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

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



  1. this is sort of twisting reality.

    the definition of market order is ALWAYS based on the individuals viewpoint. " i can see the order" " i can profit from seeing the order" and obviously " the order, that i and only i see, allows me to profit"

    the randomness crowd looks at the market from a detached viewpoint generally-- the market as a whole, not individually based.

    it's the SUBJECTIVE v. OBJECTIVE dilemma----- the subjective seeing of order is meaningless to the objective observer.

    and NO your profit from thinking you see order thereby predicting the market--- has no bearing on the subject. you're the outlier to the whole.

    surf
     
    #291     Mar 9, 2008
  2. Nice idea but that would require original thought from both sides and that is something impossible to accomplish thus far in these hollowed pages.
     
    #292     Mar 9, 2008
  3. MAESTRO

    MAESTRO

    Humans typically explore the universe from their ego stand point of view. It is virtually impossible for an average person to imagine that the collective behavior of many target oriented, highly knowledgeable and intelligent people is as random as the Brownian motion of gas molecules! If a 1000 people would be asked to trough darts into a dart board the overall density pattern of darts around the central circle would be very close to the Gaussian distribution curve. Despite the fact that each and every person would do his/her best to aim the darts they will still lay in this magnificent pattern! And nothing could be done about it! The link below is demonstrating this process using a very interesting applet.

    http://demonstrations.wolfram.com:80/StatisticalMechanicsOfMoney/
     
    #293     Mar 9, 2008
  4. Jerry030

    Jerry030

    Yes, true but the point of trading the markets is to make money at a rate above the average common available return (or random if one wants to call it that.)

    That goal has nothing to do with an appearance that a person calls by any particular name or what they say their intention was. If a given person can hit the smallest circle in the bulls eye 9 times out of 10 over 1000 attempts in contrast to the average of all other participants who do so .01 times out of 10 over 1000 tries, we can say their skill in hitting the bulls eye is very exceptional.

    You can call it luck, they can say it is skill, another person can say its divine intervention or that they can do so because they are actually a visitor from another planet disgusted as a human.....many theories can be offered...BUT the fact remains that their accuracy is exceptional.

    A debate on semantics may have some academic or entertainment value but from my perspective the debate isn't likely to produce much profit. I trade for profit. The questions of importance from that perspective are:

    1) Can exceptional profit be extracted from the markets?

    2) How can that be done?

    I'll take a stab and defining exceptional profit:

    Profit Factor: > 3.5
    Number of Trades: > 200
    Ratio of Bars in Trades to Total Bars in Time period: < 10.0

    What do you think?

    Jerry030


    FOr example the retrun on my IRA is better than the prime rate but nothing that one could call an exceptional rate of return on my money.
     
    #294     Mar 9, 2008
  5. #295     Mar 9, 2008

  6. what you are describing is the same paradox in mathematics that occurs when one attempts to computerize and automate what used to be paper calculations and HP Scientific Calculator equations done in long form, or simply written out.

    what you are also describing is the same thing that programmers tasked with in generating testing data to apply towards their programs, and using random generators have observed,,

    in both cases, the conclusion is very similar to what you stated in your premise, that these random number generators and random data generators are not that random in their originality or creation of said data sets...

    now, what lacks credulity, it the notion that one can then apply this phenominon to real life trading and real life data collction...
     
    #296     Mar 9, 2008
  7. Jerry030

    Jerry030


    In terms of your question "commonly available return = random "

    I was trying to get out of the academic and theoretical into the practical. Labels and what they mean to each person is often a hindrance to practical communication and real world results.

    So for purposes of my question lets assume that there is a common rate of return on investment activity. If it is impossible to trade the market with any predictive ability them by definition the results are random if the market is random. Regardless of the random or non-random nature of the markets, if we take several million individual investors, a few thousand mutual and hedge funds and average their returns we have the average or commonly available return. I suspect this is somewhere between 0 and 10%, but that is just a guess.

    Would you agree?

    Would you agree that a group or method that can get a 50% or 100% return is not common or average?

    This is independent of the definition of the market as random or non-random...just what is the average and non-average return on account.

    Thanks for the reference to the book but the lack of mathematical sophistication of the American populace has little to do with making money in the market. It all just labels and terms being tossed about with no intrinsic meaning, at least in terms of profitable trading.

    In the Dark Ages Catholic theologians used to debate how many angels could dance on the head of a pin. Any angels observing the discussion probably laughed themselves silly since there probably was more value in seeing an angel or talking to one than debating their size. IMO, the value is making money not debating the best label to attach to the market from theoretical terms.

    So the question is better framed through the following logic:

    a) What level of profitability would indicate non-random investment performance?

    b) Does that level exist?

    c) If it does then the market is non-random.

    d) if is does not then until it exists the question can’t be answered with certainty

    200 years ago I'm sure physicists debated if it was theoretically possible to fly like a bird. They likely gave up the debate when some bicycle makers form Ohio invented the airplane in 1903.

    Jerry030
     
    #297     Mar 10, 2008
  8. Jerry030

    Jerry030

    Well yes, if it's human trading or looking at the market. If you have a non-human, say a neural network doing the looking the question is can it learn something that it can use to predict future market activity?

    The nice thing about the more sophisticated of these is that they turn what they learn into predictions which can be tested against the future and a conclusion made based on their ability to generate profit. This eliminates the subjective entirely since the NN has no emotions about its performance or what is sees or learns at all.

    Jerry030
     
    #298     Mar 10, 2008
  9. ammo

    ammo

    i see this argument over and over thru all posts in several differ,ent journals,there never is an answer, its like philosophy, if we base our argument on an if rather than an is, we never are sure of our results,i must complemrnt this thread as even tho no ones sure, no one has pre tended to be sure with an insult to make thier unproven point, carry on in your quest to quantitively figure out the mrkt,if you get an answer,we'll move on to more complicated subjects, women,logic,god ,longevity,health, all the answers are here, it's just a question if they are right or just raise another question, if its gonna rain and you are working outside , do you get a half hour notice? usually u do....the mrkt is the same way ..if you just stand back and look you usually know something is not rosy, trust your feeelings here.. go with it..there is some proof always but if you go with what you feel ,not think, its always there to warn you

    /
     
    #299     Mar 10, 2008
  10. Jerry030

    Jerry030

    I I assume you are a discretionary trader?

    If you trade a mechanical system there are is no looking or feeling involved, only outcomes. My system trades without any human intervention at all. I could spend a month camping in the wilderness without any communication and as long as my PCs had electric power and an Internet connection it would continue to day trade the markets.
     
    #300     Mar 10, 2008