Fibonacci Retracements

[Philby]

There is a recorded GET session by Nate MacCarthy on the web site covering Fibonacci Retracements and XTL techniques and in short Nate points out that price generally retraces to the 38% point. A number of examples are given to demonstrate this. In fact I pulled a number of local equities up and tested it out and gotsimilar results.

However in the Precter & Frost book on Elliot waves in the ratio analysis section in Chap 4 p127 they say that retracements go back 50% - 618% in general. The logic behind this also makes sense.

Can anybody throw some light on this difference of opinion / fact ?

[MR]

With respect and deference to Frost and Prechter's book on Elliott Wave theory, when was it first published? 1979, '81? Not many of us had a personal computer when that statement might have first been penned. Secondly, if you explore the totality of Bob Prechter's later writings, he probably communicates with more clarity on that 50-62% general retracement notion.

With respect to coworker, Nate, his Web-Ex presentation is not complete but a quick overview of the subject matter. When taken in context with all our other AGET works we find a multitude of retracement chart patterns where 50 to 62% retracement patterns, as well as other retracement values, are the standard. For example, I have shown charts where a 75% retracement is doable. I have shown where in the SP500, Live Cattle, High Grade Copper, some might focus in 75% retracements. I have shown in the US Bonds a certain type of 90% retracements I find very attractive. The point is there also are other retracement patterns unique to a market and at times they can be the norm. Only experience and observation helps define this. If you wait for 50%-62% retracements because they "generally" happen, you miss many opportunities and can get prematurely stopped out at times.

We need to better clarify this broad statement, "retracements go back 50% - 618% in general." It may sound like a true general statement, but traders don't bet real money on generalities, we bet them on higher possiblities and probablities and a whole gambit of other variables when trying to better define which retracement is most appropriate for a trade setup.

[soylent]

Fib's...

Well, I can't say why the instructor didn't talk about other levels, but it's generally accepted that 25, 38, 50, and 62. are the invisible magic retracements.

Anyhow, probably from the instructors view-point, .38 is a sign of a strong trend just experience profit taking and since he was talking about using XTL -- he probably was pointing you to the best opportunities to use XTL. .62 retraces are weaker trends or basing patterns. And, .25's will happen 'during' the heavy trending.

ALSO (while I'm on a roll...):
Note that Fibonacci support (invisible lines) are not as strong as real support (visible support lines drawn from tops/bottoms). When analyzing trends, the more converging lines, the stronger the possible move.

As an example, here's INTC. You can see it bounced on .38 (a strong indication of a continuation) but the long-term trend was extended so it double topped. This sent it down to a .62 retrace. It triple-topped. We are now bouncing down to .38 AND hitting our visible upcoming support line.

If we lose .62 and that visible support -- look out -- all hell will break loose for INTC and all the indexes...

Best of luck on your journey!

[fabrizio]

Gentlemen

A) Please do not follow Pretchers book as a bible. If you really want to understand EXACTLY what RN ELLIOT meant , please refer to:
The Wave Principle (1938)
by RN ELLIOT .

B) According to TJ approach - and Marcus is a archibishop this Church- the retracement occurs within the indicated Fibo Ratios . And they do. And You can safely6 trade safeguarding your capital.

C) I am an Humble believer of this church and I follow TJ approach , albeit in a less "safe " manner.

In waves of the same degree I use .386; .500; .618
I repeat it is LESS SAFER than what conceived by TJ.
For if the movement exceed a .618 can go straight to a 1:1 or worst to 1.272 and above. And it does, as happened to me the day before yesterday on EMINI.

So my stop were wide and my day ended with a significative loss.
I I would have followed the TJ approach, it would have been a considerably profitable day.

However the situation was exactly the opposite yesterday on ES on 5 Min when an hefty profit was harvested going EXACTLY against the market and triyng to find a top that actually was not a W4- W5 unfolding but a B Major that gaved me the whole space for a movement of a C which was at least among 1.62 to 2.62 of B to A;
But the main point was that I was supported in this by an almost flat 5/17 osc that eventually turned on my side ..... I will never be tired enough to say that labelling means nothing if you do not focalize on:
-Ratios among waves of same degree
-OSC
-Localize Elliot.

Hence -correctly- down there , was a nice mob on which proximity I closed my positions.

[fabrizio]

ALSO (while I'm on a roll...):
Note that Fibonacci support (invisible lines) are not as strong as real support (visible support lines drawn from tops/bottoms). When analyzing trends, the more converging lines, the stronger the possible move

I beg your pardon .

Can you clear what you mean by fibvonacci invisisble line?
And are ypu quoting somene lese or just reporting your own opinion?

Thanks

[soylent]

> Can you clear what you mean by fibvonacci invisisble line?
> And are ypu quoting somene lese or just reporting your own opinion?

hmm... I don't really know. Surely I may have read something like that somewhere and I'd like to give credit if it is due (or, share blame, if that's what you're thinking! <smile>).

But, in all honesty, it may just be my own opinion.

What I mean by it is:
Psychologically, 'NON TRADERS' (which I would refer to as anyone that doesn't look at this stuff daily) will follow pivots.

Because they probably BOUGHT at the last pivot and want to get out at 'break even' thus confirming the pivot.

And, while I believe fibonacci is psychological, too... (e.g. Non-traders will take an 'acceptable loss' at a .38 retrace because the law of the order of nature (fear/greed) says they will) and .62 shows enough of a sell off to bring non-traders back in to 'catch the dip because it's a bargain now!'.

MORE people will follow the visible pivots than the 'invisible' fibonacci.

HOWEVER, I should clarify, in INTRADAY markets, you primarily compete with other professional (amateur) traders that will follow the invisible fib lines as closely as the pivots. I don't intra-day trade except to get a good entry price on a stock I want to go-long (I'll watch the action and set a buy LIMIT on a FIB line and catch a quick selloff bounce to a good fill price -- most of the time).

Thanks for all your posts Fabrizio -- I always enjoy what you have to say!

Best of luck.

[fabrizio]

Thanks to you Soylent

Your thought are wise , although I 'm not on your side 100%.

Traditional TA and Dynamic TA.

Pivot - or point or swing- is just an elemenet of how waves relates among them.

I personally would not dare whether Pivot are "psichologicall levels or not.

In the other hands, law of nature, I can easily affirm that Fibo extensions are absolutely "firm & concrete" in the price movement.

Well the simple and "affordable " approach of AGET is just this . All is ruled by Fibo extensions , ratios among waves of same degree . I 'm sorry I insist in saying same degree , for if you mix two diffrent degrees you are in trouble...

Let me be blaspheme simplifiyng all:
12345 , 3 impulsive subdivided in sub 12345 .....omissis.......... and 2 corrective subdivide in 3 (ABC)......omissis.
That's all : the greatest "metafora" of life.

Indeed Mr RN Elliot made all more complex and charming , but TJ skimmed "the 70% too complex to be traded and kept the practical 30%" To make money.

Plesae do not misunderstand me: I'm not inloved with either Mr. TJ or Rn Elliot.
Just like to express once more how effective is the approach.

And there is no other soft able to accomplish in trading what safely does AGET.

With all my admiration for Miner and others.

And how simple is , once you are able to "read " the chart and the ratio and the Osc..

Thanks for the patiente to read.

[marginman]

great summary fabrizio.....invisible fib levels or merely less frequently fib levels include:

21% and 89% for retracements, the weak 21% retrace is associated with strong impulse waves, the 89% with weak say a w1 impulse wave. Also u can find these retrace level depending on which pivot point u choose within an impulse wave. As Fabrizio accureately stated, markets are fractally symmetrical, so a w3 subdivided into 5 smaller waves offer pivot points from which u can find confluence for a w4 correction at say 50% retrace for the entire w3 or 89% from the sub w4 label within the overall w3.

1.382, 2.764, and 4.236 for extensions, the last one is a default in AGET.


Kindest Regards, Dean

[soylent]

ADP Example of Trend vs. Fib

Fabrizio ---

Thanks for your reply -- I always enjoy your teachings and have much to learn from you and others in this forum.

I wanted to show an example of what I mean by visible / invisible lines and how they interact.

Perhaps you will have better thoughts on this -- or, suggestions about why I would/wouldn't take a trade based on pure FIB #'s vs. FIB + TRENDs.

Here are a couple more thoughts -- open for debate!:
a) When learning intraday fibonacci (anything shorter than DAILY bars) learn the indexes or mega caps (>30B Market Cap and > 10M daily volume).

(actually, I'd say learn the megacaps /indexes first -- period)

Reason: Big caps are more reliable on an intraday basis. Smaller caps follow well on DAILY bars and can be used intraday for entry timing but often 'dry up' so much from 10-2 that nothing happens and a big sell order can really tank the stock...

b) Always know when news is coming out.
Never take a trade into EARNINGS for a company. Even if you know it will be good -- it's 50-50 at best. I've actually found the corrolary to this to be true: If a stock drops into a WAVE 4 ahead of earnings, it's often good to take the trade because the bad news is priced in. If a stock rallys into a MOB level before earnings -- get out! It may go up -- but, it will surely correct into a wave 4 for a safer entry.

c) SLOPE:
Don't expect your WAVE 5 to rise faster than your WAVE 3. As the ADP shows, the buggers tried to buy and ride a .38 bounce into a quick break. If WAVE 5 is steeper than WAVE 3, expect it to fail early.

On ADP, even though the first MOB tries to show an early break -- momentum is dangerous...

Any more rules / observations anyone else might like to share?

[soylent]

INTC Example

OK -- I got to playing some more and I wanted to show how all this applies to 'XTL' and trends vs. elliot waves.

The original question was regarding .38 retrace in the XTL class.

Fabrizio: I think you summed it up very well in saying 'comparative wave size'. And, 'Localized Elliott'.

So, here's a sample from Oct on INTC using a sub-section of a big wave 3 (localized elliott) showing a .38 retrace. I don't have XTL on -- but, I'm sure it would say we are 'trending'.

My purpose here is to answer the original Question about XTL and .38 retraces / continuations and showing the 'help' one might get if they drop a trend (visible support/resistance) lines over the whole equation.

I encourage any feedback -- I am still learning all this...

Thanks all.

(One note: You'll see the first breakout fails -- this goes back to the elliot wave 4 not being 'complete' because the Oscillator did not pull back to zero / near zero...)

[fabrizio]

Soylent

I think you are learning well and fast!

Most of the people approaching AGET do not use the Localize Elliot: actually it can save your........many times.

I'm glad you brought it to the attention with your chart.

Your trading thoughts and sugggestions are wise and good.

About the W5 slope: the focus-IMO- is always on the relation between the prior movements. Logically a strong trend must vibrate not showing "doubts" . As soon as it does he can reverse at anytime ...

Look what is happening on the S&P weekly from end nov up to now.