Kinematics, Calculus, and other higher maths

P

pound_sterling

I am curious if anyone has any resources, or has researched themselves, the applications of kinematics (velocity, acceleration etc..), differential calculus, or actual statistical applications to one's trading.
In terms of kinematics it does look as though the concepts in their essential principles could be successfully applied to the markets. Calculus, differential specifically, looks to be a further reach to derive any sort of useful tools; specifically identifying and defining a formula of any use.

Thoughts? Links?
 
I've done a C++ program that records all trades (actual fills not orders) at all prices. it then prints the volumes out across the prices as a normal distribution curve at the end of the week. the idea is that the market always regresses to the mean (fair value, the area of highest volume). if it's going to break out, the volumes lean towards one end of the price spectrum and the distribution sort of gets "pregnant" - it develops a bump then splits off into another distribution. this way you can play ranges and anticipate breakouts.

I've seen a money markets firm where they had an algorithm that worked out the fair value using bid/offer fluctuations but as bid/offer is becoming more and more of a mugs game I would steer clear.

I prefer a statistical approach to mechanics, as there are too many constants in mechanics, whereas people can show up in the markets at certain levels and it's not constant. statistics to my mind pays more heed to the randomness of the market.
 
trust me, none of that so-called higher math or Mechanical engineering courses that you speak of will help you beat the market.

As a matter of fact it won't even help you when you try to get a decent paying Mechanical engineering job here in the United States. Assuming you might be looking for work in that field. I've been there and done it.
 
Quote from skepticaltrader:

trust me, none of that so-called higher math or Mechanical engineering courses that you speak of will help you beat the market.

As a matter of fact it won't even help you when you try to get a decent paying Mechanical engineering job here in the United States. Assuming you might be looking for work in that field. I've been there and done it.
More...

Ditto . . . ditto . . . and ditto.
You can't mathematically calculate Market sentiment or direction. Math of ANY kind applied to the Market equates to prediction and that is a totally inconsistent proposition.
 
I hold the belief that taking an outright position is tantamount to making a prediction. When I go long 5 contracts of Feeder Cattle, I am predicting that prices will rise. Or, if you like, I'm betting that the mathematical expectation of establishing a Long position is sufficiently greater than zero to warrant taking the risk. ($win * ProbWin) >> (($lose * ProbLose) + commission + slippage)

In my system backtesting experiments, I find that calculus based estimations of price, or the first derivative of price, or the second derivative of price, don't increase trading profits. Instead, profits are maximized when the trading system focuses on the more modest goal of predicting
  • the SIGN of the first derivative of price
For this task, a meat axe does a better job than a scalpel or a laser.
 
Quote from ProfLogic:

Ditto . . . ditto . . . and ditto.
You can't mathematically calculate Market sentiment or direction. Math of ANY kind applied to the Market equates to prediction and that is a totally inconsistent proposition.
More...

I found that particularly comical. Thank you, you made my day!

Thanks to all who have responded up unto this point, your experience is invaluable.

I can definately see the merit of statistical analysis of various components of a traders function. From win/loss to trade entry it definately warrants further investigation. With that said how helpful have you found your statistical analysis of volume/price? Have you further applied statistical analysis to its win/loss ratio (so to speak)?

In regards to "mechanics", i refer to kinematics, i disagree regarding the problem with constants, i am not referring to complex equations; but rather the "simple" equations relating to velocity and acceleration. Just looking at a chart one should see the correlation between the two. The problem i am currently investigating is the proper application of these formulas to the market. I'm certain someone else must have endeavoured to apply the wisdom of the physical sciences to price action.


Lastly, in regards to calculus, the point made by Horribillicus hits upon my ambition. To identify a formula or derivative of various variables (and corresonding functions) related to trading to help build and identify trading opportunities. I am not merely looking for a holy grail but rather more intelligent tools to apply to my trading career.


To say that a mathematical investigation of the markets is futile is to state that trading it itself is futile. Regardless of your trading style you invoke mathematics at some point or another (you may be using derivatives and not even know it). So with that said, i hope this conversation continues as there's a lot of very intelligent people in this forum with a variety of expertise.
 
Statistics, certainly, but about fear/greed indications and your system metrics.

But if you think "higher math" can make sense out of what are *nearly* random processes, you are fooling yourself IMHO.

What works is trader psychology, and deriving useful trading systems from that does not take much more than HS algebra.

Good trading to all.
 
Quote from pound_sterling:

I am curious if anyone has any resources, or has researched themselves, the applications of kinematics (velocity, acceleration etc..), differential calculus, or actual statistical applications to one's trading.
In terms of kinematics it does look as though the concepts in their essential principles could be successfully applied to the markets. Calculus, differential specifically, looks to be a further reach to derive any sort of useful tools; specifically identifying and defining a formula of any use.

Thoughts? Links?
More...

If you want to use those for predicting the price, it will not work.
If you are trying to predict time of the reversal, top or bottom, you are on the right track.
Markets are symmetrical and predictable.
 
Quote from ProfLogic:

Ditto . . . ditto . . . and ditto.
You can't mathematically calculate Market sentiment or direction. Math of ANY kind applied to the Market equates to prediction and that is a totally inconsistent proposition.
More...

What about volatility? Volatility is a form of sentiment, and has predictable characteristics. And for math, how about probability and statistics?

-segv
 
Quote from segv:

What about volatility? Volatility is a form of sentiment, and has predictable characteristics. And for math, how about probability and statistics?

-segv
More...

Volatility is not a form of sentiment, it is a result of sentiment.

I never said Math didn't have predictable characteristics. I said CONSISTENT predictability.

Probability and statistics are a perfect examples. Both Math based and neither offer consistency. Both will show you a percentage that COULD occur to be accurate or happen. I remember my dual Ph.D. in Math professor laughing in class saying, "Statistics lie and liars use statistics".

I would much rather "Read the Tape" of the Market and read where price is going then TRY to predict it. Of course that is just my humble opinion.

As far as my original statement being funny . . . I imagine it was ignorantly hilarious.
 
You must log in or register to reply here.
Back
Top