How Does Statistics Apply To Trading?

Discussion in 'Strategy Development' started by First Point LLC, Jan 14, 2013.

  1. I know some of you are probably thinking...."Wtf? Is this guy serious?" I am...

    Currently, I trade using a simple method of studying price action and tape reading which serves its purpose pretty well. But I keep hearing about statistics, which has gotten me curious enough to start studying statistics on my own. But I still haven't figured out how you use it w/ trading. The only form of statistics pertaining to the markets I understand is calculating moving averages.

    Where would be a good point (even books to read) that would explain how statistics and mathematics overall is used in trading?
     
  2. Two main topics would be a good start:

    1. Statistically Significant Sample Size

    2. Distribution Types

    The basic idea is to design your trading system around something you think the market does with regularity.

    Then, code a program that trades this idea automatically.

    Then run it over past data.

    Statistics come in at this point, when the two topics above must be reviewed.

    Does your number of trades encompass a statistically significant sample size, and are the win / loss, profit amount vs loss amount distributed in such a way to give you future confidence?

    :)
     
  3. Wide Tailz gave you a good response but the issue you will have is the quote above. These types of simple methods usually require a very "Subjective" level of interpretation and therefore are nearly impossible to capture in a trading algorithm that would allow you to apply statistical analysis. Most traders think they have a system but in reality it is usually a judgement call based on experience.
     



  4. Probability for Dummies by Deborah Rumsey

    Evidence-Based Technical Analysis: Applying the Scientific Method and Statistical Inference to Trading Signals... by David R. Aronson


    Read those two books and you will have a solid foundation of how statistics applies to trading.

    If you need to get more in-depth find out the textbooks used by major ivy league curriculums (grad and undergrad) and get a hold of their textbooks
     
  5. ===========
    F Point;
    take your favorite stock or commodity. Do [m] mean ,[m] median,,
    [m]mode study on many time frames. You will find some good wisdom doing that, if you do it enough ; No substitute for plenty/plenty of work. You will not see much if you dont do much work on that.

    Wisdom is profitable to direct:cool:
    Hope this helps it helps me.

    murray t turtle
     
  6. eurusdzn

    eurusdzn

    "Wide Tailz gave you a good response but the issue you will have is the quote above. These types of simple methods usually require a very "Subjective" level of interpretation and therefore are nearly impossible to capture in a trading algorithm that would allow you to apply statistical analysis. Most traders think they have a system but in reality it is usually a judgement call based on experience."

    Slave2Market,

    Would you please elaborate as I do not understand. For example, I thought that a method described by EOD close crossing above ma(5) as entry and EOD close crossing below ma(5) as an exit, with a stop loss, running against 100 stocks over 20 years is "objective" and easy to code. Therefore I could apply
    "stats 101" to this data. No?
     
  7. Bob111

    Bob111

    sure..but you will be disappointed with results :)
     
  8. eurusdzn

    eurusdzn

    Yes I agree, I mention it only as an example to frame a question.
    Whoops, I. Did not see the quote S2Mwas referring to.. never mind.
     
  9. A "moving average" is not "statistics"... it's an average...
    A primitive, high school level construct.

    Statistics are methods that tell you whether something is meaningful...
    As opposed to being "fooled by randomness".

    You cannot succeed as a trader...
    Unless you have a very strong gut feel for stats...
    And can differentiate from randomness.

    You must know when you are gambling... most have no clue.
    So a good place to start is knowing gambling math cold.

    [​IMG]
     
  10. vicirek

    vicirek

    Moving average is more than that. It is part of signal processing body of knowledge and depending on the form being used it is one of the band pass filters. Superficially it looks like simple high school concept. It is powerful tool but with questionable applicability to financial markets because of inherent lag.

    Statistics has very limited and ancillary applicability to financial markets because of inherent randomness of market signals combined with man made randomization (computer aided) introduced to markets on purpose to confuse outsiders even more.
     
    #10     Jan 14, 2013