B.Sc. Physics, Final year project – Stock Market Avalanche Dynamics?

Squashie

... or dynamics in general if that topic is too confined. I'm just wondering if people here think that would be a viable final year project to do in a physics degree? — or would such a proposal most likely be rejected? Considering it's not really physics.

I've always loved physics, however after 3 years of it I found myself become quite jaded and now lack the motivation for it. I'm only interested in certain areas. Have deferred my final year and since undertaking an add-on mathematics B.Sc. this year, have found myself become much more interested in other mathematical sciences.

I dread the thought of going back to physics next year and having to push through your typical quantum or spectroscopy project. If my proposal was to highlight that my stock market analysis would make use of various physics models, numerical methods, computational work, programming, could you see it being accepted?

krd

It would definitely help you get a high paying job straight out of college.

But any physics supervisor would probably punch you in the face for even suggesting it.


If you could get away with it, you could definitely walk into any bank anywhere in the world and they'd put you to work on one of their derivatives scams. If you could show you have a predictive model, that works, they might given you money to run your own thing.

But you'd definitely deserve a punch in the face for even suggesting it.

Could you do it as part of your maths Bsc?

rgunning

I certainly remember Peter Richmond in the School of Physics in TCD doing a course on something like this for our senior sophister year. It was pretty interesting, and I know he had a post-grad doing a higher degree in it too.

Depends where you are doing the BSc I guess.

Turtwig

Of course! There's entire disciplines of physics associated with that kind of stuff. Nothing is really physics, as physics can be applied to just about anything nowadays and why not?

Talk to your course coordinator and find out who does research in any area of interest to your project. Look at it this way, if you don't try then you won't get what you want. No point in doing an FYP you have no curiosity or heart in. Most faculty love it when students come up with their projects that they are keenly interested in.

nesf

It's possible that you might be left do an inter-departmental project with someone from Applied Maths or Economics. Possible. Depends hugely on the department and the people in them.

krd

Malty_T wrote: »

Of course! There's entire disciplines of physics associated with that kind of stuff. Nothing is really physics, as physics can be applied to just about anything nowadays and why not?

For God sake, please........Next you'll be saying the laws of physics are just a social construct. I can't remember who made the argument - some English professor. Scientists could understand fluid dynamics because men were solid, because of their penises, and women were fluid - because of their menstrual cycle. That's the kind of thing that provoked Sokal.

Markets are not a physical phenomena. Economics, is not physics. Psychology is not physics. Literary theory or gender studies are not physics either.

Ever since someone discovered that certain equations in thermodynamics could be used in trading stock options, there's been a craze for hiring physicists to apply their maths to the market.

And now because manufacturing has largely left for Asia. Someone with a physics degree is far more likely to end up working in finance than applying their skills to the production of industrial materials. The kind of stuff physics graduates used to do.

They even have physicists with specialities in signal processing, scouring market data looking for waves. And I think their results might be along the lines of if you look at a cloud long enough, you'll see a face. Like the economists who believe in the "business cycle" - because economic activity goes up and down - and if you squint your eyes you can see sine wave in there.

What physicists are doing with finance now days, is just as bad as what the post-modern literary theorists were doing with science a few years back.

I can't wait for some economist to pull the leg of the physics world, just like Sokal did with the literature crowd.

Some ground breaking work on the wave/particle duality of credit default swaps. I bet someone is already working on it.

rgunning

I think you are incorrectly using Sokal's experiment to denigrate a area of research you don't agree with.

The motives were completely different - Sokal was analyzing and exposing the methods with which certain thinkers in the liberal arts were attacking the foundations of scientific thinking - many with good intentions. These post-modernists were submitting their views on science without checking the validity of their scientific arguments with people who understood the equations and theories that they were loosely applying.

In other words, they didn't understand the tools that they had used, because they weren't trained to use them.

Sokal did not pull this prank because he thought their work was unimportant, on the contrary. He just wanted it done properly. He welcomed the analysis of scientific objectivity and outlined many times the reasons why he thought it was important.

The physics of the behaviour of stock markets, social response, etc. is a different story. It is a field created within the sciences; by both economists, physics, and those trained in both fields ; and peer reviewed in the same way commenurate with research in both fields.

Physics in economics, unlike the post-modernists who Sokal attacked, are using tools that they are trained well to use - they are just applying them in a different area and seeing what happens. I'm not sure of one part of your argument - surely it is better to have physicists who are trained to spot the difference between real and spurious signals observing these things, rather than have economsits who are not trained "squinting and seeing sine waves" as you put it.

And to say that a supervisor would "punch you in the face" for suggesting such a thing is just ridiculous; equally so that you would deserve it for asking a question.

krd wrote: »

For God sake, please........Next you'll be saying the laws of physics are just a social construct......

krd

rgunning wrote: »

I think you are incorrectly using Sokal's experiment to denigrate a area of research you don't agree with.

The motives were completely different - Sokal was analyzing and exposing the methods with which certain thinkers in the liberal arts were attacking the foundations of scientific thinking - many with good intentions. These post-modernists were submitting their views on science without checking the validity of their scientific arguments with people who understood the equations and theories that they were loosely applying.

I think Sokal was right in what he did. And I think he made a real cultural contribution with his paper. It's not something I want to get into explaining. But the post-modernists weren't engaging in pseudo-science or charlatanism. They were playing with text. .....Some people though, were getting ahead of themselves and believing they weren't. Like Alain Badiou, using set theory to illustrate his ideas on ontology. Badiou, is not a mathematician.

In other words, they didn't understand the tools that they had used, because they weren't trained to use them.

That would be Badiou. And probably worse, he encouraged a few generations of his graduate students to get involved in the same psuedo-maths.

Sokal did not pull this prank because he thought their work was unimportant, on the contrary. He just wanted it done properly. He welcomed the analysis of scientific objectivity and outlined many times the reasons why he thought it was important.

It helped clear the air. And it was funny. It's important to have a laugh now and again. What was equally funny was Igor and Grichka Bogdanoff doctoral thesis. Sokal was suspicious that this was the po-mo crowd getting revenge.

The physics of the behaviour of stock markets, social response, etc. is a different story. It is a field created within the sciences; by both economists, physics, and those trained in both fields ; and peer reviewed in the same way commenurate with research in both fields.

Behaviours may on a certain scale, in certain given conditions, mimic physical processes. They aren't physical process.

Where physicists have a serious edge on economists is having better maths. They have precision tools.

Physics in economics, unlike the post-modernists who Sokal attacked, are using tools that they are trained well to use - they are just applying them in a different area and seeing what happens. I'm not sure of one part of your argument - surely it is better to have physicists who are trained to spot the difference between real and spurious signals observing these things, rather than have economsits who are not trained "squinting and seeing sine waves" as you put it.

If you haven't already, read Nicholas Nasseem Taleb's Black Swan - the extended edition if you can get it. Certain processes - having all the data, will not allow you to reverse the process. For example, if you take an ice cube, and study it melting into a puddle, and you collect a vast amount of data on the process. If after that, you can take a puddle, and use your formulas and models, to construct an ice cube from it. There is something wrong with your thinking.

At the minute there is a lot of money in finance for physicists. And as long as someone is paying them, they may turn puddles into ice cubes. They could do worse - the could provide tools that claim to eliminate systemic risk, when instead they colossally magnify systemic risk.

And to say that a supervisor would "punch you in the face" for suggesting such a thing is just ridiculous; equally so that you would deserve it for asking a question.

I was joking. You didn't seriously think I was being literal, did you.

rgunning

Yeah, I don't understand why Sokal was criticized for it. I think that his motives were altruistic and light-hearted, whereas the reactionary attacks on him were just ego-related and mean-spirited.

Also, I see what you mean about the whole reversability argument and the specious reasoning it leads to. The only thing I would say to that is that, with all that ice-cube related data, you would have a pretty good idea of how the next ice-cube may melt. As long as the agents in the process are unaware that they are being observed and have no reason to change their behaviour, then the process may repeat itself.

I suspect that there is a lot more work of this type going on behind closed doors than one may think. Google and Facebook have a lot of data on how your social behaviour, the like of Paddypower - with the advent of on-line gambling - now have a lot of information on how you gamble, etc. I doubt they are just sitting on that data for the craic of it.

Do people act as agents that one can model with certain physical processes? I would say, as long as they don't know they are being observed and you have a high enough population, that it's plausible. But if so, it is a mimicry, as you say, but that doesn't mean the whole thing is useless. You can't "solve" an individual person from a group's reaction.

Ironically, the one place I don't believe it is applicable in any successful way is the stock market. The agents here know that they are being watched. All the whole thing is crooked anyway.

And I know you weren't being literal about the "punching" thing - hence the quotation marks. I just don't think that many supervisors would actually react very badly. In most cases, a BSc project is not going to contribute much to their body of research. It can be a bit of a chore coming up with these projects, so if a student comes with a pre-formed idea that they just need some analytical guidance on, they would be delighted. They would probably only refuse if they didn't feel they could help or thought it might be over their heads. I really should have made that clearer.

ILikeBananas

Squashie wrote: »

... or dynamics in general if that topic is too confined.

If you are interested in this sort of thing but you think your supervisor might balk at the idea of a financial type project then might I suggest doing it in the area of hysteresis.

Originally this phenomena was studied in Magnetism so it's definitely suitable for a Physics degree but recently there has been a lot of interest in applying the mathematics of this to other areas, including Finance.

Perhaps you could study the general theory and then use specific examples that apply to both magnetism and finance.

Edit: If this sounds like something that might interest you check out Harbir Lamba and Rod Cross. I was thinking about doing my dissertation in this exact area but I ended up applying it to an area of microbiology instead.