54 replies to this topic

[thakid]

Where are you getting the idea for the 8 year gold cylce (i'm not saying there isn't)? Is this something you have come up with yourself...using Hursts original analysis techniques...or are you merely quoting authorities? I'm just trying to better understand the situation here no offense intended in any way ( :

[SevenOfEleven]

Where are you getting the idea for the 8 year gold cylce (i'm not saying there isn't)?

Initially, after reading some Stan Harley work on gold, I was asking someone a question regarding historical charts and gold and I happened upon it as an answer to my own question.

I then noticed that sharlynx was tracking it. At first it appeared (to me anyway) in promotion of the 'gold bull' a couple of years ago or so. Recently it was updated by someone as having been broken (the gist was - 'we are in a longer term bull than the normal 3 years up rally on the 8 year cycle). I posted a response to this over at capitalstool.com (in the Stools Gold forum) as that notion being incorrect based on my own work.

I then cross posted one of the posts from that thread over here at Traders Talk:

The 8 Year Itch, Be home by ate...

and to a couple of other forums. My reason for cross posting was that I have recently been seeing what appears to by a cycle inversion on the shorter POG cycles and that there was a 'sell' on the 8year (read: long cycle washing out shorter cycles).

Is this something you have come up with yourself...using Hursts original analysis techniques...or are you merely quoting authorities?

I use my own methods of tracking and interpreting cycles, though 90% of that is based on generic market cycle analysis and Hurst.

If you look at some of the earlier posts in the 'cycles' thread over at Capital Stool you'll see a system I was developing using sine waves at the octave - I keep that to myself now as it works so well I'm looking into copyright/tradmark possibilities, but you'll see it in it's infancy if you dig around.

In any event, Hurst tends to focus on a set of market cycles (ref page 33 'The Profit Magic of Stock Transaction Timing). You'll notice starkly that 8 years is not amongst those common cycles. My simple answer is that gold is not a standard stock. In fact - it's a commodity, and even in that sense it is unique. My more esoteric answer is - it's been happening, plain and simple.

I've seen folks use fast fourier transforms for the POG and come up with everything from 7.1 years to 9 years as the 'dominant' long cycle for gold. I just break everything in life that I can down to it's simplest form and count the darn candles on a chart and if you check the link above you'll see why my system gave a 'sell' on the 8year.

I'm just trying to better understand the situation here no offense intended in any way (  :

No offense taken. 100 people can look at a thing and describe it 100 different ways. Many of those descriptions will completely conflict with others. If they all had to place a bet on their description, then, in financial terms (which what we're interested in) the one with the most money has given the correct description. Does that mean it's the right description ultimately? I don't know. All I can say is that when it comes to markets - the money is what counts to me, and I stick to and refine methods that make me money. If someone can make money doing things 180 degrees differently than me - hey, more power to 'em.

Also, I don't get paid for my charts so for anyone who doesn't like them; I'm glad for the feedback but surely don't feel pressed to have to look at them because I'm only human and paid or not I'm subject to err.

Best Regards and I wish you well in all your trades,

-Seven

[stockbucks_coffee]

Good thread here...I'm too have seen the 8yr cycle at the Sharelynx website...and even though POG and xau/hui have reached their 200dma, which in the past years have marked m/t bottoms, I'm looking at couple years down the horizon. POG and xau/hui are at levels which should be at m/t bottoms. But as in earlier post in another thread, I have not been so nervious about whether this is the 'bottom'. If this the bottom then the 8yr. cycle have finally inverted 'up' from the past gold secular bear market and we are in fact in a secular bull market for the gold. but if the 'bottoms' drop from here for POG and xau/hui then I can only suspect that we are heading into some deep [bleeeep]...deflationary problems..not just here but possible globally. at best it's a deep this-inflationary issue (which is what Japan has been in the last several years.) time will tell soon as to whether POG and xau/hui hold around this bottom support levels. my assumption is that they would not be 'down' for long...they will eventually hold and move up. SC.

[stockbucks_coffee]

typo again... from But as in earlier post in another thread, I have not been so nervious about whether this is the 'bottom'. TO But as in earlier post in another thread, I have not been so nervious about whether this is the 'bottom' UNTIL NOW...it may not be the bottom yet.

[SevenOfEleven]

Good thread here...I'm too have seen the 8yr cycle at the Sharelynx website...and even though POG and xau/hui have reached their 200dma, which in the past years have marked m/t bottoms, I'm looking at couple years down the horizon.

It's interesting, that's for sure - would love to see more cycle discussion.

Even though I already said I track the POG (the metal) I should probably chisel it harder into the marble by stating that I don't track cycles on the XAU or HUI. Doesn't mean they're not there - just means I don't track 'em, and it doesn't always appear that they track the POG 1:1. Primary reason: I trade physical.

POG and xau/hui are at levels which should be at m/t bottoms. But as in earlier post in another thread, I have not been so nervious about whether this is the 'bottom'. If this the bottom then the 8yr. cycle have finally inverted 'up' from the past gold secular bear market and we are in fact in a secular bull market for the gold. but if the 'bottoms' drop from here for POG and xau/hui then I can only suspect that we are heading into some deep [bleeeep]...deflationary problems..not just here but possible globally. at best it's a deep this-inflationary issue (which is what Japan has been in the last several years.)

I couldn't have said that better myself. One thing I might addend is that I do see a possibility for inversion of the coming 5 years down on the POG 8year cycle but I'm not fixed into it having to happen directly on an inversion point. That is to say, I could see 2.5 to 3 years down then a flip. My reason here is purely symmetry (I did say I break things down to their simplest denominators) on the POG. The large basin we've carved out has a three year left arm on it - so why not a three year right arm to match? Just a speculation there.

time will tell soon as to whether POG and xau/hui hold around this bottom support levels. my assumption is that they would not be 'down' for long...they will eventually hold and move up. SC.

Some of those people I mentioned that I respect are still looking at this as a longer term bull market correction. I believe the expectation in that arena is a retest of the highs (fakeout kinda thing) followed by a frightening drop, another rally followed by another drop (basically a subdivided 5 wave ABC pattern of some sort correction) which is to be followed by a hUge upward drive (think of the past three years as wave I up, the corrections mentioned as wave II down, to be followed by a massive wave III up). This same line of thinking, to me, portends a rather deflationary bloodbath of a depression to come so I'm rather hoping they're wrong.

One other view by 'Major Crapper' over at capital stool shows a chart with 52.3 year cycles on the POG (it's on the last page of the 'cycles' thread in the 'Stools Gold' forum) that, if all follows as it has, would have us downtrending (with rallies) for years to come. The chart is segmented 'low to low' with the last low being the early '70s and the high in between being 1980. So, tack 52.3 years on to 1972'ish and you get us heading into the next low around 2024. That's darn bleak if you're a gold bug.

Cycles themselves are nothing new. In fact, they may be the oldest things around. We orbit the sun about every 265 days. We rotate on our own axis about ever 24 hours. These are dependable enough that we base just about everything we do on these basic cycles. The sun goes through an 11 year sunspot cycle. There are extreme weather cycles. When a doctor takes your resting pulse he's depending on a basic cycle theory to tell him/her if something is wrong with you. Hook an EEG up to your noggin' and you'll see very consistent cycles. So, it makes perfect sense to me that whether it's subconcious or not, our financial markets tend to have a lot of cyclicality burried in them. It seems only natural to me.

-SevenCents

[thakid]

Intersesting. I know Mr Harley personally, i studied underneath him for several years...great guy ( : I worked with Clyde Lee too (though he did the majority of the programming) on developing a sin wave system (advanced fourier with price value projections) along the lines of what you speak. We have perfected it and it works wonderfully. I even have got it to the point where we can tell the exact probability of direction for of each of the cycles, its very powerful. Cost about 5-10k to develop not to mention the time i spent testing it. I was debating whether i should break it out on the forum because very few people truly understand the power of it, and it only tends to confuse those who are not familiar with hurst. Heres some intersting stuff about the system. maybe if you show me yours i 'll show you mine Heres the basis of the system, i am the only person i know of who has the code and clyde will not allow me to share it with anyone but he will allow me to post the signals. It took me 1 year to get it to the point where it is now. I question if even clyde can do better... but he is the Guru so i am indebted to him for he has helped me tremendoulsly. Remember my mathmatical knowledge is limited so i'm hoping you will bear with me if i am a little slow. Heres the jist of it. One of the arguments (particularly by mathematicians) against the use of sinusoidal elements to describe the market is that cyclic data is NOT stationary and therefore is not of any value in measuring what the market might do in the future. That cycles are not stationary is quite evident in this study. That cycle information IS valuable for anticipating market movement is in my mind just as obvious from a casual observation of this study. That certain cycles fade to nothing and that those fades occur when the market is trending is very valuable to me. That certain cycles develop significant amplitude and that when such happens then those cycles tend to follow/predict the market for some useful period of time is very valuable to me. This analysis uses a new tool that I am releasing as soon as I can get it up on the website. Those of you who have followed the "new-gun" development will recognize that this is basically what new-gun does. The tool calculates and plots the instantaneous value of a selected period cycle on the current bar based on an analysis of the phase and maximum amplitude of that particular cycle over a period of two cycles back from the current computation. At the end of the data, using the maximum amplitude and phase of the given cycle, then indicator continues forward with the cyclic calculation. One study consists of 8 cycles that were pointed out by the FOURLINE application which is a conventional Fourier Line Spectra approach to definition of useful cycles. The other study consisted of 8 cycles that were pointed out by the FOURLINE_MEM routine which uses Maximum Entropy Method for definition of cycles. Look at the bottom of the chart and see which analysis routine is in use. This computation is performed on every bar for every cycle that is defined. The limitation of two cycles for each application is because Tradestation only allows four plots total for any indicator and since I want to see the Exponential Moving average of the derived cycles then we get two cycles and two averages. What do you think?

[SevenOfEleven]

Intersesting.

I know Mr Harley personally, i studied underneath him for several years...great guy ( :

I worked with Clyde Lee too (though he did the majority of the programming) on developing a sin wave system (advanced fourier with price value projections) along the lines of what you speak. We have perfected it and it works wonderfully. I even have got it to the point where we can tell the exact probability of direction for of each of the cycles, its very powerful.

Cost about 5-10k to develop not to mention the time i spent testing it.

I was debating whether i should break it out on the forum because very few people truly understand the power of it, and it only tends to confuse those who are not familiar with hurst.

Heres some intersting stuff about the system. maybe if you show me yours i 'll show you mine

Heres the basis of the system, i am the only person i know of who has the code and clyde will not allow me to share it with anyone but he will allow me to post the signals. It took me 1 year to get it to the point where it is now. I question if even clyde can do better... but he is the Guru so i am indebted to him for he has helped me tremendoulsly.


Remember my mathmatical knowledge is limited so i'm hoping you will bear with me if i am a little slow.

Heres the jist of it.

One of the arguments (particularly by mathematicians) against the use
of sinusoidal elements to describe the market is that cyclic data is NOT
stationary and therefore is not of any value in measuring what the
market might do in the future.

That cycles are not stationary is quite evident in this study.

That cycle information IS valuable for anticipating market movement
is in my mind just as obvious from a casual observation of this study.

That certain cycles fade to nothing and that those fades occur when
the market is trending is very valuable to me.

That certain cycles develop significant amplitude and that when
such happens then those cycles tend to follow/predict the market
for some useful period of time is very valuable to me.

This analysis uses a new tool that I am releasing as soon as I can get
it up on the website. Those of you who have followed the "new-gun"
development will recognize that this is basically what new-gun does.

The tool calculates and plots the instantaneous value of a selected
period cycle on the current bar based on an analysis of the phase and
maximum amplitude of that particular cycle over a period of two
cycles back from the current computation. At the end of the data,
using the maximum amplitude and phase of the given cycle, then
indicator continues forward with the cyclic calculation.

One study consists of 8 cycles that were pointed out by the FOURLINE
application which is a conventional Fourier Line Spectra approach
to definition of useful cycles.

The other study consisted of 8 cycles that were pointed out by the
FOURLINE_MEM routine which uses Maximum Entropy Method for
definition of cycles. Look at the bottom of the chart and see which
analysis routine is in use.

This computation is performed on every bar for every cycle that is
defined. The limitation of two cycles for each application is because
Tradestation only allows four plots total for any indicator and since
I want to see the Exponential Moving average of the derived cycles
then we get two cycles and two averages.


What do you think?

Sounds very impressive, you've obviously had much more experience with markets and cycles than I.

I come from a software engineering background, though, as with you, my math knowledge is limited. That's probably what got me into software - the ability to layer functions for ever increasing levels of complexity. If you can add, subtract, multiply, divide and read reference manuals - you can write higher level math software.

Prior to software I studied music. I always found harmonics and the concept of 'sympathetic vibrations' fascinating.

If you can locate a given set of harmonics in a given composite and they point you to a primary, it's confirmational of that primary. With markets there are harmonics that fade in and out, highlighting different primary cycles.

If you use simple fourier transforms to locate both primaries and harmonics, discarding all that does not confirm within a given timeframe, you tend to hone in on a very strong primary cycle. That is to say, one of high relative strength or persuasion.

Next, you remove all but a select set of harmonics, and add a few that may or may not be present, as an overlay. In doing so you end up areas that repeat in their overlay pattern (picture a piece of lace ribbon). There are very specific areas that tend to fall very tightly the turns (e.g. top or bottom of a cycle). Because the pattern repeats, it's possible to use this pattern to forecast up to two cycle turns forward with high accuracy (exactly how much, I won't know until I've done a lot more back testing).

Every new bar that comes into whatever timeframe you're examining tends to alter, even if only slightly, the harmonic structure. Thus, the system can adjust it's forecast as required right up to the point where the event actually occurs (nothing new here - most systems I know of benefit from constant adjustment to market flux).

This system is only a partial building block to be used with other systems because it does absolutely nothing by itself that forecasts levels; only the time at which turns are likely to occur. So, it's to be used in conjunction with systems such as Hurst to enhance timing.

The internals are built entirely upon sine waves. It's how they're used and updated that creates a ribbon which does indeed augment and diminish it's pattern in order to dynamically hone in on it's target.

The only thing really proprietary is the system of harmonics I use. There is more than one set that is used depending on a few variables, and the system can automatically use it's own generated signals to determine what set of harmonics to use at a given point in time. In that sense it's a bit SKI-like, but the resemblence ends there.

How's that for windy?

[thakid]

Cycles1

Cycles2

DollarCycles

Cycles3


Heres some past cyclical work i did on the $Dolla and the Cross Rates.

I think if you can crack the dollar then gold is an easy game. The dollars especially easy to figure out due to the fact the answer to it lies in the EUR...GBP and JPY...due to their inverse relationships (uh-oh theres that word again) .

Breaking the cycles down like this is IMHO what RN Elliott was dabbling in years ago. I think he had a different method to the madness but yielded similar results (all roads lead to rome). All in all you can basicly break down every market into 3's and 5's (and sometimes 6's if its a vvvery strong trending market) .

Even works on the USD/JPY now that is a frightening market to trade

I have some better pattern breakdowns too, these are just some cycles i was posting for a few guys on the elliottwave message forum to help them with there wave counting.

AlexB and AlexN have been friends of mine for some time.

[dougie]

Why should one assume that fallinng gold prices equate with deflation or disinflation. i see that assumption an awful lot. in fact there have been significant periods of defaltion in which gold did just fine. A quick google will show those periods.

[SevenOfEleven]

Why should one assume that fallinng gold prices equate with deflation or disinflation. i see that assumption an awful lot. in fact there have been significant periods of defaltion in which gold did just fine. A quick google will show those periods.

I don't even see any (what I would consider) serious instances of deflation aside from the great depression. If I'm not mistaken PMs actually went into the spiral with everything else; they were simply the first things out of the gate blazing once the dust settled. Any historians on board?