ETF (mostly) trading system development

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

  1. I´m not expert at write code but i have some basic knowlodge of amibroker and Mestatock. I think i can translate the rules to both languages. Dont ask me about tradestation code.... i'm not that good.
     
    #11     Aug 13, 2018
  2. traider

    traider

    Can you run your system on a random selection of other ETFs? Choose a period that is as long as possible.
    Or if you can post your code, I will try to replicate it.
     
    Last edited: Aug 13, 2018
    #12     Aug 13, 2018
  3. I usually do that test in related markets like (SPY, IWM or MDY) and see if the system makes profit and if the equity line have a positive slope. I dont expect to achieve the same results....

    Also, i dont test the system in other non related markets because i dont follow the idea that one size fits all.
     
    #13     Aug 13, 2018
  4. Basically the system have two parameters that can be optimized (i never change, with a small exception the ATR_STOP formula (in theory you could, but i choose not to do it), so i don’t considered it a parameter that can be optimized).

    The parameters are:

    • The IBS threshold
    • The number of days in the Highest High Close
     
    #14     Aug 13, 2018
  5. As you can see the system make money with all the range of IBS threshold value. I found a stable range between 20-35.

    I choose the original value that I tested (20).
     
    #15     Aug 13, 2018
  6. Regarding the number of days in the Highest High Close i found a stable range between 3 and 6 days.
     
    #16     Aug 13, 2018
  7. In order to evaluate real time performance i will provide real-time signals on my twitter account under the hastag #QQQIBSM1.

    From time to time i will also provide here the updated results of the system.

    Any comments or critics? Or can we go to the next one?
     
    #17     Aug 14, 2018
  8. traider

    traider

    Can you compare the system's metrics to buy and hold for the various time periods. (esp sharpe)

    2000-present
    2005-present
    2010-present

    Are you using your own custom developed backtester?
     
    #18     Aug 14, 2018
  9. This is what you are competing against, QQQ buy and hold annual returns:

    upload_2018-8-14_9-57-34.png

    You are way under almost every year on your system so it is a lot of work and commission to underperform simple buy and hold. You started out beating it first 3 years then consistently underperformed so you need to rethink the strategy.
     
    #19     Aug 14, 2018
  10. Thanks for feedback. Will elaborate more on this and the previous post later in the day, but the way you see it is one way, the way i seet it is that buy and hold had a negative performance of minus 15%.
     
    #20     Aug 14, 2018