Curve-fitting to success?

Discussion in 'Automated Trading' started by RStrauss, Sep 30, 2024.

  1. Hello RStrauss,

    Yes, I mean it.

    I do not have an answer for you because I have no proof that curve fitting trading systems made me money trading.

    So I can not offer you great advice.

    All I can say is keep on trying your best.

    I tried developing algos for about 3-4 years buddy, I could not find anything worthwhile to invest money that I trust to make me money.

    I was not curve fitting. I only curve fit 50% of the data set.
     
    Last edited: Sep 30, 2024
    #21     Sep 30, 2024
    RStrauss likes this.
  2. RStrauss

    RStrauss

    #22     Sep 30, 2024
  3. RStrauss

    RStrauss

    Thanks! Agree.
     
    #23     Sep 30, 2024
  4. RStrauss

    RStrauss


    So you are a discretionary trader now? There has to be some sort of trading rules that you are following to place your bet. Can you not code those rules to automate your trades?
     
    #24     Sep 30, 2024
    SimpleMeLike likes this.
  5. Peter8519

    Peter8519

    According to most back testing literature, curve fitting(tweaking parameters to give the best results) is a pitfall. Just go with a simple and robust strategy and run it with good and bad year data. In good year, rack in profits but in bad, it gives the least loss.
    upload_2024-10-1_8-58-35.png
     
    #25     Sep 30, 2024
    RStrauss likes this.
  6. RStrauss

    RStrauss

    Thank you Peter for sharing! want to tap your mind further on what's considered a simple strategy.

    For instance, say if I use a simple MA cross-over strategy with 3 filters (any of the 3 conditions, if met, would not result in a trade). Is that considered simple enough? What's the maximum number of variables (used for optimization) should there be in a strategy that's considered simple?

    Thanks in advance!
     
    #26     Sep 30, 2024
  7. Peter8519

    Peter8519

    Don't use moving average as it is lagging. Just scan the price within a certain period and see the trend if it's up or down.
     
    #27     Sep 30, 2024
    oshjdf and RStrauss like this.
  8. Good Evening RStrauss,

    Yes, I am discretionary trader now.

    I need food in my belly @RStrauss. I need food in my belly @RStrauss. Algos do not guarantee me food in my belly every night.

    I work +50 hours on my regular job and have to take care of my 2 dogs after work. I do not have the time, energy or ambition to be programming trading algos anymore. It is too much work and the reward is sooo little.

    My only rule is make money that I need to eat food everyday and get rich.

    No, I can not code the rules in an automated trading systems format, because I do not know what to program the entry signal to because I am guessing and doing my best.

    Doing my best in my trading is all I need.

    I never coded and back test a strategy that I trust to put food on the table at night and make me rich.

    They may have algos that make traders money out there, but I never see any and I am tired of looking.
     
    Last edited: Sep 30, 2024
    #28     Sep 30, 2024
    RStrauss likes this.
  9. themickey

    themickey

    There's a contradiction.
     
    #29     Sep 30, 2024
    SimpleMeLike likes this.
  10. themickey

    themickey

    100% discretionary trading is too much work and the reward is sooo little.
     
    #30     Sep 30, 2024
    SimpleMeLike likes this.