- Joined
- 14 April 2007
- Posts
- 317
- Reactions
- 0
This is another logical issue, not a problem with the Zig Zag indicator. You can't identify any point as a peak or trough until you've had some degree of retracement after it. At that point you can go back and identify the peak or trough, but of course it's too late to buy or sell at that turning point (naturally performance will be good if you do use the peak and trough bars as signals).
For example, if you say a turning point is defined by a 10% retracement, then you can only generate a buy or sell 10% above a trough or below a peak. As long as you only use the bar where the peak or trough is determined, and not the peak or trough bar itself, then you can safely backtest it (mind you, that's not all that easy).
So you can write alternatives to the Zig Zag indicator to do the same thing, but you can't get around the logical issue that a peak or trough can only be known in hindsight after some amount of retracement.
GP
GP,
As always, I appreciate your time in replying.
I totally agree with all your points. I was led to believe that the Zig Zag indicator behaved inconsistantly in backtesting and one of the reasons given was the problem with defining the last reversal point (using future bars). However, as you say this makes perfect sense as to why.
In my situation I just want the ability to identify how many bars have passed since each of the recent reversal points have occurred, so using the Zig Zag is probably the way to go.
Can I ask why you commented that it isn't easy to backtest the Zig Zag indicator? Any insight into potential problems would be most helpful.
Kind Regards,
Chorlton