I analyze macroeconomic issues from a fundamental perspective, and I analyze market behavior from a technical perspective. Original macroeconomic analysis can be found here and both macro analysis and commentary can be found on my Caps blog. If you like or appreciate my analysis, please add yourself to my Following List

## Thursday, January 6, 2011

### Five-Wave Structures Revisited: The Identification of an Impulse Wave

Last year I tackled this subject (see Not All Five-Wave Moves Are Impulses: A Short Treatise on Elliott Wave). And I want to tackle it yet again from a slightly different perspective. This is not simply some academic exercise in Elliott Wave Theory. There is an end purpose here with some very relevant implications.

But let's start at the beginning.

There are three references for this post and you really should read them first:

1) Not All Five-Wave Moves Are Impulses: A Short Treatise on Elliott Wave
2) Another Impulse Wave Study: A Look at the 1974-1975 Low and Rally
3) Historical Count: 2002-2007

The key question before us right now is: "Was the move from March 2009 to April 2010 an impulse wave?". I have said no before. But recently a lot of large counts in the blogosphere have changed to say that the move was in fact an impulse (including, I believe, EWI. EWI subscribers please confirm). The idea is that the move is an impulse which is forming the A leg of a Primary Degree zigzag for P2. On the other hand, if the move is in fact an impulse wave, then we have the possibility that the correction is over and a new secular bull market has begun.

So, you can see that the answer to this question is extremely relevant and will either open or close several possibilities.

I will try to answer this question from the ground up. Meaning that I have studied impulse waves from a number of different time periods to not only show their characteristics from the textbook descriptions, but how they have actually manifest in the equity markets. What are their differences and similarities? What are low and high probability characteristics? Does the Mar 09 - Apr 10 waveform fall within the norm of observable impulses or is it an extreme outlier?

First, how is an impulse wave identified? Reference 1 above gives some very good guidance beyond what you read in the Elliott Wave books. Here is a summary of my 'lessons learned' in identifying impulse waves through historical market analysis. These are my takes on the EW guidelines to how 'the look and feel' of an impulse wave should manifest.

1) Dynamics in the third wave. The third wave should nearly always be the longest and strongest in the equity markets. Sometimes it is the fifth wave. It is almost never the first wave. The first wave is a 'throw-away' wave that signals a primary trend change but is almost never recognized as such. Hence their tendency to be mostly retraced by Wave 2's. In my studies I have seen that strong 'first waves' are almost always the strong reaction in a corrective wave

2) Wave 2 is almost always a deep retrace wave with 50% being typical and between 38% and 62% being 'in norm'. A wave 2 that does not retrace 38% is likely not a 2 (and the structure is likely not an impulse). There is an exception to this tendency, which is in the middle of an extension (particularly a third wave extension), the series of 1-2's that develop, the new 1 always goes past the 1 of the next higher degree and the 2 becomes shallower on each subsequent extension. The whole sequence has a very pronounced accelerating quality to it.

3) Wave proportionality and alternation. Wave 2 should be in proportion to wave 4 (if wave 2 lasts 30 minutes and wave 4 lasts 5 days, then something is incorrectly identified) and wave 2 and 4 should alternate in form (tendency not a rule), but absolutely should in severity. If 2 is a very deep retrace wave, 4 should be shallow. If 2 is a complex wave, 4 should simple, etc.

The bottom line is that simply 'counting from 1 to 5' is not sufficient to identifying an impulse wave. The internal wave structure must conform to be a valid impulse wave.

From reference 2, I sought to answer the question "The theory is that 1969-1974 was a vicious up and down bear market with a *severe* drop from 1973-1974 (about 50%). So perhaps the wave forms get distorted in an impulse wave after a severe drop? Let's see if this idea is backed up by the historical record."

If you read the post, the conclusion is no. Even after a severe drop, the impulse wave that signaled the bull move was a very clean impulsive wave form. It displayed all of the tendencies listed above that we look for in an impulse wave.

But that wave (1974-2000) was a Cycle Degree impulse wave, whereas 2002-2007 was a Cycle Degree corrective wave. So perhaps the impulse waves within a larger degree corrective wave tend to be more distorted than they are in a larger degree impulse wave? Seeing as how my position is that the current wave is a large (either Primary or Cycle degree) corrective wave, refutation or confirmation of this idea is extremely relevant. Especially to those that claim the Mar 09 - Apr 10 is in fact an impulse forming the A leg of a zigzag.

That is exactly what reference 3 above sought out to do. It was a detailed study of the impulse waves within a large corrective wave. Please refer to that post for the detailed wave counts. Here are the conclusions:

-- Cycle B is a large Cycle Degree Zigzag and Primary A and Primary C are clean impulse waves. The Intermediate / Minor / Minute degree impulsive subwaves (1, 3 and 5) upon examination are also clean impulse waves.

-- There were two cases (Int 1 of P-A and Int 1 of P-C) where the wave structure was not a clean impulse. But these waves did morph into Leading Diagonals and the subsequent Wave 2 retraces were ~62%-78% as expected

-- There was only one wave with a highly distorted impulsive waveform, and that was Int 5 of Primary A. Despite being distorted, Minor 2 was still a deep retrace wave and we see acceleration in Minor 3. We do not observe very small retrace 2's until the Minute degree. This behavior is clearly the exception and not the rule

-- In all, the impulse waves even in the middle of a large Cycle Degree corrective wave still exhibit the characteristics of an impulse (see above) and are readily identifiable as such.

Okay.

So now what I have demonstrated is that via historical analysis, from both a Cycle Degree impulse wave and a Cycle Degree corrective wave, looking all the way down to Primary / Intermediate / Minor / Minute, etc. waves that the impulse waves are seldom distorted and the vast majority display all of the characteristics of an impulse wave that we look for.

This means that any 'impulse wave' that does not display the characteristics listed above is likely not an impulse wave.

So let's re-review the March 09 - April 10 impulsive wave count.

First, this count is invalid. There is overlap between 4 and 1 in the middle of the third wave. This isn't a leading diagonal or an ending diagonal. But a rule break in the middle of what is typically the strongest wave. Not so great, huh.

But let's ignore this rule break for a moment and see how many rules are bent.

-- Wave 1 is by far the strongest wave in the sequence. This is almost never the case. This is clearly a departure from norms.

-- Wave 2 does not even come close to retracing 38%. This is a major departure from all historic impulse waves.

-- Wave 3, beyond ignoring the 1-4 rule violation, shows pronounced deceleration as the wave progresses. Whereas the norm is for a Wave 3 to accelerate through the wave and especially in comparison to a Wave 1. This does neither.

-- Tackling the violation, the end of Wave 3 should be in Jan 2010. This is a huge problem. There is no way to make a valid impulsive sequence for Wave 3. And then Wave 4 is almost exactly the same size and type of correction as Wave 2.

This wave is disqualified by being an impulse because of the major rule violation in the middle of the 3rd wave. But even if we can somehow justify this (which we can't), I have never seen an impulse wave with this many tendency violations.

So my previous conclusion still stands: The March 2009 - April 2010 wave is NOT an impulse wave.

The odds of this 5 wave move being an impulse despite all of the tendency violations is so small (maybe in the low single digit percentage chance) that it should be way down on the list of alternate counts.

Anybody who insists that this wave is an impulse, or who favors an impulse count for this wave over more likely (yet ugly) alternatives is doing a disservice to the entire Elliott Wave community.

There are two major ramifications to this conclusion:

1) The March 2009 but does not signify the start of a new secular bull market. Bull markets up are long term upward impulses, and the first wave of ANY impulse of ANY degree MUST ITSELF BE AN IMPULSE!!!. This is basic Elliott Wave 101.

2) The count for a Primary 2 (operating with in a Cycle C down framework) CANNOT be a zigzag (which is a 5-3-5). Whether we are in a Cycle X or a Primary 2, the count is more complicated and cannot legitimately be counted as a 5-3-5 impulse-three-impulse.

I will save my thoughts for the large count possibilities (and they are ugly) for tomorrows post. But the 'appeal' to a 5-3-5 count for P2 is that it could be ending here. And I think that is almost certainly not the case. I think we will have a top here, but I think there is almost no chance that it is 'the' top.