Use random walk to make $$$

Discussion in 'Strategy Building' started by Thumama, Aug 5, 2008.

  1. My question does not imply that I believe in non-random markets. I was merely trying to question his initial hypothesis.

    To list the ways to profit from randomness, you must first assume the market is random. What if the market is non-random; or alternates between random and non-random periods ? Then all the methods you derive based on the initial assumption are essentially flawed.
     
    #11     Aug 5, 2008
  2. MAESTRO

    MAESTRO

    I am very sorry to say that, please do not be offended, but your statement clearly indicates to me that you are still not quite there yet in understanding of randomness. Randomness has nothing to do with the predictability or with the anticipation. Unfortunately, there is no reading material that I can suggest to help in this matter. One has to do his own research to grasp it.
     
    #12     Aug 5, 2008
  3. Brandonf

    Brandonf Sponsor

    That's what I said. It spends the majority of its time being random, but sometimes its not. But life can be like that as well. My dad was in Vietnam in 1971 and the guy immediatly next to him stepped on a landmine, he lost his entire leg and my dad was ok. Another time his tent was mortared but he had just left to use the bathroom so he was ok. Two years ago my brother was in Iraq and his Humvee was hit by an IED. There where 5 guys in it, 1 of them died, 3 of them are perminatly disabled, my brother suffered a broken wrist and concussion. A lot of things can be random.
     
    #13     Aug 5, 2008
  4. MAESTRO

    MAESTRO

    GOD BLESS YOUR FAMILY!
     
    #14     Aug 5, 2008
  5. eagle

    eagle

    You confuse many people with the term used. You mean a sideways market, and a high tendency to continue in one direction.

     
    #15     Aug 5, 2008
  6. Thumama

    Thumama

    Correlation has nothing to do with randomness. High correlation doesn't imply nonrandomness. Autocorrelation does. The randomness I'm refering to examines the degree of relationship between serial (not parallel) moves.

    Thumama..
     
    #16     Aug 5, 2008
  7. Jerry030

    Jerry030

    You may want to look at Complexity Theory to understand how the market alternates between random and non-random behaviour.... and then plan your trading on that basis. Various measure like the Hurst Coefficient are useful in this process.
     
    #17     Aug 5, 2008
  8. Thumama

    Thumama


    The idea I'm thinking about is not related to Shannon's method. As to Shannon method itself; which is older than me, If it works we would have some Shannon_based billionaires by now.
     
    #18     Aug 5, 2008
  9. Thumama

    Thumama

    Couldn't agree more except for the trick.
    Testing randomness must cover the whole time-series. Slicing the series into predetermined parts will surely produce trends and lead to the rejection of randomness. The questions are where do you make the cut and what's the length of each part. Ironically, the timing and length of those parts are random themselves !

    I can not switch between random and nonrandom mode for any time-series because switching time is random itself.
     
    #19     Aug 5, 2008
  10. Jerry030

    Jerry030

    Lack of knowledge of an event does not prove that it did not happen.

    If anyone made a billion dollars from a defined method or process that could be used by others would you:

    1) Except them to announce or share it?
    2) Announce it on ET of all locations?

    If you or anyone does then I've got some really great luxury condos in South Florida that I can tell you for certain will double in price in the next 60 days if you buy them from me now.
     
    #20     Aug 5, 2008