Do you see patterns in Random Walks?

[Antisyzygy]

Quote from MAESTRO:

Dear all, once again, could we please stick to discussing the topic of this thread which is RANDOMNESS at large. I promise we would all gain if we try to stay focused.

Of course.

The market isn't random. Its a chaotic system. I.e. it is a highly nonlinear system that has feedback loops which exist sometimes and other times seem to not exist. These sort of systems appear random, but are not.

The issue is that human beings are not always rational. They act inconsistently many times, however sometimes they act together when they get scared, or confident. Then you have groups of humans in corporations, or banks, or investment firms acting as one unit which also drive prices. The problem is that no human is the same as the next, and no human group is the same as the next, and human beings drive the market. Furthermore, not every human being has the same information at any given time, and even if they were totally rational it would be impossible to duplicate eachothers actions.

If you were to model the universe and the minds of human beings you would be able to completely predict the market's move.

The best way to look at the market is using nonlinear dynamics paired with probability and statistics. This is also known as "chaos theory". Most traders are on the "Fractal" side today, however I actually believe this is the best starting point in spite of having limited concrete evidence myself. Empirically, you can look at any timeframe in the market and if you ignore the time-stamps or make it so you can't see what timeframe you are in, you cannot tell the difference many times between 1 minute timeframe and 1 day timeframe. This is a symptom of the market being "fractal".

Take a look at Fractal dimension indicators. As the dimension approaches 1.5 the signal approaches the same measure of dimension as total randomness. As a signal approaches a fractal dimension of 1 they tend to be trending, and as a signal approaches a fractal dimension of 2 it tends to by cyclic.

Check out this link for a good basic starting point :

http://www.vanderbilt.edu/AnS/psychology/cogsci/chaos/workshop/Fractals.html

Then check out this book for theory that will help you understand the market scientifically :

http://www.amazon.com/Fractal-Geome...dp_top_cm_cr_acr_txt?ie=UTF8&showViewpoints=1

 

[Antisyzygy]

Quote from Samsara:

I think I understand Mike0085's position a bit. Long post ahead -- doing this without proofreading just to get it out in one smear.

First off, Nietzsche was far from an empiricist (nor was he an existentialist), though he shared its attack on transcendentalism. There's a certain tradition in philosophy that often begins with the Skeptics, runs through Nietzsche, then Heidegger, then basically corresponds very well with much of what people are chewing on today that crosses many boundaries of inquiry, including theoretical physics and even Eastern philosophy.

On the other hand there's the anti-realist strain in analytic tradition that fits well with formal logic, linguistics, logical positivism, and much of the functional sciences. Most Americans from the particularly "hard-nosed" sentiment fall into this tradition, and then stop at positivism, often never understanding the big problems with positivism we've since explored. You state you're a pragmatist, but I'm betting you actually fall in this camp (which is directly opposed to Pragmatism).

When I talk about things like a skepticism about the "real world" it doesn't mean that we float in god's dream and operate outside of factual existence. It's rooted in a epistemological position that there is no correspondence between language, including mathematics, and this so-called "outside world". Our knowledge operates within a network of other concepts that we have deemed true, and "truth" is not a property of the outside world that somehow grants itself to our statements. We add and refine concepts within this network, but in doing so we never get closer to some absolute truth out there. This process is what Kuhn talks about in <a href="http://en.wikipedia.org/wiki/The_Structure_of_Scientific_Revolutions"><i>The Structure of Scientific Revolutions</i></a> and is formalized in the <a href="http://plato.stanford.edu/entries/pragmatism/#PraTheTru">pragmatist theory of truth</a>.

This is what Maestro hinted at when he talked about changes in scientific knowledge over time, and about the Big Bang, and which is also why I took back the "correspondence theory" dig.

So, getting back to randomness -- my whole arc in this thread has been disagreeing that we're getting outside of Plato's cave by uncovering randomness as a "sky hook" to transcend the structural limits of our comprehension. Rather, randomness could be a functionally better fit within what we know now. You would agree with me along these lines, I think, but not with me on the Pragmatist stuff.

edit: I may be misreading what role randomness plays in this new cosmology anyway; no idea since I haven't yet read into the texts that Maestro has presented. Regardless, I think this is, maybe, a fair way to frame two commonly opposing positions when addressing a claim about whether randomness is a fundamental property of all things, or a measurement of what we don't know.

Even mathematicians and some other scientists, who use a form of language called "mathematics" to understand reality, realize its an approximation to reality. There are paradoxes that exist which do not makes much sense using our "math" language. There's even something called "Godel's Incompleteness Theorem" which states an uncountably infinite number of axioms exist in any system of mathematical logic. Axioms are things accepted as true that enable your math to work. The biggest one is the "Axiom of Choice", however many logicians and mathematicians do not accept it as truth and formulate their own alternative system.

However, the system we have made a consensus on is still the best approximation we have to understanding reality, and even being imperfect it enables us to build GPS satellites, transfer data via electromagnetic waves, and calculate things on a computer a human mind would take several lifetimes to calculate.

 

[Antisyzygy]

Quote from SrRuthenate:

Actually no highly intelligent being, including yours truly, would put up with the nonsense and boredom of completing a PhD and sitting hours on end listening and working mostly on trivial crap or teaching undergraduates basic science. Capisci?
PhD programs are the refuge of those that can't hack reality and learn from reading books (once if at all on weekends..). Why am I wasting my time with you?. Because I enjoy punching bags and crushing egos.

What were you? An art PhD? Try for a minute to realize that there is a universe out there that exists which is not totally understood, and then stretch your mind for a minute to realize that many people came before you and spent their whole lives studying it and discovering things you do not currently understand nor could you understand all of it even if you had two lifetimes. A PhD is about studying as much of it as you can. Try learning a mathematical discipline sometime and then get back to me. The methods used in indicators you take for granted in trading were essentially invented 400-4000 years ago. Imagine what people invented today that could be applied to trading? My specialty in grad school was wavelets and image processing. Many so called "traders" or "quants" implement wavelet methods without even understanding what they really are. Maybe if they tried to educate themselves in science and saw what it would do for them (even if it shows them what not to use) they wouldn't be such anti-intellectuals.

 

[Antisyzygy]

Quote from DontMissTheBus:

Fair enough - I thought the previous guy who claimed to be a in part-time program is in a finance/economics program; upon rereading, it's in math - which I suppose could be the case. I speak only for the finance/economics/insurance/etc programs.

The anti-intellectualism on this board is absurd. Moreover, it's from the crowd that could barely reason out of a paper bag (but surely, knows by intuition where the dow SHOULD be).

I am the previous post I believe you are mentioning. I am indeed a math major grad student. I work for a trading PLATFORM company which is broker independent. I mainly do coding support. However, I trade Forex myself moderately successfully. I make more than I lose, and spent quite awhile working for a private account making about 20 percent a year returns (it was a team effort, I just coded things). Anyway, the anti-intellectualism is what I wanted to address. I am sort of over-the-top sometimes when it comes to needing to express my opinion, but I get tired of coming here for information and seeing nothing but dilettantes making up lies about what it is to be a graduate student. I know what its like, because I am one. I am not trying to piss on anyone's credentials, experience, etc. I am just trying to gather information and share some in a mutually respectful environment.

 

[Antisyzygy]

Quote from SrRuthenate:

The stock indexes complex (that's right worldwide!) are heading to make a deep meander low around Sep 19 2011. Have any of the geniuses, PhD's and other ilk any idea of what I'm posting about?. If so, has anyone got the faintest idea of how low in price and deep in volume the meander will get?. How much can the FED and the ECB jointly attenuate such coming meander?.
You may raise your hand and speak or STFU and go back to your PhD studies. Pardon my bluntness, but most here are not even in the neighbourhood to even post. See you all on the 20th.

Why don't you piss off and get into a forum covering the topic you want to cover? This is a topic about seeing patterns in random walks, and it developed into more of a conversation about theory. There is no need to be a prick, dickhead.

 

[SrRuthenate]

Quote from Antisyzygy:
Why don't you piss off and get into a forum covering the topic you want to cover? This is a topic about seeing patterns in random walks, and it developed into more of a conversation about theory. There is no need to be a prick, dickhead.

I see you missed the short bus too. No need to slobber all over me, I'm Richard, they call me Dick the 13th. Must be the inches. Like that pattern?.

 

[henrik_freebsd]

It has been said before, look at the Price-Step plane and check for a normal distribution.

 

[SrRuthenate]

Quote from Antisyzygy:

I work for a trading PLATFORM company which is broker independent. I mainly do coding support.

How much to bring me a cappuccino from the italian bar?.

 

[Samsara]

Quote from Antisyzygy:

Even mathematicians and some other scientists, who use a form of language called "mathematics" to understand reality, realize its an approximation to reality. There are paradoxes that exist which do not makes much sense using our "math" language. There's even something called "Godel's Incompleteness Theorem" which states an uncountably infinite number of axioms exist in any system of mathematical logic. Axioms are things accepted as true that enable your math to work. The biggest one is the "Axiom of Choice", however many logicians and mathematicians do not accept it as truth and formulate their own alternative system.

However, the system we have made a consensus on is still the best approximation we have to understanding reality, and even being imperfect it enables us to build GPS satellites, transfer data via electromagnetic waves, and calculate things on a computer a human mind would take several lifetimes to calculate.

Certainly, no disagreement there.

 

[Samsara]

Quote from underwater:

That's where I came from too, but I'm no longer certain about that position. Have a look at MAESTRO's arguments 1 and 2 earlier in this thread. I found them to be a highly interesting read. Maybe we can continue the discussion from there.

http://www.elitetrader.com/vb/showthread.php?threadid=225224&perpage=6&pagenumber=37

I'd like to address another point. To many in this thread, random means 100% random and totally unpredictable. For example, the fact that volume clusters around the open and close of a trading session was used as an argument against randomness.

However, there could still be a random component to the volume distribution, while the distribution is obviously weighted to the open and close. That random component would make it impossible to make absolute predictions even if we had the perfect computer. We could however still make predictions with a probability attached to it.

Excellent suggestion and post, in my opinion.