Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Quantum Economies: Phyisical Modeling of Economic Systems

  1. Nov 16, 2009 #1
    Hello, this article caught my attention:

    "Physical Modelling of Quantum Economies"
    http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1304630

    It intents to develop dynamic economic models based on physical systems. So for example: it introduces the concept of "price-space"; substitutes homo oscillans (oscillating man) instead of homo economicus; "wave function of the economy" obviously analogous to the physics counterpart, it even borrows the letter ψ! For some reason, the author reserves the Hamiltonian operator for "further application", hoping that maaaybe someday somebody expands the theory and actually finds some use for it. Further comparisons are drawn to Schrodinger's equation... excited states....

    It doesn't seem all too crazy for me. Could this be succesful? To compare economic behavior with physics?

    Also, I am fairly familiar with the mathematics used for the classical economic models described in the paper (i'm an economics student) but I don't quite understand the math used in the "quantum economy" part. Help with it please? Would you like to share your impressions?

    mariana
     
    Last edited: Nov 17, 2009
  2. jcsd
  3. Nov 16, 2009 #2
    It's about who you know, not what you know. Maybe that's business, but how many people in the world have a good command over mathematics? Would it make sense to create a mathematical model to explain the economic actions of a mathematically poor people, especially theoritical physics?
     
  4. Nov 17, 2009 #3
    Im sorry, I dont understand what you mean. You mean does it make sense to create a model to explain decisions of people in an economic framework? Well, its some of the things economics does, I believe the mathematical approach is different to that of physics, and precisely what is discussed in the paper, to adopt and apply concepts used to physics in order to make more accurate descriptions in certain conditions. I have yet to further stufy the formal aspects of it, but my impression is that this idea could have some success.
     
    Last edited: Nov 17, 2009
  5. Nov 19, 2009 #4

    Fra

    User Avatar

    I didn't read that paper but I can make a personal general comment.

    IMO, there are strong parallells to QM and the foundation of modern economy (which is game theory).

    Each actor in the market are assumed to act, rationally to maximize it's own benefit. This is used to infer a global dynamics of economy. The market is a game, where assumptions of how each player acts, yields predictions for the overall game evolution.

    There are parallells to physics here, where each physical systems acts only upon it's own information about it's environmnet!

    I think some of the best analogies, that can be used to explain QM conceptually are in terms of games. For example how a player acts not upon a random choice, but rather that he single action is based upon the entire space of possibilities - spreading your risks and increasing your chances of victory.

    So I think the general connection is very interesting, but this is still an active research area. And I would not say that it's just that economist can learn from physics, it's also the other way around and physics can learn to see the laws of physics with a different abstraction. As we know, the geometrical abstraction has been dominating for a long time. Here the game theoretic angle is IMO a much more plausible and natural abstraction.

    This angle actually also comes with a very interesting interpretaion of the wave function in QM. It's a intrinsic information the player has about it's environment. The rational action from game theory can be exploited to infer a physical action of a physical system, by constructing a measure to be extremized. This is what the least action principle is in physics.

    But this is not mainstream interpretations, but I personally think this is among the more interesting developments.

    /Fredrik
     
  6. Nov 19, 2009 #5
    It has a looong way to go... why does it seem to never get it right? Models in economics appear to be a better tool for abstraction/education but... do they help for (a decent level of) prediction?
     
    Last edited: Nov 19, 2009
  7. Nov 19, 2009 #6

    Fra

    User Avatar

    I'm not economist, but the general reasoning behind in game theory is generally very intuitive, but I think the regular descriptions are very idealistic, so it's not mature yet IMO. In particular the idea that there exists objective measures of rationality or objective rules of the game.

    At first, one may think w/o objective measures we loose predictability, but that's not true. It just means that predictions are relative, and justified relative to each player. I think each player has his own idea of the game. And the only way to compare the views are for the players to interact and the objectivity is IMO only emergent as a resulting negotiation.

    But I mean the general framework of "game theoretic" thinking, is IMO much more natural to me at least, than the geometric thinking or other abstractions. That's what I like about it, but I agree that there is alot yet to complete.

    To predict a global economy is clearly not an easy task. We can't even predict the global weather :)

    But in the game theoretic angle, I think part of the point is that there IS no perfect predictions, this is why the modelling itself is part of the game. This is my view.

    I personally think the game theoretic view should be combined with an evolutionary view, where the rules of the game itself is changing. My view is that the rules of the game are relative and encoded in each player, so tha the population of the players encodes the variety of the rules. So the ambition to have objective rules, or observer independent laws of physics, only makes sense when considering a population in equilibrium. A changing population, indicates that at some level there is an evolution of the game. It's like a negotiated democracy defining the rules.

    But even if we understand this better, predicting everything is imopssible. I think this abstraction implies that part of the evolution is unpredictable, and we simply have to act on the feedback the future gives. I think this is also the key to explain the action of some systems, as it's indifferent to information that it doesn't see. The rationality measure must scale with the scale of the player.

    A player might occure irrational or "stupid" relative to another player, but where in fact, the problem might be that the problem is tha the second player tries to impose HIS measure of rationality to another player - which is wrong.

    Edit: my point in this last comment is that the scientists trying to do the game theoretic inferences here are not external observers juding the game, they are themselves just players in the same game!

    /Fredrik
     
  8. Nov 19, 2009 #7

    Fra

    User Avatar

    My objection to the fixed rules, and action spaces of game theory are direct parallells to my objection to current quantum theory NOT beeing a proper intrinsic measurement theory.

    So in that sense I think both are plagued by the same issue. But I think that the game theoretic abstraction to me at least is far more natural, plausible and intutitive than some other geometric abstractions.

    In geometry, the least action principe can be assodicated to geodesics on a manifold.

    In game theory the least action principle can be associated to a principle of minimum speculation in the sense that the evolution of the game is like a random walk.

    It's somehow two possible abstractins of the same thing.

    Just like there IMO is no global objective manifold, there is no objective global game.

    There are only inside views, and in the inside view, the evolutionary perspective is apparently unavoidable.


    /Fredrik
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook