Fully automated futures trading

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

  1. Oh okay that's interesting.

    GAT
     
    #2831     Jul 9, 2021
  2. #2832     Jul 9, 2021
    djames, Kernfusion and wopr like this.
  3. blink18

    blink18

    Does IB Gateway need daily re-start/re-login or weekly?
     
    #2833     Jul 9, 2021
  4. d08

    d08

    Weekly but IBCAlpha avoids all that given you have opted out of 2FA.
    Personally, I have automated restart via IBC daily (cron), to avoid memory leaks and other issues.
     
    #2834     Jul 10, 2021
    blink18 likes this.
  5. blink18

    blink18

    Do you suggest any security measures when 2FA optout? IB sugests static IP with IP filtering + secondary username + SSL encryption. Anything else?
     
    #2835     Jul 10, 2021
  6. ValeryN

    ValeryN

    One cool thing about their MFA optout is that you can't withdraw money using it. So an intruder is kinda stuck trading your account.

    Re IB gateway - another option to automate is to put it in a docker container, configure health rule to a presence of a Java process inside and set auto restart if it dissapears. That's what I do.
     
    #2836     Jul 10, 2021
    blink18 likes this.
  7. Magic

    Magic

    @globalarbtrader

    Rob, are you able to call up your realized sharpe ratio over the lifetime of running your system for each individual instrument?

    I was wondering what the realized sharpe is for VIX particularly, since there is risk premium embedded in the term structure.

    In the same vein I would expect bond and equity futures to have higher avg sharpe than commodity futures because some of the natural yield produced should be captured by your carry rules.

    Just curious if this turned out to be the case over your trading timeframe or if performance of risk instruments are not very distinguishable from the commodities when trading with a momentum/carry framework.

    Would appreciate input from some of you other systemic guys as well if you have data on this particularly through the last 5-10 years.
     
    #2837     Jul 11, 2021
  8. Not easily and I'm not sure how valuable an exercise it would be. For starters, even with 40 years of data from a backtest, I don't see any statistically significant differences in performance between different instruments; so using a mere 7 years would be a total waste of time. Secondly, the returns in eg bond futures especially, would be heavily influenced by the long term secular upward trend in these instruments which are unlikely to be repeatable.

    GAT
     
    #2838     Jul 12, 2021
    Magic likes this.
  9. Magic

    Magic

    Okay, thanks. That pretty much answers my question. Interesting that the backtest didn’t produce any noticeable difference between instruments. Since long bonds/equity (similarly short vol) has a moderately positive sharpe ratio vs. SR a lot closer to zero for constant one-way exposure to commodities I would have expected rules controlling portfolio volatility and reducing exposure to weakening momentum to enhance that differential further. Or even if TF rules added the same benefit to each asset class, the traditional assets should still be noticeably ahead.

    I didn’t expect there to be no difference over a relatively long time period. Perhaps a lot of the avg returns from buy&hold risk asset portfolios come from mean-reversion or un-expected upside shocks that can’t reliably be forecasted in a systemic manner?
     
    #2839     Jul 12, 2021
  10. I didn't say there was no noticeable difference, I said there was no statistically significant difference. Not the same thing at all; you need a massive Sharpe Ratio difference for statistical significance.

    GAT
     
    #2840     Jul 12, 2021
    sef88 and Magic like this.