Absolute need for walk-forward testing? (If given enough data)

Discussion in 'Automated Trading' started by RStrauss, Apr 18, 2024.

  1. RStrauss

    RStrauss

    Nope. My strategy doesn't hold position overnight so I could only get so much out of a day.
     
    #31     Apr 18, 2024
    taowave likes this.
  2. RStrauss

    RStrauss

    As said, I trade HSI futures.
     
    #32     Apr 18, 2024
  3. long

    long

    Do those results come from the same size position the entire timeframe? Or did the system increase bet size as the account grew?
     
    #33     Apr 19, 2024
  4. RStrauss

    RStrauss

    1 contract throughout.
     
    #34     Apr 19, 2024
    SimpleMeLike and long like this.
  5. Sounds like a lot of needless fees.
     
    #35     Apr 19, 2024
    RStrauss likes this.
  6. long

    long

    You should adjust it to trade a percentage of your account balance and see what it does. And report back.
     
    #36     Apr 19, 2024
    Picaso likes this.
  7.  
    #37     Apr 19, 2024
    RStrauss likes this.
  8. RStrauss

    RStrauss

    Just thinking out loud - usually any strategy would go through irregular cycles of mini-series of gains and losses. if the decision to increase position sizing is based on account balance (which increases with series of gains and decreases by series of losses), the timing to adjust the position sizing upward has a high chance to coincide with the end of winning cycle and the beginning of a losing cycle - this would amplify the loss because of higher position sizing. What'd you think?
     
    #38     Apr 19, 2024
  9. taowave

    taowave

    LOL!!!!!!!!!!

    Did you read the replies??





     
    #39     Apr 19, 2024
  10. long

    long

    I’m a discretionary trader and never ran an automated system. But I’ve used autotrading on a SIM account. You are correct, the timing of the position size increase makes a huge difference in volatility and can increase drawdowns. But…every SIM I ran had long term returns that dwarfed constant position size. My theorized fix was to start the system trading as a very small percentage of the account equity. By the time it got to the point to increase position size it would be big enough to not go negative in a drawdown. I’ve never found an algorithm that would continue more than 6 months into the future though. And that 6 months never performed as well as when back tested. Long story short, increasing position size on a profitable system will raise your returns exponentially. That’s if the system performance doesn’t break down. It won’t take long to test it, report back to us.
     
    #40     Apr 19, 2024
    SimpleMeLike likes this.