How to Auto Trade Support and Resistance?

Discussion in 'Automated Trading' started by greaterreturn, Jul 6, 2008.

  1. Please only reply to this thread if you personally trade automatically around support and resistance.

    A test system to trade support and resistance works but has some problems.

    And it works great to buy/sell break outs from the S/R lines either above or below.

    Here's the problem. Close to half the trades are whip sawed around the S/R lines since there's congestion around them.

    Specifically, the strategy right now sets a range around S/R lines -- 8 pips. (Forex) So it buys any close of a 1 minute bar that crosses over 4 pips above the S/R line and sells any close that crosses below the S/R line.

    That way, profits come both from the bounces off the line AND when the market blows through the line.

    HOWEVER, when it does blow through, it gets hit with a 8 pip loss from the whipsaw.

    Worse yet, it whip saws several times sometimes around the S/R line that create 8 pip losses several times ina row before prices head up or down from there.

    So how to remove or reduce those whip saws so that the strategy generates much more profit?

    Any help?

    Wayne
     
  2. wenzi

    wenzi

    Take a look at the book "Mechanical Trading Systems"

     
  3. I checked out that book on Amazon and Google. With some pages removed, it still appears that he explains the psychology of support and resistance (which I understand fully) without any specifics about how to avoid getting whipsawed around them.

    Still, his reminders about the psychology have stimulated my thinking and maybe some ideas to try.

    Specifically, he reminds that some measure of risk/reward must be found--even in chop markets--to justify each trade.

    Sincerely,
    Wayne
     
  4. Another thing that book refreshed is the reversion to mean tendency.

    Of course, I already filtered out the trending periods. I'm working on choppy periods.

    So during these choppy periods, I should only play the "bounces" of S/R back towards an average.

    Right now, the strategy tries both bounces and breakthroughs of S/R. That's silly.

    Playing only bounces also improves the ability to compare risk to reward. The distance back to the average (reward) can be compared to the stop loss (risk).

    In contrast, with breakthroughs it can be near impossible to estimate a reward. But I will research that anyway. Maybe the distance to the next S/R line could statistically work as a potential reward. Ideas?

    Someone jump in if this is going wrong.

    Thanks!
    Wayne
     
  5. jellob

    jellob

    I am researching on some thing very similar. Rather than S/R, I use MAs. Nonetheless, I think that sooner or later you will run into the same problem that I have. Since the success of any system depends on the underlying security, it is paramount that the stock or future oscillates predictably between trending and consolidating. If you like, PM me and we can exchange notes.
     
  6. wenzi

    wenzi

    The book had a small section on whipsaw filters in trading systems.
     
  7. I greatly appreciate the tips and ideas on this thread.

    After an epiphany Sunday, my chop strategy now works beautifully.

    I confess that S/R's are a part of the solution. But the key, just like anything, is how to use S/R's correctly.

    Sincerely,
    Wayne
     
  8. Get the book "The Encyclopedia of Trading Strategies" by Katz and McCormick
     
  9. What pair are you talking about? If its EUR/USD or GBP/USD then 8 pips is just noise.
     
  10. The key to using S/R (or any kind of price resistance) is to determine the strength of the resistance.

    Re-occuring swing top/bottom prices (double/triple bottoms or top) is generally a very weak price pattern (actually all chart patterns are weak signals to use, as itself).

    The key is to use volume and more-so bid/ask data.
     
    #10     Aug 12, 2008