ETF (mostly) trading system development

Discussion in 'ETFs' started by QuantpTrader, Aug 13, 2018.

  1. Well the question is why did you outperform so well initially and then most years underperform? That is the focus to figure out how to correct the system. QQQ had a really bad/shitty 3 years so that hole might be where your comparison makes it look better annualized but year in and year out over past 10 years should be the focus. Also are you outperforming in bearish years because you have short signals and QQQ is just B&H but regularly underperforming bullish years. That might be something to build on....
     
    #21     Aug 14, 2018
  2. traider

    traider

    It looks like an always long or 0 system. No short
    I suspected the same thing as you, system is overtrained on 2000 - 2005. That's why I asked for the sharpe in a few periods for comparison to BuyHold.
     
    #22     Aug 14, 2018
  3. The system is long only. I understand your point but the final objetive is not to outperform qqq with a single system. My goal is to have a portfolio of systems and that portfolio provide consistent returns with the lowest risk possible.

    This is only the first system. That portfolio should have a historical mdd under 20%, no more than 1 consecutive year with negative returns. Also rolling 3 year returns should be above 10%. Also with a portfolio of systems its more easy to use leverage and boost returns.
     
    #23     Aug 14, 2018

  4. As requested.

    As a standalone system without any leverage I found hard to bit the returns of the 2009 bull market. Also that’s not the objective of only one system (maybe a portfolio of systems).

    I explained my main objectives with a portfolio of systems in a post above.

    If we considered adjusted risk returns, system seems pretty nice as a standalone.

    Besides that, I don’t find a major, if any, deterioration, in the main metrics of the systems.

    I currently use a custom developed software that was developed by the trading company where I worked, but I thinking in buying Amibroker, since chart stuff I have to do it in Excel.
     
    #24     Aug 14, 2018
  5. traider

    traider

    You don't gain much from active trading. This is buy and hold 2005 - Aug 2018

    RETURNS
    130.36%
    ALPHA
    0.03
    BETA
    0.43
    SHARPE
    0.75
    DRAWDOWN
    -25.45%
     
    #25     Aug 14, 2018
  6. DD of 25% for B&H since 2005?
     
    #26     Aug 14, 2018
  7. traider

    traider

    you have risk of execution at open. Open is the most volatile time of the day.
    You need to trade in and out , with fees.
    BH gets you the same sharpe ratio. Why bother?
     
    #27     Aug 14, 2018
  8. Can you deal with a 50% DD of B&H, I can't.

    Also, I think you are missing the point or i haven’t clarified enough.

    I’m a firm believer that only one trading system you will not have success. You have to have a portfolio of trading systems, as much uncorrelated as possible.

    My main objective at portfolio level:

    1. Equity curve as smooth as possible.
    2. Maximum of one losing consecutive year.
    3. Consistent return (AR 10/15% range).
    4. MDD under 20% and as close to 10% as possible.
    5. CAR/MDD of 1 or above.

    These are metrics without the use of any leverage.

    With only one system most of these will be impossible.
     
    #28     Aug 14, 2018
  9. traider

    traider

    I understand diversification effect of multiple systems. But right now we are evaluating this single system as opposed to simple Buy and hold.
    I won't replace buy and hold with something that has more complexity + optimization when it doesn't give me a much better sharpe ratio.
     
    #29     Aug 14, 2018
  10. I lived 2 bear markets at this point. I would not give the benefit of the doubt to B&H, I just can imagine my portfolio with a 50% DD… at least I feel comfortable giving the benefit of the doubt to this system.

    Let's add one more system.... it have a lower AR than the first one.
     
    #30     Aug 14, 2018