GARCH, etc.

Discussion in 'Options' started by aPismoClam, Jul 14, 2006.

  1. OK, geeks. Who wants to help a seat-of-the-pants trader get a handle on volatility calculations, not necessarily for options.

    I'm talking, the basics.

    Let's start with your definition of volatility.

  2. It would be my pleasure, first though a couple of quick questions

    Do you want GARCH, AGARCH, EGARCH or QGARCH?


    do you prefer Heteroscedastic or homoscedastic?

    You know that gets me to thinking

    Jesus (also known as Jesse) Christ was Heteroscedastic.
  3. Well Okay then;

    On a serious note, since you already have a reference for Vol, I can offer a good reference work for understanding Engle's GARCH (I assume you want GARCH and not the other flavors).

    "Non-Linear Time Series: Non Parametric and Parametric Methods"

    This is from the Springer Series in Statistics and it is used in many schools to teach the subject.

    Authors J.Fan and Q Yao.

    Published 2003 by Springer
  4. Since you posted not in a math but in a trader’s forum let’s start with the bottom line:

    GARCH has much to do with complex math and nobel prizes, but absolutely nothing, zero, zippo, nada, nix, null with trading or making money.

    Save yourself the time and don’t even think of digging into this.

    Just my suggestion.
  5. Curious why you say this.
  6. To discourage him/her from wasting time studying GARCH, again, from a practical viewpoint. To discuss the math the Wilmott forum may be more appropriate.

    But feel free to discuss GARCH here, of course. I just posted what I consider to be the trader's bottom line.
  7. You know its not the part of your comment that is wrong that I object to, it is the lack of context that kind of irritates me.

    Actually professionals do use GARCH, HOWEVER in order to make use of it, one has to have a very different understanding of it than one can usually obtain by reading a textbook, and/or discussing it with people of unknown expertise at a billboard site like ET.

    By the way, volatility is a measure of how quickly price varies over time.

    Back to what I was saying

    Robert Engle is credited with developing an approach called ARCH or AutoRegressive Conditional Heteroscedasticity. This precursor to GARCH is still used to model economic variables. Its value was that it provided a (relatively) quick estimation process that permitted the covariance matrix of a variable to vary with time. Before 1980 or so, you couldn't really do that.

    ARCH was popular with Options pros because it allowed you to calc the value of an option under the assumption that volatility varied with time and with respect to previous price "shocks".
    The underlying assumption of ARCH is that volatility always reverts back to its long term historical average.

    Assuming that you prob don't know what "covariance" is:

    Covariance is a measure of the way that two variables move with respect to one another.

    Okay then I don't think I can say it any simpler than that. At some point you're going to have to read some of the materials so that you can ask your own questions.

    I hope it helps

  8. Well that’s great. Now could you share a recent trade and explain how (G)ARCH made it work?
  9. Alright then:

    I have had a few moments to think and cool down

    This is a mathematical tool. In the hands of a skilled person (like me for instance) it can yield an answer that can make the difference between winning and losing. In the hands of an unskilled person or a person whose understanding is limited (you for instance) well, it just sits there and you have nothing to show for your time.

    What you have so far is a few references and some background information. If you want to make use of this mathematical tool YOU will need to do some footwork. YOU will need to develop an understanding of the process. The next step isn't up to me, its up to you..

    I don't mind helping you or anyone else as long I have the feeling that YOU are meeting me halfway. Asking me to wrap it with a pretty bow doesn't strike me as sincere interest. Your comment strikes me more like when I put a "for sale" sign in front of my house and the neighbors come over to get a look. They really aren't interest in buying my house, they just want to see the inside of my home without having to invest anything more than a few minutes of their time.

    Let me know if/and/or when you have a real interest in the subject.

    Now if you folks will excuse me, its Friday night, and I can think of a few things I'd like to do.

    #10     Jul 15, 2006