Mathematically Predicting the Future?

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

  1. MarkBrown


    I think anyone can make money at the end of the day and it doesn't even have to be human to do it, likewise they can loose.

    Math has no emotions, it doesn't care about anything but yet provides a answer. You can take or leave that answer. Let's put the math question this way.

    Instead of trying to prove math can predict the markets let's consider way's it can be made to fail. In the end we will be left hopefully with what works. The good ole trial and error learning method.

    Is it a fact that on catastrophic news to the SP that we can expect Bonds to spike down? If this is not true then we have eliminated one thing that we trust does not work. We continue on this path until we have no excuse to not own the world. After eliminating everything that does not work, that is.

    #41     Oct 29, 2007
  2. I agree math is very important in trading. For instance when I put on a trade I calculate what I am expecting to get as a profit on the trade and then i calculate what my expected loss is. If the R/R is good I take the trade. I do all of this addition and subtraction in my head (my teacher will never believe it):D

    But seriously other than addition subtraction a little bit of multiplication and division you don't need any other types of math in trading. I mean im taking algebra and calculus and to be honest can't come up with one good way to predict where prices are headed tomorrow from it.

    People are weird creatures. They don't want to think things are mostly random. They need an explination for the way things act whether it be because of the planets lining up or by some complicated math equation. But can you blame them? I mean we have explanations for everything else in this world why not the markets? It all comes down to supply and demand folks. Markets are not a pure science like physics and math they are AT BEST a social science.
    #42     Oct 29, 2007
  3. MarkBrown


    you r as u say a young trader. good 4 u. you have nothing else to learn. mb
    #43     Oct 29, 2007
  4. Mark-

    "Mathematically is it possible, can man define the markets and surround it with a mathematical fence? Pull the market around on a leash like you were walking a dog?"

    I’d really hate to rain on anyone’s parade that thinks defining the market is improbable let alone practically impossible to achieve. A mathematical model that predicts future market movement not only does exist but it has pivot times, defines expected movement (i.e. shape), along with levels of support and resistance as illustrated below.

    Additional examples sit in the archives on this web site and have been sitting there for more than a year. You just need to do a little research to find it.

    The model not only predicts intraday movements but can isolate specific turning points days in advance down to the hour and major turning points months in advance down to the day. I’m sorry if the skeptics find this hard to swallow however, this model has been in existence for the past 10 years.

    Due to the proprietary nature of the work details cannot be revealed. I write this simply to demonstrate that the model does exist and that the market is not as random as one would have you believe.

    My2 ¢¢
    #44     Oct 29, 2007
  5. MarkBrown


    i think if anyone is going to have a math model that will predict the future. it would be clearly a buy it sell it model. not something that has perfected the art of confusion. like support will be found tomorrow at 9:30 around the 1500 level and resistance will be found around 12:00 at the 1505 level etc.

    no work here expected the math model in my mind would just say i am long from here or short from there and thats it plain and simple. subject to no interpretation at all in fact what would be better is if the thing automatically placed the trades and made money which was deposited to my account even while i sleep.

    the prediction part would be great cause then i could possibly have the model leverage the trade size according to the confidence levels.


    i think at least it is possible to have a math formula that will predict the future and hang it all out by saying buy or sell unconditionally.
    #45     Oct 30, 2007
  6. Good discussion. When you say "buy it sell it", are you referring to a SAR type of system that is always in the market (Stop And Reverse) ?
    Clearly this discussion should also review the various "premises" embodied in any winning trading system which by it's nature must be able to successfully "predict" the future price:
    1) reversion-to-mean
    2) trend-following
    3) break-out
    4) trend-reversal
    5) momentum
    6) ??
    #46     Oct 30, 2007
  7. gnome


    Years ago Futures Magazine did a study on "predicting the market with artificial intelligence".

    Their final conclusion... "there is a small degree of predictability for the next bar, only".
    #47     Oct 30, 2007
  8. Hello MarkBrown,

    mathematically speaking: Yes, it is possible to "model" the market...but the concept has one significant flaw:

    The number of variables to describe the system is far too big!!!

    On top of that, the variables are changing constantly, so that your models needs to adapt itself all the time.

    I've studied physics...basically speaking "the science of modeling the world as we see it". Here is something to think about: The physics of this world has not changed since the dawn of time, but yet we still havn't found "The theory" that describes it all!

    The only chance in my opinion might be to view the market as a complex system and see if it is possible to draw out some useful information about the system itself...much like for e.g. temperature in thermodynamics.

    If you are looking for "The Indicator" that describes it all and makes you one hell of a lot of money...your journey will be long...maybe too long for you to live...

    You should check out some of Van Tharps materials. He and Tom Basso have shown that a RANDOM ENTRY SIGNAL can be profitable dealing the makets with the appropriate position sizing and money management in place! So don't devote too much time with entry signals, since its only a part of of a whole trading system.

    Btw, I actually made the mistake of searching for that killer indicator...I even wrote my master thesis on it, which turned out to have a major postdictive error in it! Now I've mentally moved on to try to view system development in a larger picture than just the entry. I'm still researching, but hopefully I'm on the right road to success.

    I hope I could give some insight and all the best for your future trading!
    #48     Oct 30, 2007
  9. Mark-

    The subject of this particular topic is: “Mathematically Predicting the Future?” correct?

    The question at hand that we are discussing is: Has it Can it be done?

    Or colloquially, “Mathematically is it possible, can man define the markets and surround it with a mathematical fence?”

    This is the question that I addressed. I came to the table to illustrate just a fraction of what is currently being accomplished by the program traders. If you saw the entire picture it would be an eye opener for you.

    Mark you gone off on a related tangent here and are now discussing systems.

    Syswizard beat me to the punch. Although my question is why bother to do all the work Mark? You're describing one of the better reactive trend following systems. You can buy a myriad of those online for your generic black box or talk to the people at Sybase they have a great one, set it and forget it.

    Maybe a better way to formulate your question is by asking does the holy grail of forward looking trading systems exist today? Sure it does, for today at least, that is until the sniffer programs track it down and the wonks start to program against it.

    #49     Oct 30, 2007
  10. I think its probablly a fallacy to think you need to predict the market to make money from it. That's probably the toughest way. Anticipate (or guess) and then manage accordingly to what actually happens.

    Having said that the market certainly seems to have a 'memory' I guess that's because the participants all have memories.

    I'd settle for the one bar into the future that Gnome mentions, now if you can just tell me next months high & low on the S&P I'll be a happy camper.:)
    #50     Oct 30, 2007