## Friday, September 2, 2016

### Treacherous Triangles

I have decided to post this update intra-day today. From yesterday's post, we were expecting upward movement, potentially, in wave 4, upward, of a downward contracting diagonal. Well, upward movement is what we got. However, yesterday's post provided clear invalidation points, in that such a wave 4 could not become longer than a wave 2. And, in fact, that invalidation did occur on the lighter than expected jobs report. As always, diagonals and triangles must prove themselves, and a downward diagonal did not.

So, during live chat, I re-studied the potential triangle pattern and settled on an hourly chart which appears below. What was initially bothering me was the fact that a contracting triangle should have only one complex leg, and I kept coming up with two complex legs - until I found another pattern.

 SP500 Hourly - Triangle Re-Established

The pattern that was missed was that - inside of the circle-c leg - there was yet another valid contracting running triangle, which was in the (b) wave position. And there was, indeed, a five-wave sequence following it down to the (c) wave of circle-c. You'll even note within this triangle that the e wave, up, overlaps the (a) wave down, making the triangle valid. That now makes a simple (sic) leg for the circle-c leg of the triangle.

So, circle-a is simple, circle-b is complex, circle-c is simple, and waves circle-d and circle-e should be simple zigzags if a triangle is to play out in valid form. And, again, circle-e must come back down to overlap wave 3 to the left.

Ah, you say, "that downward overlap could have already happened, right?"

Well, it could have, but I don't think so. Here is a very short term chart of the SP500 5-minute chart counting the circle-d, wave up.

 SP500 5-minute Chart

This count had actually carried over from yesterday. We had counted waves 1, 2, 3, 4 and part of 5, up, which seemed incomplete.  As far as can be told right now, we finished wave A, up, of circle-d this morning. Then, we got an A, B, C down both to the 38.2% retrace level, and where C is nearly exactly, 1.618 x A. We think this is the B wave, down, of circle-d up. We even called the internal wave iv of C as a triangle, and the wave v, down, on the mid-line of the channel.

If prices land on or slightly above the high on the close, it may wrap up the C wave, up, of circle-d, up, and in fact, wrap up the circle-d wave.

Circle-b and circle-c now provide the clear invalidation points for wave circle-d if this is indeed a running triangle. And it is hoped this example shows the value of an Elliott Wave count. We clearly noted where the diagonal option would invalidate. And when it did, we quickly and swiftly reacted. If your Elliott analyst is not giving you clear invalidation points, you are only getting half the story!

We apologize for originally missing that internal triangle of circle-c. As far as I know, no one else has pointed it out prior. But we wanted to get back on track, and keep you up to date.

We hope this helps clarify things as the long weekend in the U.S. approaches.

1. Thank you for this intraday update. I do have a question. In yesterday's update, you showed how the downward waves from 2187.94 to 2160.39 and from 2183.48 to 2161.35 were both impulse waves that subdivided by 5. However, in today's update, those two waves are now labeled wave b and wave d of the newly discovered contracting triangle. I thought that all waves of a contracting triangle had to be corrective waves that subdivide by 3, and could not be impulse waves that subdivide by 5. What am I missing? Thanks.

1. Hi. In the Aug 29 - Aug 31 five wave down sequence, just move the wave shown as 4 one wave-peak to the left, and place a 5 at the bottom of the next wave down (currently unlabeled - as if 4 was a flat wave), and then that 5 becomes (a) down, followed by (b) up until the gap down. And then the gap down wave become (c) down in five more waves. It all still fits like a glove, but that's the problem with triangles, the waves can get so contorted and overlapping it may initially be hard to sort them out. Good question!

2. Thanks again, Joe!

3. Chapeau Joe. Have a great weekend

4. Thanks Joe. Would this triangle be wave 4?

1. As far as I can tell, if the triangle completes properly, it would be wave 4.

5. Thanks TJ. The newly identified triangle is indeed a large piece of the jigsaw puzzle.

Need your help on a basic question - what are the Elliott wave rules of zigzag pattern. Neely wrote that "termination of wave-b will complete at 61.8% of wave-a or less". Is this a rule or guideline base on your experience?

1. Sorry. I missed out the reference, which is as follows.

Page 5-20 (Mastering Elliott Wave)

2. The only 'rules' for zigzags are 1) that they must be comprised of a 5:3:5 pattern, counted as a-b-c, where 2) the b wave may not exceed the start of the a wave, and 3) the b wave can be 'any three', including a flat, a triangle, a double or triple combination, or another zigzag, double zigzag or triple zigzag.

The guidelines are that 1) often c = a, c = 0.618 x a, or c = 1.618 x a, or c = 1.272 x a, 2) often multiple zigzags form a near perfect channel while single zigzags usually do not hit the exact boundary of a channel, sometimes failing at the middle channel line, and 3) most-often, and usually, c exceeds the end of the a wave, but 'very occasionally' c can fail to move beyond the end of a.

Hope this helps.

3. Thanks for sharing these rules and guidelines. They are certainly helpful. Appreciate them.

6. TJ- in leading diagonal- does 3 have to be smaller than 1. is that a cardinal rule or a guideline.
And when you say 4. is this the 4 of the move from 1990s or from 2000s.

thanks

1. sorry meant 1990s or 1800s

2. In a contracting leading diagonal, wave 3 'must' be smaller than wave 1, wave 4 must be smaller than wave 2, and overlap wave 1, and wave 5 'must' be smaller than wave 3.

In an expanding leading diagonal, wave 3 'must' be larger than wave 1, wave 4 'must' be larger than wave 2, and must overlap wave 1, without traveling higher/lower than the end of wave 2, and wave 5 'must' be longer than wave 3.

Hope this helps.