Models Behaving Badly

Discussion in 'Wall St. News' started by MKTrader, Nov 22, 2011.

  1. http://finance.yahoo.com/blogs/dail...-led-mf-global-collapse-people-174954374.html

    Interesting take on MF Global. This professor has built models his entire life, in both physics and finance, and had this to say:

    "Taking models too seriously, especially in the social sciences, it's sort of a form of idolotry...you think you can write down an equation that's going to explain the way people and markets behave."

    Amen. Academic economists should take note as well.
     
  2. I guess I was thinking something else, regarding the title of this thread.
     
  3. srsly, this thread was kind of a let down lol
     
  4. I was hoping this was sort of a "girls gone wild" thread with hot models.

    Severely dissapointed. :(
     
  5. I didn't come up with the title, but it's known as "clever ad copy to get your attention." Since it was posted under Wall Street News, didn't you think it may have been related to the markets?

    Seriously, three people reply with the same asinine comment?

    There are a million other websites out there with the stuff you're looking for. Sheesh, I knew a lot of juvenile types (including "adult" juveniles) frequented this site, but didn't know it was this bad.
     
  6. nitro

    nitro

    Thanks for posting that. It is interesting that this is the current view. My own model is doing extremely well, but I have a theory about why people go in and out of favor with backing mathematical models in markets.

    Since I don't have to answer to anyone, I am completely free to be creative and to continue to plow forward when things don't go well. That means that as I learn things, I am very quick to incorporate theory and ideas into changes in the model. The one thing that doesn't waver is my core belief that it is possible, but that any failing is my own simplistic view, and that the more general view is not incorporated into the model that would make it more correct in more markets.

    I suspect that there is a culture of mistrust of the mathematics and theory where these people work because they have huge bets on and the pressure is to make money with a model that is working now. Changing it is seen as suspect at best.

    Never in my life have I been more confident that my model captures a large part of the dynamics of the markets that I trade, but I am constantly on the lookout for evolution of markets, and that I must evolve with them as well. This is extremely complicated, but it is not impossible.

    One of the things that I have learned is that there is always someone that has figured out something deep, and just because I can't do it doesn't mean it can't be done.
     
  7. It is not the models that misbehave, it is the market. :)

    Models that expect mild randomness simply fail to understand the chaotic nature of markets.

    A good rule of thumb is: if the model is based on log normal markets... it is useless.