Why do I see "Trends" in Randomly Generated Data?

Discussion in 'Data Sets and Feeds' started by Rahula, Feb 21, 2008.


  1. Hi RD,

    Do you have any further information on this lawsuit?

    Regards,

    surf
     
    #221     Feb 26, 2008
  2. http://infoproc.blogspot.com/2007/07/algorithm-wars.html



    Surf, I just don't see why posters around here rip on you so much. Your "cynical" beliefs are not much far from my own (if I'm interpreting correctly). A professor once said, "A cynic is someone who sees the truth for what it is." Words that stuck over the years.

    People need to start reading more books like richard ney's "wall street jungle" and "trading with the enemy: cramer." Once you understand and agree that the playing field is not level, then you begin to model around that hypothesis.
    --------------------------------------------

    http://www.arezzotrade.com/wall_street.php

    "When you start to think you've got the special sauce, someone else (probably)
    has it too,"


    "David Viniar, chief financial officer of Goldman Sachs, told clients in a conference call Monday detailing Goldman's own quant funds' steep losses that the unusual dislocation in stock prices last week was a 22 standard deviation move. In other words, it should not have happened in the entirety of human history if returns are what statisticians refer to as log normally distributed. Similarly, the 1987 stock market crash should not have happened either. Clearly unusual moves can and do happen, meaning that basing models on even a 150-year history of stock prices can miss disastrous outliers. Simons acknowledged as much in an Aug. 9 letter to investors detailing the 8.7% loss over the first six trading days of August.
    "We have been caught in what appears to be a large wave of de-leveraging on the part of quantitative long/short hedge funds," he wrote.
    Renaissance ignored several requests for further comment about its performance and operations."

    The sneaky spurious fat tails keep jolting the Gaussian modelers.
     
    #222     Feb 26, 2008
    beginner66 likes this.
  3. RedDuke

    RedDuke

    #223     Feb 26, 2008
  4. #224     Feb 26, 2008
  5. Another professor once said, "A cynic is someone who sees "a" truth for what how it fits into his corner of his universe".

    People need to start thinking more by personally validating what they read.

    The playing field has never been level from the day each of us entered this world and will only become level when we die.
     
    #225     Feb 26, 2008
  6. Interesting commentary. So, could you tell us in broad strokes (nothing proprietary, of course) how you trade the markets?
     
    #226     Feb 27, 2008
  7. Maestro,

    I need to disagree with your well accepted idea that it is the masses that move the market, therefore studying their pyschology is beneficial. At one point in history this was likely correct, however now, due to monster capital pools controlled by the very few, crowd psychology is no longer relevant in the financial markets--the crowd is trend followers by default, always too late. I believe the premise of behavioral finance as it relates to market prediction is fundamentally flawed, much like TA itself, due to the above.

    regards,
    surf
     
    #227     Feb 27, 2008
  8. MAESTRO

    MAESTRO

    If you are asking me, then the answer is very simple. I use the game described in my paper to construct multiple strategies utilizing differences between distributions exhibited in the markets vs. Gaussian distribution. It only works well (no losing days at all) if you are involved in hundreds non-correlated games at the same time. I think it is as far as I can go with it. The clues are all in the paper.
    Cheers,
    :cool:
     
    #228     Feb 27, 2008
  9. MAESTRO

    MAESTRO

    In the traditional sense, you are correct. Traditional behavioral finance like TA is wrong. However, I use totally different approach that, in my opinion, is free of the mentioned by you short comings.
    Cheers
     
    #229     Feb 27, 2008
  10. MAESTRO

    MAESTRO

    Later in the day I will post a very important paper that will shed some light on the subject. I just need to find it.
     
    #230     Feb 27, 2008