Degrees of Freedom

Quote from Grob109:
I'm glad to hear that you are saddling up with another person. I'm sure he is an adept trader as well. You both will have your work cut out for you to recover from the commodities false start (which still puzzles me).

Actually, I don't think she trades at all. She got her Phd in Russia in the mid-80s in optimal control, physics, nonlinear dynamical systems, etc.

What do you mean by commodities false start? When I first started out ages ago, I went directly towards studying all sorts of "advanced" mathematical methods (way before I ever found ET, or forums) and had no understanding of the underlying traditional finance work. So, I had to double-back and learn all the traditonal economic/easy stuff, so as to get a better understanding and know what concepts I could bring forward and that were actually useful. Now I'm going circling back and unifying all the things I've learned into a rapid-prototyping framework that allows me to tie all this stuff together and see it in action and spend the minimal amount of time hand-coding as possible.

ISAGA papers and being a panelist and a thoeretical physics background were all interwoven with my trading carreeer with began right after grad school. This combo precluded my becoming entranced with applying anything but logic to the money making opportunity. Logic turned out to be the only pathway to iterative refinement of effectiveness and efficiency for making money based upon the potential that the market offers.

....snip...

Look at the last five chronological shifts in market analysis techniques and strategies. Which one is still predominent in ET? Why? What is wrong with that picture?

Do you think theoretical physics helped you with specific formulas, or as simply a way to formulate and solve problems of any kind?
 
Quote from stephencrowley:

Actually, I don't think she trades at all. She got her Phd in Russia in the mid-80s in optimal control, physics, nonlinear dynamical systems, etc.

What do you mean by commodities false start? When I first started out ages ago, I went directly towards studying all sorts of "advanced" mathematical methods (way before I ever found ET, or forums) and had no understanding of the underlying traditional finance work. So, I had to double-back and learn all the traditonal economic/easy stuff, so as to get a better understanding and know what concepts I could bring forward and that were actually useful. Now I'm going circling back and unifying all the things I've learned into a rapid-prototyping framework that allows me to tie all this stuff together and see it in action and spend the minimal amount of time hand-coding as possible.

The start you are describing above. As time passes and you look at stuff like Norbert Weiner and Richard Restak you will find out how impressive the first steps are. The mind organizes for you and not becuase you are on a critical path that you conceived of nor designed.

Everything positive sticks in the mind and unlike computers, the mind does not process the "negative" part of an expression.

Try reading ET with the understanding that nothing negative sticks.

So you have set up from the beginning a very circuitous path. This is different than taking laps to continually deepen a trust that is capable of being refined repeatedly to acheive effectiveness and efficiency. That is. stemming from a model or paradigm.

The kind of kind of baggage you speak of goes with the turf for most people.

Your coach will recognize the starter books I suggested to Thunderdog. They are both Russian authors that P.G. Bergman (Manhattan Project) gave his selected students as backups.


Do you think theoretical physics helped you with specific formulas, or as simply a way to formulate and solve problems of any kind?

We will probably be on differing wavelengths vis a vis theoretical physics. I speak from an orientation IBM made when it had 80% of the worled market. IBM assembled folks in Poghkeepsie and later Eurpoeans in Kilchberg Switzerland. We participated with leading world class physcists to train to think. IBM took very seriously there equivalent of Drink Coca Cola which was the word "THINK". I had 5 hours a week with Bergman for quite a while while working in triads as he insists. Specific formulae and "simple" ways to formulate and solve......are not what is on the table when it comes to attacking opportunities. This is not quant kind of financial industry stuff.

Nitro gave you a list of math subjects. They were unbelievable in many ways as you saw by the silence that followed.

If thinking is a good thing to be doing and be very very skilled at, what do you consider to be the thrust specifically with respect to right now.

Long ago at IBM it was maintaining and increasing an 80% share of the world market.

Do you seriously think anyone looking at the potential of the markets of the financial industry looks at anything other that the potential of those markets to transfer money into accounts ownable by any participant? How could it ever be construed as a game?

Picture a conveyor belt. Make it work optimally.

 
Quote from Grob109:
Nitro gave you a list of math subjects. They were unbelievable in many ways as you saw by the silence that followed.

Quote from nitro
Modern Algebra, Calculus, Topology, Number Theory, Theory of Computation.


These are good topics, I'm not in the habit of bothering people about specifics so I didn't inquire further.

I'm getting better at calculus, solving PDEs and soforth, mostly simply by reading lots of papers and solving stuff with maple, which is a great tool because it shows the work and helps me understand how it got from point A to point B (for some reason pencil and paper doesnt compute).

I didn't know it was called "Modern Algebra", but after doing a quick google search I've encountered most of this stuff before and am somewhat familiar with it. When I encounter notations and concepts I dont understand I just keep reading, find some references, poke around on mathworld/wikipedia...etc.. it's not hard to figure out after you learn how to use all the resources available.

Just understanding the notation/language is one thing, and then being able to have a firm grasp of the language without having to consciously think of the symbols you are seeing.. I'm improving in that area as well, it's similiar to learning any other language, soon you begin to dream the stuff and you no longer have to internally translate.

Topology is another intesting topic I've became familiar with thru the process of osmosis.. I never read any books specifically dealing with topology.. maybe I should. Do you have any recommendations?

I know nearly nothing of number theory, but it sounds interesting.

My main thrust is building up my toolset, which includes understanding, code (interface to data/order placement, easy part), analytical/computing enviornment. This is further seperated into offline analysis... exploring data that is collected online.. when models are built in the offline-environment, they must be deployed into standalone executable code . My main thrust is making all of this process is painless and easily changeable as possible and I'm quite happy with my current setup.

I've probably spent $3K on grad/post-grad level books in the past few months from many different areas of math, all of which I've gained something useful from.. so I'm no stranger to reading tought material ( probably easy material for some of you physics wizards), so I'm really looking forward to some formal study.
 
Quote from mahras2:

I would like to develop some discussion on the optimal number of degrees of freedom a system may have which would not be curve fitted. What has your experience been in playing with different degrees of freedom? How did you go about obtaining the right balance of being optimized so that it can extract profits the system is capable of but not so much that its curve fitted and will fall apart in the future?

It is directly related to your pain threshold.

John
 
Supersymmetry doesn't exist right now, no supersymmetry, no string theory. So I don't see how string theory will help to develop an useful system. Maybe I'm missing something...?
 
Quote from mahras2:

I would like to develop some discussion on the optimal number of degrees of freedom a system may have which would not be curve fitted. What has your experience been in playing with different degrees of freedom? How did you go about obtaining the right balance of being optimized so that it can extract profits the system is capable of but not so much that its curve fitted and will fall apart in the future?

Barclay Roundtable Discussion with Richard Dennis and William Eckhardt. Q&As on backtesting, optimization & curve fitting, degrees of freedom, discretionary trading, etc. Go to: http://www.tradingblox.com/forum/viewtopic.php?p=3648 and scroll down about half way to the Mon Sep 01, 2003 7:23 am post.

Eckhardt's hedge fund performance is at: http://www.iasg.com/mainframe.asp? Select the "Manager Database" tab near the top of the page on the left. Scroll down to and select Eckhardt Trading-Standard, where you'll see his equity curve since 1987.
 
Quote from BENG:

Supersymmetry doesn't exist right now, no supersymmetry, no string theory. So I don't see how string theory will help to develop an useful system. Maybe I'm missing something...?


you have physics background?
 
What does string theory have to do with trading?

Whether supersyemmtry exists or not (i dont know much about it) shouldn't matter.

Probably the most useful symmetries will be in costs functions/optimization which have nothing to do with quantum mechanics, unless you are optimizing some quantum system or something.

Quote from BENG:

Supersymmetry doesn't exist right now, no supersymmetry, no string theory. So I don't see how string theory will help to develop an useful system. Maybe I'm missing something...?
 
Quote from stephencrowley:



These are good topics, I'm not in the habit of bothering people about specifics so I didn't inquire further.

I'm getting better at calculus, solving PDEs and soforth, mostly simply by reading lots of papers and solving stuff with maple, which is a great tool because it shows the work and helps me understand how it got from point A to point B (for some reason pencil and paper doesnt compute).

I didn't know it was called "Modern Algebra", but after doing a quick google search I've encountered most of this stuff before and am somewhat familiar with it. When I encounter notations and concepts I dont understand I just keep reading, find some references, poke around on mathworld/wikipedia...etc.. it's not hard to figure out after you learn how to use all the resources available.

Just understanding the notation/language is one thing, and then being able to have a firm grasp of the language without having to consciously think of the symbols you are seeing.. I'm improving in that area as well, it's similiar to learning any other language, soon you begin to dream the stuff and you no longer have to internally translate.

Topology is another intesting topic I've became familiar with thru the process of osmosis.. I never read any books specifically dealing with topology.. maybe I should. Do you have any recommendations?

I know nearly nothing of number theory, but it sounds interesting.

My main thrust is building up my toolset, which includes understanding, code (interface to data/order placement, easy part), analytical/computing enviornment. This is further seperated into offline analysis... exploring data that is collected online.. when models are built in the offline-environment, they must be deployed into standalone executable code . My main thrust is making all of this process is painless and easily changeable as possible and I'm quite happy with my current setup.

I've probably spent $3K on grad/post-grad level books in the past few months from many different areas of math, all of which I've gained something useful from.. so I'm no stranger to reading tought material ( probably easy material for some of you physics wizards), so I'm really looking forward to some formal study.
[/QUOTE]

Volume is not a leading indicator of knowledge.

Others are commenting on symmetry in an advanced context.

Lets say, bear with me, that trading is a use of time that has some value.

Most people will agree that the value is determined by the connection of the trader to the market and the degree to which he participates.

I feel that this is the kernal that anyone can begin with and then advance themselves in the trading world.

What does this kernal of knowledge lead to? Different things for different people.

If a person chooses the route to extract the potential of the market as effectively and efficiently as possible, then he has a set of tasks that are fairly clean and which colud even get to be elegant.

Of of the keys to this effort is to make discoveries as you go along.

When my kids were going through the forms of scholastic prep every other room of their home was floor to ceiling in shelving of books according to the fields of study the books were about. they were discoverers all the while.

Books were always being made as well.

Ploughing and growing and feeding and breeding were daily chores and they lived where there were no keys for doors because there were no keyholes.

Discoveries can be made best in the absence of noise and confusion.

What tools are needed to discover how things work? My kids loved screwdrivers and cooking stuff we grew. Lisa colud hardly drag her tool box around. John cleaned the Mercedes engine weekly. She adjusted her VW valves about every month I think. John played golf about two to three inches deeper than anyone else and hs goal was to be on the green off the tee no matter what; Lisa played the same strokes where birdies were good enough; skiiing for both was straight down.

The fun in everything is knowing how it works and how to work it.

John and his cousin would never play monopoly without calculators. When they ran companies for me they used calculators too. Now, they have their own assortment.

What are you going to discover about the market? It IS going to be based upon your orientation to discovering things.

The market is unique and, it turns out symmetric in terms of variable performance. See Scoring.

The money is there being offered. You want to make money. You NEED to know how it works.

The two themes, making money and how the market works, go hand in glove. The rules for trading are set by the operators of the markets. That is, there are accounts and agreements on the accounts and a set of rules for making orders and keeping money in your account.

Its discovery time for you, I think.

Get that money out of the market into your account(s).
 
Back
Top