a trading strategy based on Elliott Wave Theory - page 235
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
However, I would be more modest - the proposed method is the most effective on the lower timeframes and ineffective on the larger ones.
It's good that you're such fun connoisseurs of the union of the word 'beer' and the word 'drink'. :)
I'm more of a connoisseur of the union of intellectual powers. And I'm not too shy about beer.
And I was shocked because statistics show that the two words apart
are hardly ever used separately. How many sigmas is the probability of such an event? :-))
But I hope you won't attribute that to your post.
You came up with the topic, so you had fun with it.
Yes, even earlier than that.
This just confirms my long-standing belief (now proven scientifically :) that the breakdown of a data stream into equidistant intervals has neither justification nor practical value. Just a result of the purely psychological attitudes of the West.
But splitting into the same equidistant intervals, but by price, has its rationale. That's why kagi and renko have been around for about 2 centuries, and traders are guided by support and resistance levels rather than time of day. Although news, for example, is released at the same time, but even this "main" market driving factor does not create a time cycle. The market has its own timing.
... paid attention to the fact that from TF=60 min the correlation between adjacent readings is almost zero (see your fig.), i.e. the Markov process, starting from that moment, is born for EURUSD into a Wiener process. Isn't it interesting? People are working on the "days", but there is nothing to do here, starting from one hour!
It is an interesting point. On the one hand, it is true. On the other hand,
you could put forward a theory that only the big movements matter,
and the small ones represent noise. And the large movements,
are of a random nature and the small ones are not, not random.
True, but only for the very first differences. And FAC in the end is the sum of these differences, and this is exactly what is very important. From the first count the process develops, forming deterministic chaos in some places. By "process evolves" I mean the model that you originally laid out: the subsequent count depends (precisely depends) on the previous one.
Contrast this process with noise, for which I gave calculations and for which everything you wrote earlier is true. I.e. a complete lack of memory, which is what my algorithm confirmed. The values of the next reference are completely random, unlike the proposed model for calculating a Wiener process. After all, here also the difference between counts is completely random - and as a consequence, zero memory.
Sergey, maybe this is not a pure Wiener process model. Are there any other variants of modeling a Wiener process?
PS: I think I'll be able to make a more convincing argument soon after "in-depth study" of Chaos :o)
What's all this pathetics? Do you think this is a better way to present the results
Sergey, it was just a joke. No pathetics, just picking up your vocabulary :o))))
to Yurixx
I'm just copying a graphical object to the buffer. Then I open any image editor and copy the picture from the buffer. It's very simple and not long at all. No screenshots.
to North Wind
It's good that you're such fun connoisseurs of the union of the word 'beer' and the word 'drink'. :)
Not only am I a fun connoisseur, but I'm also a fun practitioner, unlike some quiet theorists :o))))! "Hougarden, for example, is my favourite variety, although, no way to discuss it....
... paid attention to the fact that from TF=60 min the correlation between neighbouring samples is almost zero (see your fig.), i.e. the Markov process, starting from that moment, degenerates for EURUSD into a Wiener one. Isn't it interesting? People are working on the "days", but there is nothing to do here, starting from one hour!
That's an interesting point. On the one hand, it is. On the other,
one could put forward the theory that only large movements matter,
and the small ones represent noise. And the large movements,
are of a random nature and the small ones are not, not random.
That may be true, but I take the exact opposite view. The big moves reflect just the "Jedi power" as I jokingly call it. One only has to follow that power and understand its nature.
What happens, we'll see... :o)))
Sergei, you... take it easy :) Mathematics is a hard science, after all.
For example, if we consider a "pure" random "noise" process, as you presented it and as I understand it: X[i]=sigma[i], where sigma is a normally distributed random variable with zero expectation, then FAC constructed for first differences, DOES NOT identically equal zero!
1. This is easy to prove mathematically rigorously, and it contradicts what you claim (see above).
2. This fact is not an indication of a hidden relationship between the first differences of a noise series (because you probably would immediately want to create a revolutionary "New Chaos" theory exploiting this property of a "pure noise series"), but only an unavoidable consequence of the mathematical procedure of taking these differences.
3. you have to be careful, critical and understand what you are doing.
4. The Wiener process has had a mathematical definition for a long time and there is no need to invent a new one, because it will not give you anything new - you will only lose time.
To Northwind 28.01.07 22:55
one could put forward the theory that only large movements matter,
and the small ones represent noise. And the large movements,
are of a random nature, while the small ones are not.
First of all, it is not a theory but a hypothesis. The theory will come later, after it has been built and confirmed by practice. Secondly, did you yourself understand what you wanted to say? "...large movements matter... large movements, have a random nature...", "...small movements represent noise... the small ones... are not random." Well, either random and irrelevant or not random and relevant.
It seems to me...
... paid attention to the fact that from TF=60 min the correlation between adjacent readings is almost zero (see your fig.), i.e. the Markov process, starting from that moment, is born for EURUSD into a Wiener process. Isn't it interesting? People are working on the "days", but there is nothing to do here, starting from one hour!
IMHO, from a purely mathematical point of view, the "Cascade of Bifurcations" describes the pricing model most completely.
A bifurcation cascade (Feigenbaum sequence or the period-doubling scenario) is one of the typical scenarios of transition from order to chaos, from a simple periodic mode to a complex aperiodic mode with an infinite period doubling. The Feigenbaum sequence has a self-similar, fractal structure - increasing any region reveals similarity of a selected region to the entire structure.
Analysis of transition mechanisms from order to chaos in real systems and in various models has revealed the universality of relatively few scenarios of transition to chaos. The transition to chaos can be represented as a bifurcation diagram (the term "bifurcation" is used to denote the qualitative restructuring of the system and the emergence of a new regime of its behaviour). The emergence of the system into an unpredictable regime is described by a cascade of bifurcations, one after another. The cascade of bifurcations leads sequentially to a choice between two solutions, then four, etc.; the system starts oscillating in a chaotic, turbulent regime of successive doubling (number?) of possible values.
All TFs are therefore meaningful and can be used for trading. The important thing is to use meaningful inputs in the right amount and not go out of your script, or they double up too quickly :). IMHO
That Northern Wind 28.01.07 22:55
one might theorize that only the large movements matter,
and the small ones represent noise. And the large movements,
have a random nature, while the small ones do not, are not random.
Firstly, not a theory, but a hypothesis is put forward. The theory will come later, after it has been constructed and confirmed by practice.
Well, if you fall into demonic formalism, then yes, theory and hypothesis are different things, for
for some people. At the level of housewives there is no difference.
Secondly, did you yourself understand what you wanted to say? "...large movements matter... large movements, have a random nature...", "...small ones represent noise... the small ones... are not random." Well, either random and irrelevant or not random and relevant.
It seems to me...
Always perfectly understand, what I speak about. And there is no contradiction here.
Something similar is described by Malderbrot. In the simplest case, if you take
one random process and overlay it with another random process, but
on a smaller scale, both in magnitude and in counting, you get
roughly what I'm talking about. In this system, with two independent
processes, the smaller scale process just has to follow
the larger process and that makes it less random.
On the spider, a well-known breeding ground for sentiment ideas, there is much and passionate discussion of
that very sentiment. From my point of view, large movements
is sentimento. But it must be said that the Russian forex market does not
like a real market, if only because the quotes are not formed
and the sentimento is different.
That may be so, but I take the exact opposite view. Big moves reflect just the "Jedi power" as I jokingly call it. One just has to follow that power and understand its nature.
Well, yes, the "Jedi Force" is the big moves, but the vector of movement of this force,
it's incidental to outsiders. Accidental in the sense that the laws of its movement
and it's unlikely to be.
For example, if we consider a "pure" random process "noise" as you presented it and as I understand it: X[i]=sigma[i], where sigma is a normally distributed random variable with zero expectation, then FAC built for first differences WILL NOT identically equal to zero!
1. This is easy to prove mathematically rigorously, and it contradicts what you claim (see above).
2. This fact is not an indication of a hidden relationship between the first differences of a noise series (because you probably would immediately want to create a revolutionary "New Chaos" theory exploiting this property of a "pure noise series"), but only an unavoidable consequence of the mathematical procedure of taking these differences.
3. You have to be careful, critical and understand what you are doing.
4. The Wiener process has already had a mathematical definition for a long time and there is no need to invent a new one, because it will not give you anything new - you will only lose time.
A Wiener process is one of the varieties of random processes. It, as well as noise, must satisfy a number of properties, including (not all are listed):
(a) zero mathematical expectation.
(b) the "gain" must be a random variable
If you look closely at my graph, you will notice the following:
(1) It is not FAC in pure form, but its modification for my task based on autocorrelation, what I tirelessly reminded earlier
(2) FAC and is not equal to zero for noise, and its value varies on average from 0.1 to 0.8 (all seen on graphs)
I equate it to zero in words, using criteria (roughly enough) - 0.1 or 0.8 is much less than 1.9. Anything less than this value is considered zero, i.e. the strength of the connection between the counts is zero. (I remind you of the purpose of my criterion and the statement itself)
I also found out that the Wiener process is not modelled in this way. This is a very crude approximation. And if we approach formally, the model you propose is not a Wiener process at all.
So, Sergey, I have all my love for mathematics. And as for creating a new theory, I already gave, as it seems to me, good advice - don't limit Nature and, therefore, your consciousness.
PS: Just don't ask me what I am expanding my consciousness with. :о) Ok, I'll tell you... beer! :о)
If the discussion about Pastukhov's thesis has come to an end, then maybe it is time to sum up?
Because fellow traders are wondering what to do now. :-))
And if this result is positive, then maybe we can move on to the next stage - to discuss a practical strategy?
And if it is negative, does it mean the final verdict ?