Which PNL curve looks better?

Discussion in 'Strategy Development' started by mizhael, Jun 3, 2010.

  1. traderhf


    Looking at the output, it seems that # of trades in system 1 >> # of trades in system 2.

    With above in mind, I might go for system 1, since statistically its more significant, and performance will be less prone to missing some trades.

    Just looking at 2 params you have given and not knowing anything more about the markets/system behavior etc., I will go with system 1 (as a personal trader and lever it up). From a fund's perspective - for whom DD is of paramount importance, I would go for sys2.
  2. telozo


    system 1 looks like it trades a lot. if you didn't include commission and slippage, you might be in for an unpleasant surprise.
  3. doesn't different parameters and configurations make it a "different" system?

    anyways.. trade them both
  4. Sorry folks. I should have explained a bit clearer.

    Both systems have almost the same entry signals.
    But the exits are different.
    These systems are for FOREX.

    Both have a larger trailing-stop loss with an reentrance period. (So here I use a trailing-stop loss plus a reentrance after certain number of days.)

    Therefore there are two parameters, the {trailing-stop percentage off from the peak} and the {number of days before reentering}.

    The system A has a smoother equity curve because its trailing-stop is wider.

    The system B has a rougher equity curve because its trailing-stop is very tight, it's 1% for FOREX.

    Any more thoughts?
  5. Trailing stops do not determine roughness of equity curves. It is market volatility that does, in conjunction with the trailing stop.

    Besides that, with the same market and volatility, a wider trailing stop should produce a rougher equity curve because it would accept greater drawdown.

    Thus, the whole idea of what you are asking is very confusing, I don't know whether you do this on purpose, but I wonder why nobody noticed.
  6. promagma


    Given that they have the same statistical significance (same entry and same # of trades), it's just a matter of risk control. System B has better return vs. drawdown. The absolute return is a bit less, but that doesn't matter, just use a bit more leverage if you want. Also think in terms of the black swan. System B looks less prone to going really, really wrong.
  7. Okay, with very tight trailing-stop, I stopped a lot, and the equity curve has a lot of flat periods, which is shown in system B's curve.

    With a wide trailing-stop, I don't stop out often, the equity curve is up-and-down with market volatility, but at last the curve is smooth, as shown in system A's curve.

    That's what I meant by "rough" vs. "smooth".
  8. I will be happy to fade your strategies if you do not mind. they look like the sure thing, in opposite direction I mean:)
  9. What do you mean?
    #10     Jun 9, 2010