4 ms·

econophysicists? Wtf? If you want to avoid the term "economics" because most people associate it with "made up financial bollocks based on applying completely

econophysicists? Wtf? If you want to avoid the term "economics" because most people associate it with "made up financial bollocks based on applying completely inappropriate models from unrelated disciplines", then why include the "econo" part. Also, you appear to be trying to be modelling a financial phenomenon by applying a completely inappropriate model from an unrelated discipline.

jasonwatkinspdx 13y ago
Derivatives trading is now dominated by quants with a background in the methods of computational physics. Aspects of gauge theory, etc are clearly and directly relevant. Don't reject this idea out of hand because it is unfamiliar to you. It may be proven a fad eventually, but that's not at all obvious now. As a concrete example: the Higgs result depends both on machine learning in the form of decision trees to minimize the data, as well as monte carlo simulation to generate theoretical data for comparison to what is measured. Experimental particle physics is very much now an exercise in machine learning and probabilistic simulation.
- andyjohnson0 13y ago
  Physical laws exist. Objects in the real world have to follow those laws. I accept (I think) your point that ML and simulation can be used to "discover" those laws in some inductive sense. What I'm having a hard time with is the idea that markets necessarily have "laws" that are "discoverable" in the same way (which is what I think you're saying). They might do, and clearly some smarter people than me believe that they do, but where is the justification for this? @contingencies mentioned [1] a George Soros quote about classical economics being based on a false analogy with Newtonian physics and markets are just the behaviour of individuals. Observations of particles tell you the behaviour of the particles, but the behaviour is fully determined by underlying physical laws. I don't see how this is true of behaviour of actors in economics, except in the obvious respect that the bodies of traders and the hardware of trading systems are subject to the laws of physics. What am I missing here? I also wonder if there is a danger of building automated trading systems that are based on ideas from econophysics, and using their resulting behaviour are proof that the ideas were correct in the first place. [1] https://news.ycombinator.com/item?id=6481663 https://news.ycombinator.com/item?id=6481663
  - count 13y ago
    Physical laws are postulated to exist, and we make models that are proven and disproven by theory. Newtownian physics was replaced by relativity, which is being muddled by quantum physics. Physics is not cut and dried, so why should economics be? To use your own example, the market is made up of the actions of individual actors, who are each governed by underlying mental models/laws (such as a probability field around making a trade, etc.). The OP is trying to discover what those models/laws ARE.
  - jasonwatkinspdx 13y ago
    > I accept (I think) your point that ML and simulation can be used to "discover" those laws in some inductive sense. Yes, we use these methods to confirm that reality matches some candidate law. I personally don't think it's very meaningful to talk about physical laws existing in some independent platonic sense, but the debate over the principle of induction and the epistemology of science is a side story here IMHO. Ultimately what we care about is that a model successfully predicts reality in the future. I think a useful conceptual division is: with some models, the internal structure of the model gives us understanding of the physical behavior. Most of what we call "laws of physics" are in this category. We can look at the algebraic expression and understand symmetries, constants, conserved quantities, etc. A second type of model is purely algorithmic and utilitarian: it gets the right answer, but the internal structure of the model doesn't provide any comprehensible information or understanding. Most ML algorithms fall in the latter camp. What muddies the issue is that experimental physics tries to validate models of the first kind, but now generally requires methods of the second kind. We're forced into these methods because of the complexity of what we're studying (particle physics can involve thousands of Feynman diagrams), the noisiness of our measuring equipment, and the rareness of the events we're searching for. When I say physics methods are very relevant to wall street, I'm talking about the latter methods more than any attempt to find fundamental theorems.
- lotsofcows 13y ago
  As with most market cycles, if you get enough people following a common philosophy, the market will conform to the philosophy... for a while. In the past 100 or so years economics has completely failed to define itself as a science. You're using the wrong tense. Anyway, we're drifting off-topic - I was bitching mainly about the title and a little about the misuse of Zipf's law. If the article was written in reverse, finishing with "this closely matches a power curve from which we can estimate x" then I would only have had a problem with the title "econophysicists".
axus 13y ago
Reminded me of "psycho-historians" :)

jasonwatkinspdx 13y ago

Derivatives trading is now dominated by quants with a background in the methods of computational physics. Aspects of gauge theory, etc are clearly and directly relevant. Don't reject this idea out of hand because it is unfamiliar to you. It may be proven a fad eventually, but that's not at all obvious now. As a concrete example: the Higgs result depends both on machine learning in the form of decision trees to minimize the data, as well as monte carlo simulation to generate theoretical data for comparison to what is measured. Experimental particle physics is very much now an exercise in machine learning and probabilistic simulation.

andyjohnson0 13y ago
Physical laws exist. Objects in the real world have to follow those laws. I accept (I think) your point that ML and simulation can be used to "discover" those laws in some inductive sense. What I'm having a hard time with is the idea that markets necessarily have "laws" that are "discoverable" in the same way (which is what I think you're saying). They might do, and clearly some smarter people than me believe that they do, but where is the justification for this? @contingencies mentioned [1] a George Soros quote about classical economics being based on a false analogy with Newtonian physics and markets are just the behaviour of individuals. Observations of particles tell you the behaviour of the particles, but the behaviour is fully determined by underlying physical laws. I don't see how this is true of behaviour of actors in economics, except in the obvious respect that the bodies of traders and the hardware of trading systems are subject to the laws of physics. What am I missing here? I also wonder if there is a danger of building automated trading systems that are based on ideas from econophysics, and using their resulting behaviour are proof that the ideas were correct in the first place. [1] https://news.ycombinator.com/item?id=6481663 https://news.ycombinator.com/item?id=6481663
- count 13y ago
  Physical laws are postulated to exist, and we make models that are proven and disproven by theory. Newtownian physics was replaced by relativity, which is being muddled by quantum physics. Physics is not cut and dried, so why should economics be? To use your own example, the market is made up of the actions of individual actors, who are each governed by underlying mental models/laws (such as a probability field around making a trade, etc.). The OP is trying to discover what those models/laws ARE.
- jasonwatkinspdx 13y ago
  > I accept (I think) your point that ML and simulation can be used to "discover" those laws in some inductive sense. Yes, we use these methods to confirm that reality matches some candidate law. I personally don't think it's very meaningful to talk about physical laws existing in some independent platonic sense, but the debate over the principle of induction and the epistemology of science is a side story here IMHO. Ultimately what we care about is that a model successfully predicts reality in the future. I think a useful conceptual division is: with some models, the internal structure of the model gives us understanding of the physical behavior. Most of what we call "laws of physics" are in this category. We can look at the algebraic expression and understand symmetries, constants, conserved quantities, etc. A second type of model is purely algorithmic and utilitarian: it gets the right answer, but the internal structure of the model doesn't provide any comprehensible information or understanding. Most ML algorithms fall in the latter camp. What muddies the issue is that experimental physics tries to validate models of the first kind, but now generally requires methods of the second kind. We're forced into these methods because of the complexity of what we're studying (particle physics can involve thousands of Feynman diagrams), the noisiness of our measuring equipment, and the rareness of the events we're searching for. When I say physics methods are very relevant to wall street, I'm talking about the latter methods more than any attempt to find fundamental theorems.
lotsofcows 13y ago
As with most market cycles, if you get enough people following a common philosophy, the market will conform to the philosophy... for a while. In the past 100 or so years economics has completely failed to define itself as a science. You're using the wrong tense. Anyway, we're drifting off-topic - I was bitching mainly about the title and a little about the misuse of Zipf's law. If the article was written in reverse, finishing with "this closely matches a power curve from which we can estimate x" then I would only have had a problem with the title "econophysicists".

axus 13y ago

Reminded me of "psycho-historians" :)