Topic: Elliott Wave Analysis - page 3. (Read 20044 times)

Re: fibonacci: I went on a serendipitous scavenger hunt through your references and somehow ended upon learning a bit of Haskell. I think it was something about simplified code beautification. My Latin is a bit rusty and didn't find much meat in English. I lost your train of thought from Fib to Gradus to Retracement.

I intend to go through my own charts and do some statistics on the previous fourth wave retracement; Wave 2 and 4 alternation; and longest of waves 3 and 5 (but my charts are already influenced by what I believe to be statistical likely). There are innumerable statistics to be obtained and it would be wonderful to remove the subjectivity. Swannell had many of these statistics but he went off to Fiji and died on us this summer. I've seen several pages of his results, but none of the raw data.

Anyway, yes, It's possible and would be very interesting to write an Elliott calculator. I can already imagine how it might work, though I'm sure there are many demons in the details. It's no simple weekend project.

You may note that Swannell added many new rules (beyond Elliott's three) to restrict the number of possible counts. I already think there's much too much variability in the corrective wave types. I'd rather say there are x impulsive types and y corrective types and when none or multiple fit, you'd have to say for example, it's 57% type a and 58% type b and only 12% type c. Whether the result would still be truly Elliott, I don't know, and perhaps I don't care, as long as it's predictive.

Re: Weierstrass: Yup, this is why I had to add some pathological parenthetical tap-dancing around my statement that sine waves were 'never' self-similar. But where are you going with that?

I have been thinking if it were possible to develop a program which parses the mtgox trade api data and enumerates the possible wave counts so that we see which possibilities are there. You mentioned that if taken that effort Elliot might not be the best choice, but there is something similar with sine functions which also forms a 1-dimensional fractal curve.

http://en.wikipedia.org/wiki/Weierstrass_function

PS: Any thoughts on my suggestion on Fibonacci Retracements?

Please challenge, ask questions, spit on the floor. Otherwise, I'm just talking to myself.

  Heya, just popping in to let you know I'm still lurking here. =)  Thank you for continuing this and explaining what you are doing. It is also nice to see someone else joined in that actually understands some of this stuff.

   Cheers

Today's decisive spike up just shy of $3, invalidated the 30 October count upon crossing above $2.8 (the grey bar). This leaves the "Swannell wave" as the most likely count.

Though ugly, it has a few advantages in retrospect. Cyan 2 moves perfectly into the price territory of the previous black wave IV (cyan bar $2.8-$3.45). It validated my extended green count and didn't have the 'squeeze' problem (three waves between $2.6 and $2. of the 30 October count. Interestingly enough, the Swannell count is the only strictly valid count on the magenta hourly scale that I've drawn since 2 November. Perhaps there's a lesson there?

No that's not me
But that's where I heard about that fact which inspired me to buy "the book of numbers" where this initially from. The guy behind this may bright but has wrong concepts about physics imo. He attributed particular importance to the number 10 and it in a circle-jerk about "quantum numerology". I view numerology as a misguided esoteric concept. (More on that later)


I haven't yet figured out if there is something as total entropy. It is said to be the golden ratio but according to Eulers theory this would hold true for any irrational number. There may be some merit to it since Phi is the most efficient way to arrange something like a sunflower daisy.



The other way around and the one is also included for some reason (In the past it was considered prime)

6 --> 1 + 2*1 + 3*1 = 6
7 --> 1 + 7*1 = 8
8 --> 1 + 2*3= 7

Is called the "Tentamen novae theoriae musicae" and is of the rare cases google finds viable information because of the weird latin name
The number used as the input is the least common denominator of all integers involved in a system. I think work very well to be used as a generalization of the so called Fibonacci Retracements.

So while it would include the commonly used ratios it would also introduce new ones and we would be able to weight them according to their gradus function value.

The only exception may be the ones including the prime factor of 5 because of the trend to have resistance levels around "whole" fractionals like 2, 2.5 2.25 and so on. In our culture we happen to use a basis of 10 number system which might suit us because we have 10 fingers but any other number system works the same and there is no particular unique property about it I know of.
The mayas used 20 maybe they were also counting their toes.  

The whole concept of numerology is based upon the numbers 10 and 12 which have no unique property whatsoever form a purly mathematical standpoint if used as a number system. Btw the number system with the best entropy is trinary which is the best approximation of the idealistic (but unobtainable) value of e. It is for example most efficient to construct a multiple choice phone menu with a tree of 3 choices.

Did you write this? It's an absolutely beautiful collection: http://www.vortexmath.com/a-fibonacci-phyllotaxis-prime-number-sieve



We experience the resonance as beautiful in a minimal resonant system? Some perfect harmony between absolute order and total entropy, maybe?



Do you have a reference to this specifically? What does this Gradus Function tell us? Are these correct?

6 --> (2+1) x (3+1) = 12
7 --> (7+1) = 8
8 --> (2+3) = 5

Ok but where?

Alright I'll try...

Fibonacci Numbers and the golden Ratio occur in Nature although not as often as generally believed. The phyllotaxis is one of the few examples where the Sequence occurs in its actual form. The floret show arithmetic progressions with Fibonacci differences. They are also a particular efficient natural prime number sieve.

Prime numbers occur in Nature every time a periodic system must minimize resonance effects. We experience these resonance effects as aesthetic the arising complexity of combinations as beauty.

There is a way to calculate the aesthetics of a particular ratio, this was used by Euler to describe the aesthetics of musical intervals as a generalization. But it holds also true for other cases were ratios occur.
The formula used describes a "Gradus Function" which attributes a rating to any integer. This rating is calculated by summing the product of its prime factors and their exponent.

Implementing said function in a memory efficient way is related the PI function, the number of prime numbers below a certain limit. This function can only exactly formulated once there is an answer to the Riemann Hypothesis and only by using an assumption of truth and precomputed values can it be improved.

This is my 1234th post  

Please, get started...

I never understood the whole Fibonacci numerology cult. No...
As a somewhat hobby number theorist I have to say the following:

The aesthetic ratios relate to prime numbers and that the ratios contained in the Fibonacci sequence are aesthetic is a mere coincidence. (or not depending on how you put it.) What is important is they are only a substet of particular aesthetic ratios.
I traveled this road, all the way over the Phyllotaxis Spiral to the Riemann Hypothesis.  

We are dealing with only a miniscule part of the matrix, don't get me started

Did you know you can enumerate the elements of the old periodic Table perfectly to the basic audioactive decay elements of the look-and-say sequence? Nothing less than a conspiracy theory hidden in the fabric of math  

Sine waves are never (not necessarily and most often not) fractal patterns. A sine wave does not look like the same sine wave on all scales (for example all octaves simultaneously below, through, and above the range of hearing). A perfect sine wave is repeating by definition, but not self-similar on any scale but one.



There's some fun Fibonacci math: 3 corrective waves + 5 impulsive waves = 8 waves. The three corrective waves consist of 5+3+5 = 13 subwaves. The five impulsive waves consist of 5+3+5+3+5 = 21 subwaves. Together that's 34 subwaves. At a higher order, five + three + five = 21 + 13 + 21 = 55 corrective sub-subwaves preceded by five + three + five + three + five = 89 impulsive sub-subwaves for a grand total of 144 sub-subwaves. You can continue this math-erbation for higher scales and extend the sequence below:

0 , 1 , 1 , 2 , 3 , 5 , 8 , 13 , 21 , 34 , 55 , 89 , 144 , 233 , 377 , 610 , 987 , 1597 , 2584 , 4181 , 6765 , 10946 , 17711 , 28657 , 46368...

But we should not be too surprised by the results of the application of the Fibonacci definition. Fibonacci as it relates to Elliott, as I understand Pretcher, is the ratio between any two elements of the sequence, not the sequence itself. In other words, the Golden ratio: (a+b)/a = a/b = 1:.61803...

2 / 3 = .66667...
21 / 34 = .61765...
233 / 377 = .61803...
2584 / 4181 = .61803...
28657 / 46368 = .61803...

If an impulsive wave is 89 sub-subwaves, should we expect a 55 sub-subwave correction to retrace 1-.618 = 38.2% ? Well sure, except when it's 61.8% or 50% or 14.6% or 23.6% or 12.3456789% or... And that's the problem I have with Fib and prices. Wave price movements, in my experience, have only slightly better than no relationship to the golden ratio.

BUT the golden ratio is an aesthetic. When I draw counts I place future counts where they 'look good' within various price suggestions and Elliott rules. 1/3, 2/3, and 3/2, tend to be more aesthetically pleasing than 1/2, and 2. I don't know why, but the market appreciates beauty.

Not good enough, since it is possible to approximate any signal with any periodic oscillation. Sine Waves, sawtooth, triangle, square waves, wavelets, curvelets, etc..


Now we are talking, yes if you consider the Fibonacci Series 3, 5 are a unique pair which can be enumerated as this impulse wave movement.
However there is also 2,3 and 1,2 which has the same property plus there is 1,1 which has the inverse property of them all.

Of course 3,5 would be the most complex and useful and the other ones are no doubt a subset but that shouldn't mean they couldn't serve as a primary wave. (As of carrier or first order or whatever it is called by Elliott)


Thanks, let me guess the site requires a credit card, ironically...

A failed fifth wave often resembles a fourth wave triangle. A failed fourth wave (1,2,3,4) resembles a corrective reversal (a,b,c,1). A combined corrective zig-zag resembles an impulsive wave somewhat. There are many such seemingly ambiguous situations, however the previous waves indicate the present wave, which all fit within a model of a higher scale.

I've had incredible confidence regarding the correction during the last two weeks of October, but have been puzzled by the sideways correction early November, which is making it difficult to confidently count the present situation. This muddy count keeps me on the lookout for a reversal, impulsive failures, and triangles. I am very confident bitcoin is at or approaching the bottom as there are conflicting trends and tension on all scales.

I really don't know why impulses are five waves rather than say seven waves. However the question is a bit backwards (http://en.wikipedia.org/wiki/Anthropic_principle). As with any fractal, if the pattern were not self-similar on all scales, there would be no pattern at all.

S3052, does Pretcher argue convincingly for the Fibonacci series? I find 2/3 and 3/2 reasonable approximates, but unreliable. Corrections to the previous fourth wave however has been incredibly predictive.

Some years ago I was similarily surprised like you about the predictive power of Elliott Waves.

A great book that uncovered the fundamentals behind it is Elliott Wave Principle
Key to Market Behavior http://www.elliottwave.com/books/ewp/default.aspx?code=ecg

This will probably answer some of your questions (I have no affiliation with ElliottWave.com or Robert Prechter)

It seems your prediction might turn out to be true after all, there seems to be some pump going on @2.46.
Whoever is doing this is still jumpy but at least he is careful enough to be safe from the shorts.

Thanks for the warning if it really gets there

Elliot seems to have predictive power where other mechanisms fail to indicate a specific direction (I use Aroon, as discussed) which was inconclusive during this day. I wonder if this is just random chance or a self fulfilling prophecy though, since I still don't get why market movements should always follow the 3 up 2 down wave pattern. It seems kind of arbitrary to me.

I understand that there is a trend for repeating corrective series in any market, but why 5 waves?

Thanks S3052. I think there is more exposure here and even dismissive/troll-y posts force me to reconsider alternative counts, which is only good. But I'll move immediately if this forum drops embedded images.



Continuing with the preferred 30 October count (extended with new data above), our little rally (black wave 4, implied but not drawn on the chart) must not breach $2.8 (the real bottom of 1.v). $2.6-ish is a more likely peak. If the fifth wave fails to break below $2, then we begin a slow but grand yearly blue wave III toward new heights.

My interpretation since 14 November is a corrective fourth wave extended/expanded flat. We are now in the c and final wave, which should exhibit five sub-waves, were hourly wave counts reliable.



Though I'm betting against it, if this rally peaks in the $2.8-$3.45, it would retro-actively suggest the "Swannell wave" count I shared last week. It would imply that the previous $3 sideways correction were a fourth wave, the drop a fifth wave, and our current rally a correction of a higher degree. Anything above $3.45 (before testing and possibly breaking below $2) invalidates all of my counts for the past month or two (reversal, month long triangle, downward impulsive failure).

The chart above is the "30 October count" applied to November data (with new font colors). The pattern since cyan wave 3, $2.1, 15 Nov, looks like an expanding flat correction. This past week's black wave C should consist of five impulsive waves up, however it is not possible to count even three valid magenta waves. If we strictly consider valid wave counts (and assume the cyan count is thus far correct) then bitcoin is in a tight squeeze. Cyan wave 4 must not cross into the price territory of cyan wave 2 above $2.8. Green 3 can not be labeled magenta iii because the following wave (green 4) would cross under magenta i. So, we count magenta wave iii extended with five impulsive green waves, leaving enough room for magenta iv to drop no less than wave i, $2.5, neatly in the suggested price territory of green 4. Bitcoin has never been so neat and tidy on tiny scales. I'm willing to reject magenta spike i (likewise, the spike down to $2 between A and B), throw out the green extension, consider green 3 to be instead magenta iii, green 4 to be magenta iv and prepare for a cyan 4 peak (about now) in the grey band $2.6-2.8.

Netrin: you are doing great with Elliott Wave analysis. Have you thought about a thread here: http://www.bitcoinbullbear.com/forum.html#/news/

Well, I'm not even the OP. I hijacked this thread from aptly named Mr. BitcoinBull.


There may be billions of ways to count waves, but not all counts are equal. I pay no heed to a chart which is blind to the April-June correction. Just as accurately, subjectively and superficially as facial recognition, I recognize April-June as the fourth wave of the yearly impulse.

I don't 'want' bitcoin to reach $0.6-1.1; That price is suggested by the fourth wave. That count and its sub-waves have proved to be quite predictive. Counts on numerous scales form a lattice which reinforce each other and subsequent counts. My counts are generally built from six degrees. I analyze hourly charts (green and magenta) in order to execute daily trades (black and cyan) which build a reliable case for monthly scale counts (red and blue).

Of course, but this is Economy, not the Speculation subforum.

All I say is that there are (at least) 2 different ways to draw the waves, and each one corresponds to a particular economic development. And I think in case of bitcoin Elliott describes this choice. We are facing a crossroad which will determine the future role of bitcoin.
I hope that your prediction turns out to be true, bitcoin would be far more useful if prices would recover quickly. But what someone wants and what could happen are 2 different things.

We can continue our debate next year when we should have enough data to draw a preliminary conclusion.