Mathematically Predicting the Future?

Discussion in 'Automated Trading' started by MarkBrown, Oct 27, 2007.

  1. yes
    convient theories

    is what they are

    Random walk....CAPM...etc
    non work
    I agree...
    it's how it's used or sometimes overlooking
    to creat a better model
    Modelers sometimes make it complicated
    Good modelers do not want to...gets harder that way...
    but I think its in the sheer observation
    U think if someone had great model...they will post it?...Noway..(ie. Gann)
    Trust me my friend you are on the right track to use math
    Check what they say about Oil...these days

    getting the emotional rise world-wide

    recall the Nasda 100 when was 5200 area
    Same thing

    DJI Index heading to 11,300
    in july
    politicans and cnbc people wont do their Jobs if they knew were the MKT is going

    they Just Tag the excuses for the MKT doing its thing

    So short the ^DJI till july
    then Wait for cnbc to excuse it in july
     
    #91     Apr 24, 2008
  2. Hey Guys
    Doctors recommended Ciggretes in the 1930 as a digesting thing + relaxing method

    1960 they recommended qoilodes
    over the counter drug to relax
    elvies dies

    at least math is the one field that uses profs
    and not empirical

    80 processers in a normal house...computer clocks...etc
    maybe Albert had something to with it?
    I do not think the FDA likes it
     
    #92     Apr 24, 2008
  3. Any update after 13 years?
     
    #93     Mar 27, 2020
  4. volpri

    volpri

    NEXT BAR …Reason for scalping LOL. SEE most of the time the market is probing back and forth. Price will go where more transactions take place.
     
    #94     Sep 6, 2021
  5. IMO...
    Fundamentally the markets are driven by psychology, external events and business success/failure. However, there are a lot of people who want to treat them as a random process.
    I would suggest that this is dangerous.

    Basic statistical calculations are unlikely to tell you that a 1 billion dollar hole is going to appear in a balance sheet out of the blue.

    I think that mathematical modeling is potentially useful for spotting odd characteristics worthy of further attention. It might also be useful for spotting the effects of other market participants using similar modeling.

    Looking backwards, you can generate a model the fits basically anything, but looking forward is a different story.
    If someone believes they have a model that predicts a small segment of the market for a short period of time I would consider that a success, but not count on it lasting long.

    How many people are running around saying, "My financial model predicted covid and its effect on the market” ?

    On a side note:. Combinations of probability distributions tend towards gaussian. (Central limit theorem). When someone is talking about modeling a big complex system, but somehow not have it be gaussian, that should suggest some extra attention.
     
    #95     Sep 8, 2021
    comagnum likes this.
  6. Overnight

    Overnight

    This is just a note to myself.

    @Overnight 12 year bump.

    Carry on.
     
    #96     Sep 8, 2021
    comagnum likes this.