Have anyone used statistical based Oscillator?

Discussion in 'Trading Software' started by manz66, Mar 18, 2004.

  1. manz66

    manz66

    One company claiming there statistical based Oscillator generates 'cleaner crossover signals, and more reliable divergences'. It is similar to macd, but statistacal based.

    The other one is quote:

    "The KEES PeakOscillator is a simple oscillator from which classic divergence signals are generated, but with two major differences. Unlike traditional oscillators which are derived from moving average with set lengths, the PeakOscillator is derived from a mathematically sound, statistical evaluation of trend which looks at over 50 different trend lengths, rather than just two".

    Is it true?
     
  2. Ditch

    Ditch

    I've used them. I didn't notice any difference whatsoever with the divergence signals from the ergodic indicator ie. By the way, e-signal has a 1 month free trial, so you can check it out yourself.
     
  3. Nice post.

    Say you were doing the nononsense thing, i.e., creatiing your own method of trading because all those sold by others are no good for nononsense.

    Apparently KEES assembled 50 different length MA's and then statistically assigned then to price patterns where they worked best. It looks like Kees plugs in the appropriate bunch for the contemporary situation just instead of using the same ones all the time.

    Kees is apparently looking for divergence as the way to make money. MACD is a combo of some MA's too. MA's of MA's it turns out. Divergence and MA's, etc are thought to be helpful in ET when you look at all the mentioned uses.

    A peak oscillator is probably a trough oscillator too. So the focus could be taking profits primarily at peaks/troughs.

    Because everyone knows that peaks/troughs may be followed by lateral trends or opposite trends, the "peak oscillator is probably not an entry tool but more and "end of something" tool.

    My bet is the sales literature for both talk about using lagging indicators to make money when divergence happens.

    Ask what is diverging from what? If an equation puts together data from the past (like say a trend continuing to unfold), then when divergence occurs between the indicator and the price or something, it most likely tells you something ended.

    OT

    To make money it is intersting to consider the present tense question: What do we need to know right now?

    Personally I have come to the conclusion that I should be the market as a fundamental consideration for making money.

    I feel exiting is only a concern when I am not any longer making money.

    So all I do is consider What do I need to know right now in the present. None of the aspects you are considering come to mind for me.

    Alternatively, I do know to always be looking for the correct thing.
    I am in the market; it is making money. I see profits accumulating. I glance around and monitor. I keep on the right side of the market.
     
  4. nitro

    nitro

    What the heck?

    You sound like a different person!!!!

    I can understand what you are saying!!!!!

    :eek: nitro :eek:
     
  5. In this sentence from the post above a word is missing:

    Personally I have come to the conclusion that I should be the market as a fundamental consideration for making money.


    After "be" put "in" so the sentence reads:

    Personally, I have come to the conclusion that I should be in the market as a fundamental consideration for making money.

    Thank you for your patience.
     
  6. As usual you didn't. I left a key word out od a key sentence.

    When you understand what I am saying, immediately give consideration to a mistake that I have made. Then conclude that you cannot understand what I am saying.

    This is a two way street:

    1. if you understand it, then I made a mistake.

    2. If you can't understand it, then it will work for others.

    Naturally, there is a dilemma here.....for you. As everyone watches, they get to observe two people. One of them is said to be posting crap and is a snake like thing that so far cannot be characterized neatly. We are still finding out who the other person is.
     
  7. manz66

    manz66

     
  8. I suppose Bollinger Bands should be considered a statistical-oscillator.

    riskarb
     
  9. nitro

    nitro

    I understood _two_ of your posts!!!!!!!

    YOU'RE ON A ROLL!!!!

    nitro :eek:
     
  10. manz66

    manz66

    I was reading SFO magazine published in Oct, 03. The author Cynthia Kase in the article claims that, using daily data futures contracts of multiple commodities including s&p 500, altogether of 47,000 days of data, 185 years in total; her systems catch three quarter of market turns compare to less than half with the traditional indicators.

    Also, her idea 'that further the market moves in relationship to volatility, the less likely the movement is to be classified as random, and more likely it is trending'.

    Another thing, she wrote about the warning line using two bar reversal and the three additional stops depended on two and three standard deviation reversals.

    To me that is interesting way to trade.
     
    #10     Mar 18, 2004