Avoiding Whipsaws

Discussion in 'Technical Analysis' started by SStrong, Apr 7, 2004.

  1. SStrong


    Hello board, newbie here and somewhat of an amateur. More of a swing trader then day trader, wanted to get the boards opinion on the following.

    I've been backtesting this system and on the small to larger cap stocks ($5 up) this has been working suprisingly well.

    Whats everyones take on this three crossover system? Its simple but it has been predicting a bullish trend reversal...

    Look forward to your replies
  2. Mind as well, post the backtested results.
  3. You know you should render to Caesar what belongs to Caesar why don't you post the site where you find these charts: it's not your own idea as far as you answered in the other post since you're not the webmaster of the site ?

  4. gnome


    You've chosen a period where the markets made a relentless advance and the methodology appears to be "working". Virtually EVERY approach will work in some certain kind of market, but this "multiple crossover" style will be shown in time to (a) NOT avoid whipsaws, and (b) be inefficient.

    Bottom line... avoid multiple crossover method. Fish for something else.
  5. taigong


    Easy--just pick the bottom or top instead.

  6. As gnome stated, adding more moving averages isn't the answer. In fact I would go so far as to say that moving averages are not "the answer" either. They are a tool which from my experience function best as a trend filter rather than as an entry/exit system.

    Think about what a moving average is designed to do. It smooths historical price movement. Think about the ramifications of that for a minute. If an average smooths historical price movement, then will it really perform well in identifying trend changes? Isn't this the trap that 95% of the new traders to this business fall into after looking at a few selected examples in a magazine?

    (to the thread starter) Ask yourself the following questions:

    Is the relationship between the three averages supposed to give an accurate picture of the prevailing trend?

    Do crossovers of these averages accurately represent trend changes?

    If you can't answer "YES" to both questions, then why will using more of the same tool improve your results?

    Have you backtested this method? If not, then why not?