Can "Noise" Be Traded?

Quote from nononsense:

An ET member PM'ed me to express his bewildrement about the different definitions advanced for the term "noise". This is an often encountered situation on these boards where many participants jump on the bandwagon juggling with stuff like "random, noise, mean, etc" while demonstrating that they don't have the least idea of what they are talking about.

Any textbook on probability theory and on signal theory will give a cogent definition of these terms.
More...


Cutten is correct.

He just took the time to point out two very different scenarios.

Cutten is able to recognize, practically, how noise affects making money.

The mean being present, for Cutten, signifies that the range of values during this condition is such that nothing else is going on and therefore the noise evaluation comes out as he syas.

At all other market times their is a dominance of other market factors and noise, while there, has no statistically significant expression. An example would be important short lived news, a time when the diversity of "reaction" is great and also affects a lot of "cancellation" of unfiltered signals.


The later could be handled in a "carrier" type discussion as well.

This example you are commenting upon from cutten is what serves to differentiate the kind of experiences that p[eople have had. I believe Cutten has been in the trenches and that you have a big library where you read a lot.

I like cutten's views because markets are empirical rayher than theoretical. Noise is a little empirical as well.
 
Quote from nononsense:

An ET member PM'ed me to express his bewildrement about the different definitions advanced for the term "noise". This is an often encountered situation on these boards where many participants jump on the bandwagon juggling with stuff like "random, noise, mean, etc" while demonstrating that they don't have the least idea of what they are talking about.

Any textbook on probability theory and on signal theory will give a cogent definition of these terms.
More...
This doesn't help. Price movement is not a communication signal.

Thunderdog got it right: "Noise" can be a very relative term.

I would state that it is a relative term.

Market noise is neither additive nor multiplicative nor "random". It is systemic. So my definition of market noise is practical: it is that portion of price movement that is untradable. By contrast, signal is that portion of price movement that is tradable.

So the answer to the thread is "no"... imo. YMMV.
 
Quote from abogdan:

The noise exists on any level. If you trade weekly charts you just simply deal with different kind of frequency and amplitude of the noise compared to, let say, 5 min chart. However, the principles of dealing with this noise will remain relatively the same. There are two major principles of treating the noise on any level:

1. Filter it (using, moving averages, FIR filters, etc.)

2. Use it to trade (employing any counter trend technique like bands, oscillators, RSI etc. that work well on detrended signal)

Those are basics, obviously. But recognizing the noise is a very fruitful exercise for anybody whether you use it or filter it. The question becomes then what signal gives you more $$$ per trade. Simple observation could be also made on that respect:

The sum of all the squiggles is always greater then a straight line connecting the ends
More...

On a practical note: let's say i use bollinger bands to measure the amplitute of the "noise". I would use lagging data to base my decisions on. i don't really like that. I wouldn't know how far back should i look in the past? I just really don't like it.. that's the drawdown of systematic trading versus discretionary trading, but maybe there's a way around.

besides averaging down :D , any other way around this, that someone knows?
 
abogdan has it right! The sum of all the squiggles is greater then a straight line so have systems that create profits from the noise------- use it as a means to profits.
 
Quote from abogdan:

Noise or imperfections are the only things that are tradable! If the markets were completely 100% efficient there would be no patterns then. The only reason we can trade is because of the noise!
More...

I've always considered the term "noise" as merely a subjective label for "incomprehensible" market movement at a given time frame for a given trader. But I think I'm also coming close to understanding the concept of noise from the above angle.

Let's assume the ideal model of movement for a near-perfectly "efficient" market = occasional gaps as new information is instantly priced in, followed by seemingly random fluctuations -- aka "noise". Well, what accounts for those fluctuations of after-noise but the net imperfections of the market participants? The more imperfect the market (in dissemination of information, quality of analysis, etc), the "noisier" the price action as information is discounted. Now, what happens if the market participants begin to discount the imperfections of the market itself, to the point where the effect on price is greater than that of the initial inflection point of new information? As more and more players are added to the efficient market model and attempt to get in on the same "discounting race", wouldn't it eventually become more and more profitable to concentrate on playing the noise rather than attempt to discount raw information itself, to the point where we come to the above conclusion? Just thinking out loud.
 
It is important to define terms, otherwise we will go around in circles forever.

1) Mathematical Noise by definition means it is indeterminate. Therefore, no edge can exist, even for God.

2) What you are talking about is the definition of Chaos, seemingly random behavior that in fact has dynamics.

If what the original poster asked is the Mathematical definition of noise or 1) above, the answer is no, even for God.

If what the poster is asking is about 2), then the answer is yes, but you have to be really clever to find the signal.

What you are saying is that there is no pure versions of 1) in the markets. I doubt it, but it may be true.

nitro
Quote from illiquid:

I've always considered the term "noise" as merely a subjective label for "incomprehensible" market movement at a given time frame for a given trader. But I think I'm also coming close to understanding the concept of noise from the above angle.

Let's assume the ideal model of movement for a near-perfectly "efficient" market = occasional gaps as new information is instantly priced in, followed by seemingly random fluctuations -- aka "noise". Well, what accounts for those fluctuations of after-noise but the net imperfections of the market participants? The more imperfect the market (in dissemination of information, quality of analysis, etc), the "noisier" the price action as information is discounted. Now, what happens if the market participants begin to discount the imperfections of the market itself, to the point where the effect on price is greater than that of the initial inflection point of new information? As more and more players are added to the efficient market model and attempt to get in on the same "discounting race", wouldn't it eventually become more and more profitable to concentrate on playing the noise rather than attempt to discount raw information itself, to the point where we come to the above conclusion? Just thinking out loud.
More...
 
Quote from nitro:

It is important to define terms, otherwise we will go around in circles forever.

1) Mathematical Noise by definition means it is indeterminate. Therefore, no edge can exist, even for God.

2) What you are talking about is the definition of Chaos, seemingly random behavior that in fact has dynamics.

If what the original poster asked is the Mathematical definition of noise or 1) above, the answer is no, even for God.

If what the poster is asking is about 2), then the answer is yes, but you have to be really clever to find the signal.

What you are saying is that there is no pure versions of 1) in the markets. I doubt it, but it may be true.

nitro
More...
Are you talking about Mandelbrot's contentions?
 
Quote from duard:

Are you talking about Mandelbrot's contentions?
More...
Not specifically, although there is a connection between fractals and chaos.

However, the markets can be chaotic without being fractal.

nitro
 
Abogdan

.........Unfortunately, there is only a handful of traders who understand this in depth. Oh, well ....

FORTUNATELY
 
You must log in or register to reply here.
Back
Top