. Similar situation here. My account reached its deepest point in drawdown by mid April and has been rising ever since. Compared to that low point is my account now some 40% up and hovering around all-time-highs. This strong rise did worry me slightly, but I made some analysis tool to look at the daily account volatility and compared that to my system settings. This all seems to be fine, so nothing to worry about, is my conclusion.
    #733     May 9, 2017
  4. Yes I put the maximum into ISAs every year - I've got a large long only equity / ETF portfolio and it makes sense to gradually move this into ISAs (bed and ISA trade). Also I don't compound my futures returns, so any excess profits I make are first used to fund my ISAs.

    Although the SR would be higher if I put absolutely everything I owned into futures, I have two misgivings;

    a) the size of the daily swings in my p&l would be gigantic in GBP terms (my long only portfolio has huge £££ swings, but I don't look at these daily, only once a year) unless I reduced my risk target down to compensate, affecting my expected returns accordingly.
    b) the stream of dividends is a psychologically more secure way of earning an income than futures MTM

    #734     May 9, 2017
  5. southall


    It would be helpful, for those of us who are new to your journal, to know the exact start date you are using.
    My guess its the start of this journal (Feb 11 2015), but not sure. Thanks.
    #735     May 9, 2017
  6. No, I started trading on April 5th 2014 and the total p&l is since that date.

    #736     May 9, 2017
  7. They are the type of returns I'd expect from someone that knows what they're doing. Well done.
    #737     May 9, 2017
  8. Fundseeder have now got working trade statistics on their site. Here are mine (futures only):

    profit factor 1.31
    average win $1,274
    average loss $1,014
    all trade average profit: $156

    These numbers are quite small because I usually trade only one contract at a time, continuously adjusting my position. However the win:loss ratio is what you'd expect from trend following.

    Percent winning trades: 51.1%

    Again you wouldn't expect this to be high; if anything slightly below 50% is normal.

    Average holding period:
    winning trades: 38 days
    losing trades: 22 days
    all trades: 30 days

    Again for trend following you'd expect winners to be held longer than losers. The fact I'm trading quite slowly is also evident here.

    #738     May 9, 2017
    f2calv likes this.
  9. isotope1


    What do you define as a trade? Is that a round trip (zero-position to X-position - to zero-position)?

    On a completely separate note, for the EWMAC fast numbers 4, 8, 16, 32, 64 (and slow = 4*fast); is there a mathematical basis to this particular progression?
    #739     May 9, 2017
  10. It's fundseeders definition - I think if I buy 1 lot on five successive days, and then sell 1 lot on the next five days, that is ten trades - even though arguably I've only held one 'position'. However I will have held each 'trade' for an average of five days - and that figure is correct.

    I think it would possible to derive those EWMAC numbers using stochastic calculus - but I'm not sufficiently interested to try and attempt it. That would assume a random walk with no trends. I get these kinds of numbers with random data (both with and without underlying trends) and also with real data.

    #740     May 9, 2017