Kaufmann Risk of Ruin

Discussion in 'Risk Management' started by virtualmoney, Nov 8, 2011.

  1. One thing missing from this conversation is that zero probability events can happen all the time and there can be infinitely many ways for a probability 1 event to NOT happen.
     
    #101     Nov 28, 2011
  2. Zero probability means that the event is not part of the probability space and probability 1 means the event is certain, i.e. it is like it has happened already.

    So maybe you want to say something else like for instance that people do not always know the probability space and they cannot measure probability correctly. This is true for trading and I think this is the reason that the probability of ruin is 1 no matter what the hit rate is and the risk-reward. The issue is not if it happens but when it will happen.
     
    #102     Nov 28, 2011
  3. Not always. For example, there are infinitely many numbers between 0 and 1 and the probability of randomly selecting any of them is zero. Yet every time one is selected a zero probability event occurs.

    Less intuitive is that there are infinitely many rational numbers between 0 and 1 and infinitely many irrational numbers between 0 and 1. BUT when drawing a random number from the entire sample space [0,1], the probability of randomly selecting any of rational numbers is zero and the probability of randomly selecting an irrational number is 1, even though there are infinitely many rational numbers that could potentially be selected.
    No, the reason the probability of ruin is 1 is because with any finite account size X, there's always a finite distance from X to zero. But even though the probability is 1, ruin will almost surely happen -- not surely.
     
    #103     Nov 28, 2011
  4. kut2k2

    kut2k2

    Untrue. The probability of randomly selecting a particular (i.e., predetermined) number is zero. But the probability of selecting any number between 0 and 1 is one by virtue of the fact that the act of selecting a number between 0 and 1 is performed. Randomly selecting an unpredetermined number is not a zero-probability event.
    Theoretically true, practically untrue. In the real world, the number is chosen mechanically or electronically and roundoff or measurement error assures the result will be a rational number.
    Certainty of ruin is averted by using a positive-expectation trading system.
     
    #104     Nov 28, 2011
  5. Wrong. All sample points in [0,1] are zero probability events; therefore any outcome can only be a zero probability event.
    I was talking theory but even "in the real world" irrational numbers can easily be expressed electronically, for example with roots.
    Your certainty of ruin can't be averted because you're a dumbass.
     
    #105     Nov 29, 2011
  6. No, you just failed the course on probability. probability can be defined in 4 different ways. You are talking about the Classical definition which is known to be problematic because:

    1. I requires all oucomes are equally likely
    2. It does not apply to infinite outcomes unless a measure of infinity is introduced for the particular problem

    No, even though you are correct about the distance, the probability of ruin is 1 because the other players in the game make it to be that.
     
    #106     Nov 29, 2011
  7. No, YOU just failed it.
    http://www.statlect.com/subon/probab2.htm

    Next time check your facts before trying to counter mine with nonsense.
     
    #107     Nov 29, 2011
  8. You don't understand what you read. You need to take a graduate course in probability to talk about these things. You also must find reputable sources like a standard text.

    The subject of the classical definition is controversial. of course, you have no clue about that.
     
    #108     Nov 29, 2011
  9. No, Imbecile Bill, YOU don't understand it because it very clearly supports everything I wrote.

    P.S. The article at the link was written by a Ph.D. in mathematics so you're wrong again, dumbass. :p
     
    #109     Nov 29, 2011
  10. Sounds like Quantum Physics stuff...Is prob(rich)=1-prob(ruin)?
    Or perhaps you can be both ruin & super rich at the same time...Parallel universe perhaps?
    http://en.wikipedia.org/wiki/Schrödinger's_cat
    In other words, is prob(zero prob event happening)=1-prob(prob 1 event Not happen) or are they totally independent?:confused:
     
    #110     Nov 29, 2011