Is this a good equity curve?

Discussion in 'Risk Management' started by axeman, Apr 3, 2003.

  1. Here is the equity curve for an ES system I developed
    this week.

    Its the simulated results of trading 3 simple (cough cough)
    systems in parallel in an attempt to make the curve a little smoother.

    Would you trade a system with an equity curve that
    looks like this?

    This is for a 4 year period, without compounding.
    1-3 contracts traded max. 1 for each system.


  2. i don't see any numbers... don' suppose you started in negative territory
  3. In my unexperienced opinion, yes, I would consider trading that. Although it looks like you have a few > 1 month drawdowns. Can you ride those?
  4. What you see is the whole curve.
    It immediately went positive.


  5. Yes... this is for my longer term trading.
    Somewhere to put my money while I trade
    shorter time frames, etc..

    It's 100% mechanical and brain dead trading.



  6. axeman, what is the maxdd for each system and what is it for the combo of all three? Why do they trade diffeent contract amounts?
  7. Aaron


    Nice equity curve. I'd trade it as long as it isn't curve-fit and you've included slippage and commissions in your calcuations.
  8. Ater trying hundreds of coded systems, ive noticed
    that one of the best ways to spot curve fitting is
    by eyeballing the curve and playing with it.

    When you have a good feel for the system, you can
    "tweak" the parameters, and if the curve
    changes wildly with small tweaks, you KNOW you
    have a curve fitting problem.

    However, if you slowly move the parameters around,
    and the curve slowly changes with the tweaks, but
    doesnt do anything crazy, then you have a solid system.

    I can tweak my variables to take only the juiciest of
    trades, and it makes less money.

    Or I can tweak it to play sloppy, and take less than optimal
    trades and the curve starts getting choppier, which
    is what I would expect. ( But still up )

    This is the kind of "curve behavior" I look for to tell me
    that I am not curve fitting.

    Of course, you can keep pushing the parameters to
    the point where it doesnt make sense and it no
    longer makes any money.

    Thoughts anyone?



  9. My thoughts on the question are that backtests can definitely help you keep confidence in the system but you still have to have ways to cut the system off if it diverges too much from expected drawdowns.
    Good luck.
    Remember, the hard part is execution and keeping to the plan.