PDA

View Full Version : Econophysics??



Oikos-nomos
01-30-2008, 07:01 PM
Could somebody please tell me what the heck is Econophysics???
The first time i read was in pitt's webpage, there was a guy there that did that thing...but maybe because of his nerdy look and the fact he already had a Physics Phd before he got one in Economics, i thought that that guy had invented it.
But today reading some post...jeez...it was there too...Econophysics!!!...

Well, if someone could explain more or less what it is about it'd be great. It's just that i can't figure it out...

Thanx

bertthepuppy
01-30-2008, 07:23 PM
Do you know who it is at Pitt? There are a few professors in the Econ dept here that are physicists first, then Economists; Feigenbaum is one of them.

Oikos-nomos
01-30-2008, 07:34 PM
Yeah, it was him.

Profgif
01-30-2008, 08:09 PM
There are several papers in the literature introducing econophysics: econophysics - Google Scholar (http://scholar.google.com/scholar?num=50&hl=en&lr=&q=econophysics&btnG=Search). I will give you my general and probably biased impressions (I have done some work in the field). The most critical of the econophysicists believe that economics is based in more or less arbitrary assumptions (like the utility maximizing behavior), assumptions which have failed empirical tests and which should be disregarded. Thus, a new science of economics should be built starting with empirical data, just like physics is supposed to have originated. In response, some econophysicists have been accused of economic ignorance (although a few of them are very well versed in economics too). The achievements of econophysics after more than 10 years (econophysics was founded in 1995) are still small, but there are some interesting works, for example in the distribution of wealth and in financial time series and financial derivatives. Such works use, for example, insights from statistical mechanics, from diffusion equations or from quantum path integrals. In my view, econophysics is more about applying a rigorous empirical and scientific approach to economic data to build empirically valid theories using mathematics as advanced as necessary, rather than blindly applying theories from physics into economics (in fact, much of standard economics appeared in the second way, with economists copying isocurves or the Le Chatelier principle from thermodynamics).

econphilomath
01-30-2008, 08:52 PM
The achievements of econophysics after more than 10 years (econophysics was founded in 1995) are still small...


Not surprising.:crazy: What is expected to be found? The great law of motion for stock prices..... Sounds like charting crap to me.


... blindly applying theories from physics into economics (in fact, much of standard economics appeared in the second way, with economists copying isocurves or the Le Chatelier principle from thermodynamics).

I'd rather have economist blindly apply physics than physics blindly apply economics....


Sorry if I went overboard, I'm a little grumpy nothing is showing in my email and my mouse broke.

filroz
01-30-2008, 09:06 PM
I have heard of some study that shows that wealth in society is distributed according to some power law common if physics

Profgif
01-30-2008, 09:26 PM
The debate is hot, see for example the response of some economists (http://www.debunking-economics.com/Papers/Econophysics/GallegattiKeenLuxOrmerod2006WorryingTrendsInEconop hysics_PhysicaA370pp1-6.pdf#ggviewer-offsite-nav-4142096) to econophysics and the (http://arxiv.org/abs/physics/0606002#ggviewer-offsite-nav-4142096)[somewhat frantic] (http://arxiv.org/abs/physics/0606002#ggviewer-offsite-nav-4142096)response from an econophysicist (http://arxiv.org/abs/physics/0606002#ggviewer-offsite-nav-4142096). While I prefer not to take positions in the debate, I have to say that there is indeed much (mostly unintended) crap in econophysics, but there are some very well prepared researchers in the field too, and they have developed quite interesting models with some preliminary results, especially in the analysis of financial time series.

bscout
01-30-2008, 09:29 PM
Not surprising.:crazy: What is expected to be found?
What about Brownian Motions?

econphilomath
01-30-2008, 10:04 PM
What about Brownian Motions?

What about them? Do you mean to imply that is something this "field" has done?

It's just another tool brought into economics to model something. The fact math and stats are useful for studying economics, doesn't mean you drop the economics (models with assumptions, concepts etc) and stay with just the math and stats! Its like just watching the time series on your computer screen until a pattern jumps out at you, Oh have a paper, when x goes up, y goes down! (we know its P and Q duh)

I'm not against questioning assumptions, and we need our theories to be based or at least be related to empirical facts, yes. But treating economics like an exact science is hogwash.

econphilomath
01-30-2008, 10:09 PM
Good links, Profgif thanks.

Read the paper, convinces me its crap. Maybe in finance with lots of data but still I don't buy the universal empirical regularity argument.

Oikos-nomos
01-31-2008, 12:12 PM
Yeah, great links and introduction to the subject Profgirl! Thanks a lot.

My God. People, please read the last paragraph of the answer to [physics/0606002] Response to Worrying Trends in Econophysics (http://arxiv.org/abs/physics/0606002#ggviewer-offsite-nav-4142096) (a critique to econophysics), made by an econophysicist:


The real problem with my proposal for the future of
economics departments is that current economics and
finance students typically do not know enough
mathematics to understand (a) what econophysicists are
doing, or (b) to evaluate the neo-classical model (known
in the trade as ‘The Citadel’) critically enough to see, as
Alan Kirman [54] put it, that ‘No amount of attention to
the walls will prevent The Citadel from being empty’. I
therefore suggest that the economists revise their
curriculum and require that the following topics be
taught: calculus through the advanced level, ordinary
differential equations (including advanced), partial
differential equations (including Green functions),
classical mechanics through modern nonlinear dynamics,
statistical physics, stochastic processes (including solving
Smoluchowski-Fokker-Planck equations), computer
programming (C, Pascal, etc.) and, for complexity, cell
biology. Time for such classes can be obtained in part by
eliminating micro- and macro-economics classes from the
curriculum. The students will then face a much harder
curriculum, and those who survive will come out ahead.
So might society as a whole.

I just wanna kill this guy...:mad:, he's too blinded by Physics' fundamentalism. This is what happens when someone irrationally believes that what one is doing is the ultimate truth, and starts building upon it. By being so blind, he's losing the essence of what makes Economics a science: the theory, and therefore the capacity to think critically about the nature of the phenomena we are trying to explain.

buckykatt
02-01-2008, 06:39 AM
/me rolls eyes

Of course, one wonders whether sociologists roll their eyes in a similar fashion when they read economists writing about sociology. ;)

Oikos-nomos
02-01-2008, 02:51 PM
mmm...i never thought of that!!! That's incredible...i always thought that sociology lacked formal reasoning...haha...i'm not very different from that physicist!!! Shame on me...:blush:

pevdoki1
02-01-2008, 03:11 PM
Eliminate micro and macro classes to teach statistical physics and cell biology?? :mad::mad::mad:

I take it he's talking about the undergraduate curriculum. The undergraduates don't give a crap about econonophysics; the majority of them simply wants to learn how markets work (so they can get a job in the private sector). Why the hell would they need courses in classical mechanics and statistical physics? Maybe statistical economics makes a little more sense? And the minority, the ones applying to graduate school, will get their mathematics done anyway.

I want to study macroeconomics.. How the hell will cell biology help me understand monetary policy? I'm interested in biology; perhaps I would even take a cell biology course if my schedule wasn't packed with Analysis, Statistics etc. But making it a requirement is absurd.

On the other hand, the kind of coursework he's suggesting will certainly develop one's critical thinking skills to the point where the person is ready for grad school. I would add courses in intermediate macro and micro and some theoretical math like analysis and topology. But where will the passion for economics come from?

P.S. I did not read the full article and I am only commenting on the selection of courses above.

econphilomath
02-01-2008, 03:21 PM
Eliminate micro and macro classes to teach statistical physics and cell biology?? :mad::mad::mad:

I take it he's talking about the undergraduate curriculum. The undergraduates don't give a crap about econonophysics; the majority of them simply wants to learn how markets work (so they can get a job in the private sector). Why the hell would they need courses in classical mechanics and statistical physics? Maybe statistical economics makes a little more sense for them?

And the ones applying to graduate school? Well, they will get their mathematics done anyway!!

I want to study macroeconomics.. How the hell will cell biology help me understand monetary policy? Granted, I'm interested in biology and read about it a little in my spare time; perhaps I would even take a course if my schedule wasn't packed with Analysis, Statistics etc. But making it a requirement is absurd.

On the other hand, the kind of coursework he's suggesting will certainly develop one's critical thinking skills to the point where the person is ready for grad school. I would add courses in intermediate macro and micro and some theoretical math like analysis and topology. But where will the passion for economics come from?

We are talking some (not all) guys who think they study economics but do not know who Adam Smith is... what do you expect! The problem with economics, and it has been discussed in another thread related to changing courses in grade programs, is that people need tools but they also need ideas. There is a trade off here and some think its leaning to strongly to tools in the first two years and thus its boring and alienates some good econmic thinkers.

I propose a must read in my opinion: "The Worldly Philosophers". After problem sets from Stokey Lucas!

Profgif
02-03-2008, 09:02 PM
This is an example of what physicists can do with "statistical mechanics" universal statistical properties of poker tournaments (http://arxiv.org/abs/physics/0703122), a recent theoretical model which seems to perform well empirically. I also think that economics is too orthodox sometimes (and sociology even more), and too much orthodoxy often implies that there may be some lack of self-criticism. I am not implying that econophysics is the answer to the empirical challenges that standard economic assumptions face, (psychology has probably more to add than econophysics), but I do think that economics needs to admit more unorthodox approaches. Finally, as a physicist, I have to say that I have studied biology extensively (which teaches a lot about the workings of complex systems), and I am currently studying more mathematics, (because knowing more never hurts), and of course, I am also studying economics, which is why I'm here... :)

pdesmmhh
02-03-2008, 10:28 PM
I don't understand why everyone doesn't like this sentiment. If undergrad physicists can be made to understand advanced mathematical models and ideas, why can't undergrad economists. Furthermore, building tools as one of you stated would seemingly be much more useful for the student who is seeking a job rather than learning ideas. Let the ideas before the grad students who want to do research in the area. I guess my point is if were hiring a physicist with a B.S. I would expect that he understand special relativity, quantum mechanics, and P.D.E.s. Likewise I would want an economist to understand advanced micr/macro, real analysis and some advance stats.

polkaparty
02-03-2008, 11:02 PM
I don't understand why everyone doesn't like this sentiment. If undergrad physicists can be made to understand advanced mathematical models and ideas, why can't undergrad economists.

I have purposely avoided this discussion thus far, but the statement above perhaps reflects the main misunderstanding between physicists and economists. As others have said already, physics and economics are quite different. Sure there are parallels, but making analogies without thinking is ignorant.

The hard sciences are somewhat nicely split between theory (e.g., physics) and applications (e.g., engineering). Economics is not split as such. Undergraduate economics majors are a diverse group and requiring all economics majors to understand advanced theory is not optimal.

Anyway, I don't really want to go any further.

trjohnson
02-03-2008, 11:46 PM
Anyway, I don't really want to go any further.

Don't do it, polkaparty! You have so much to live for!

econphilomath
02-04-2008, 12:34 AM
I guess my point is if were hiring a physicist with a B.S. I would expect that he understand special relativity, quantum mechanics, and P.D.E.s. Likewise I would want an economist to understand advanced micr/macro, real analysis and some advance stats.

I would have to disagree with this statement.

In applied jobs people do not have time to model everything so they can study things in detail. People have to concentrate on making descions and usually base them on intuition which together with experience, points you (usually) in the right direction. Knowing tons of math or advanced and specific theories will generally not help very much in everyday work. That might explain your confusion and also the amount of people who study physics vs. economics at the undergrad.

If I working in the private sector and was hiring an economist to work on real world problems I would rather have an economist :
a. use cost-benifit way of thinking
b. know why the price system is a good way to organize a society and when might free markets not get the best resultsl.
c. understand the concept of "marginality" and be able to use it in making descisions
d. idea of sunken costs
e. why monopolies might be good or bad
f. what happens to investment if the interest rate goes up,
g. what the real exchange rate is and why it might depreciate
h. why inflation is bad and how monetary policy operates.
i. comparative advantage
j. what a stock or bond is and where it derives its value.

...to name a few concepts or ideas that are way more important for real life jobs and applications and which many grad students might not know/remember because they are worried about showing the existence and unicity of and equilibrium in some otherwise useless model.

Undergrad econ is for the masses which will not have the pleasure of spending their days in abstract thought and for this reason they should concentrate on simple models with lots of intuition. They will however be producing more for society than most of us so we should respect them for that.

pdesmmhh
02-04-2008, 04:19 AM
I would have to disagree with this statement.

In applied jobs people do not have time to model everything so they can study things in detail. People have to concentrate on making descions and usually base them on intuition which together with experience, points you (usually) in the right direction. Knowing tons of math or advanced and specific theories will generally not help very much in everyday work. That might explain your confusion and also the amount of people who study physics vs. economics at the undergrad.

If I working in the private sector and was hiring an economist to work on real world problems I would rather have an economist :
a. use cost-benifit way of thinking
b. know why the price system is a good way to organize a society and when might free markets not get the best resultsl.
c. understand the concept of "marginality" and be able to use it in making descisions
d. idea of sunken costs
e. why monopolies might be good or bad
f. what happens to investment if the interest rate goes up,
g. what the real exchange rate is and why it might depreciate
h. why inflation is bad and how monetary policy operates.
i. comparative advantage
j. what a stock or bond is and where it derives its value.

...to name a few concepts or ideas that are way more important for real life jobs and applications and which many grad students might not know/remember because they are worried about showing the existence and unicity of and equilibrium in some otherwise useless model.

Undergrad econ is for the masses which will not have the pleasure of spending their days in abstract thought and for this reason they should concentrate on simple models with lots of intuition. They will however be producing more for society than most of us so we should respect them for that.

My statement was intended to address the instance where in most LA Undergrad Economics programs, one spends an inordinant amount of credit hours on useless LA classes that could be better allocated to learning more econ and math. I don't quite understand why more can't be expected in the students primary field and less in other unrelated (for the most part) fields.

polkaparty
02-04-2008, 04:32 AM
Don't do it, polkaparty! You have so much to live for!

Heh, I'm still here =p. econphilomath did a nice job summarizing what I would have said I think.


My statement was intended to address the instance where in most LA Undergrad Economics programs, one spends an inordinant amount of credit hours on useless LA classes that could be better allocated to learning more econ and math. I don't quite understand why more can't be expected in the students primary field and less in other unrelated (for the most part) fields.

This is a different argument than your earlier one.

At least at my university, both economics and physics majors must take the standard slew of non-major liberal arts courses. So if you want to deride the value of liberal arts requirements, fine, but that's not relevant to the comparison between core requirements for economics and core requirements for physics.

pdesmmhh
02-04-2008, 05:16 AM
Heh, I'm still here =p. econphilomath did a nice job summarizing what I would have said I think.



This is a different argument than your earlier one.

At least at my university, both economics and physics majors must take the standard slew of non-major liberal arts courses. So if you want to deride the value of liberal arts requirements, fine, but that's not relevant to the comparison between core requirements for economics and core requirements for physics.

Not so much, perhaps I was unclear, but my point was that Physics seems to be more demanding in its field than econ. At my university, physics majors can use their physics classes toward their sciences general education requirement and they do not have to take a foreign languager. In econ, however, econ classes can't be used toward the general education requirement in the social sciences, students must satisfy a language requirement, and there is 12 additional credits (4 classes) which must be devoted to general education on top of the standard amount required of all undergrads. Thus, one must waste an additional semester and a half on useless drivel instead of learning valuable tools. Meanwhile the physics undergrads get to take more classes in areas that they might deem useful. I wish I could have used those 18 credits spent on that LA crap on more useful classes, not to mention the money.

polkaparty
02-04-2008, 05:27 AM
At my university, physics majors can use their physics classes toward their sciences general education requirement and they do not have to take a foreign languager. In econ, however, econ classes can't be used toward the general education requirement in the social sciences, students must satisfy a language requirement, and there is 12 additional credits (4 classes) which must be devoted to general education on top of the standard amount required of all undergrads.

You should take this up with your particular university, rather than the entire profession then.

I agree that generally physics majors are "more demanding" than economics majors, I just disagreed with your evidence in support of this proposition. The difference, in general, is not liberal arts requirements (although this is one reason at your school), it is the core requirements. And the core requirements are different because of the different professional needs of the students.

econphilomath
02-04-2008, 12:38 PM
I should note the diffrence between "demanding" and "usefull". While banging your head against the wall may be demanding, that doesn't make it useful.

Valhalla
02-04-2008, 01:12 PM
I should note the diffrence between "demanding" and "usefull". While banging your head against the wall may be demanding, that doesn't make it useful.

:tup:[clap]:D:D[clap]:tup:

Mr.Keen
02-04-2008, 05:56 PM
g. what the real exchange rate is and why it might depreciate

I know quite a few economists who don't know this one. Moreover, even those who "know" would argue among themselves about what it is. RXR are tricky.:)

Profgif
02-04-2008, 05:56 PM
Very nice discussion. In fact, sometimes "hard" scientists forget that what most of the economics students need is an applied and directly useful learning approach.

I feel that I have to clarify something: perhaps I was unconsciously devious in choosing an article by Joseph L. McCauley as an example of the field ("Response to Worrying Trends in Econophysics" which contains the quoted piece about the economics curriculum). McCauley is well known to be a fanatic extremist. However, I ultimately chose to quote his response because, despite his extremism, I respect him very much as a researcher, (he is probably one of the best prepared in the field and he probably knows as much economics as a professional economist). The thing is that even in econophysics, among physicists, there have been several dead-end research lines which arose because people didn't quite understand the mathematics involved. McCauley has been one of the few that has shown (that could show?) where the nonsense in econophysics was, (and I guess that many physicists would fear to have him as a referee, because he is as hard with physicists as he is with economists).

Anyway, most (econo)physicists don't think like McCauley, and there are many who would totally agree with the arguments that have been offered here in support of an applied curriculum, (including me).

econphilomath
02-04-2008, 06:19 PM
I know quite a few economists who don't know this one. Moreover, even those who "know" would argue among themselves about what it is. RXR are tricky.:)

Good point, I really meant -to know what the real exchange rate is- and what it means when it changes... not why, thats a lot harder. Also I didn't mean explaining the nominal exchange rate, especially in the short run! Who knows right?!

Oikos-nomos
02-06-2008, 12:19 PM
Hey Profgirl,

thanks to you i searched for some material about complex systems (didn't have a clue of what it was) and it is a pretty interesting approach i must say. In fact, it makes a lot of sense when you think of an economy, as a complex system formed by an incredible large number of agents, interacting without an evident order, and making up new properties for the entire system, different from that of the individual agents, that by the way, are heterogenous. Just fascinating!
I think this approach could lead to a new way (and i hope better) to understand and do economics. And it could complement and benefit from the theory developed so far in economics.

By the way, what material can you recommend me to start studying complex systems? And what kind of mathematics are involved?

Profgif
02-11-2008, 08:02 PM
Hi,

I am not an expert on complex systems, so my “recommendations” do not pretend to be authoritative, they may serve as a newbie's introduction at best.

For some of the concepts involved in complex systems, I recommend that you read several of the articles in Principia Cybernetica Web (http://pespmc1.vub.ac.be/). You may start by the Table of Contents (http://pespmc1.vub.ac.be/TOC.html#DEFAULT), for example reading the sections Principles of Systems and Cybernetics (http://pespmc1.vub.ac.be/CYBSPRIN.html) or Evolutionary Theory (http://pespmc1.vub.ac.be/EVOLUT.html). The material covers much more than just cybernetics and it's rather interesting, although I sometimes disagree with it. Bear in mind, that there is no scientific consensus on what complexity is or implies or how to generate it and there are several different approaches. See Complexity - Wikipedia (http://en.wikipedia.org/wiki/Complexity). You may be able to find papers discussing the different approaches if you search for papers with complexity in their titles. One of the fundamental concepts in the field is that of the Complex adaptive system (http://en.wikipedia.org/wiki/Complex_adaptive_systems), you should look for that keyword too.

With regard to economics and finance, there are two fields called agent based computational economics v.g. Agent-Based Computational Economics (Tesfatsion) (http://www.econ.iastate.edu/tesfatsi/ace.htm) and agent based computational finance v.g. Le Baron's web page (http://people.brandeis.edu/%7Eblebaron/acf/). You should be able to find material looking for those keywords. See Agent based model - Wikipedia (http://en.wikipedia.org/wiki/Agent_based_model) for an introduction. I remember to have read some promising results about the interaction of fundamentalist and chartist traders in a stock market using tools from this subject, (perhaps by Thomas Lux, but not sure). In econophysics there is a field called the “minority game” The Minority Game's web page (http://www.unifr.ch/econophysics/minority/) (but it may be too unrealistic). In general, the Santa Fe Institute is one of the leading centers in complexity and the birth place of Complexity economics (http://en.wikipedia.org/wiki/Complexity_economics). With regard to book sources, I have read for example "Reality rules" by Casti. The book contains several exaggerations and mathematical mistakes, but it also provides an ambitious approach on how mathematics could be used to model complex systems. It would be too difficult to definitively pinpoint the mathematics of complex systems. It may cover fields such as chaos theory, system dynamics, the mathematics of evolution, cybernetics, evolutionary programming, game theory or computational complexity theory.

The idea of replacing standard economics equations with simulations seems rather interesting, especially if different types of agents use heuristics rules of behavior instead of optimizing rules since the last ones may be too restrictive and unrealistic. However, this field may find a lot of resistance from traditional economists, since the field not only breaks with all their tradition, but it's very difficult to prove results using simulations, (although some insights on critical behavior and transitions can be developed). Anyway, there are huge areas of engineering and physics that nowadays mainly use simulations instead of equations, so the approach is not unheard of in science. (If one day you actually like to research these fields, one interesting strategy is to camouflage as a standard economist and after you get tenure, to start your publications. It may be too dangerous otherwise.) Debunking economics is a good start point to read more about heterodox economics Economics: Debunking Economics Overview (http://www.debunking-economics.com/) .

By the way, my alias is ProfGif, not Profgirl (which probably was the result of heuristic behavior: the real brain knows that there is no point in spending the extra second optimizing your reading of "profgif" if the letters sound and seem like profgirl, if profgirl makes more sense and if profgirl was not previously corrected by anyone. Perhaps even biologically it's more interesting for your (I assume male) brain to interact with "profgirl". Perhaps a standard economist would say that these phenomena do not concern him, that they belong to "psychology", -something at which I disagree, just ask the marketing experts who make us buy their stuff every day if these phenomena don't have major economic implications).

Profgif
02-11-2008, 08:10 PM
Oops, double posting :blush:

asquare
02-11-2008, 08:24 PM
In addition to the links Profgif posted, the U-Mich complex systems group has a very good introduction to the discipline (http://www.cscs.umich.edu/about/complexity.html). An easy-to-read paper that explains the advantages of a complexity approach is "A Complex Systems Approach to Understanding the HIV/AIDS Epidemic," by Carl Simon and James Koopman. The only way I've been able to find that article online is through Google Books, in a volume called Mathematics for Industry: Challenges and Frontiers.

filroz
02-11-2008, 08:30 PM
... However, this field may find a lot of resistance from traditional economists, since the field not only breaks with all their tradition, but it's very difficult to prove results using simulations, (although some insights on critical behavior and transitions can be developed).
Simulation can be used for example to find some equilibrium points when the derivation is difficult, or in the situation when some equilibria are unstable to determine the stable ones... Se Sargent:"Bounded Rationality in Macroeconomics", 1993


Perhaps even biologically it's more interesting for your (I assume male) brain to interact with "profgirl".
:whistle:

Oikos-nomos
02-12-2008, 11:39 AM
By the way, my alias is ProfGif, not Profgirl

Mmm...sorry about that:blush:, and thanks for the reply!