Saturday, April 22, 2017

Weekly Review - Two Plausible Scenarios & Just One Preferred

Some websites want to send you back to the beginning of recorded stock price history in the U.S. to tell you why their Elliott Wave counts, which don't seem to work out when viewed under close scrutiny, must work out in the future. Just recently, one (ahem) shall we call it alternative Elliott Wave site, which claims it is objective, was projecting higher and higher third waves - which, as you know now - never developed. But, the odd thing is the authors of such sites rarely, if ever, say, "we got it dead wrong". How can a site be objective if it won't admit when it's tools and techniques just don't work? Wrong at the May 2015 high. Wrong at the 2016 bottom. Wrong at the March 2017 high.

I on-the-other-hand prefer to stick to real Elliott Wave analysis, as I understand it. And I think many, many people really don't understand it that well yet either. That's why I try to help show you how real wave analysis works - on any time frame from 5-minutes (as in yesterday's post) to weekly, monthly or more. I prefer to live with some uncertainties of true Elliott Wave analysis because I know that's what mirrors life. There is almost always a choice. There is almost always an alternate. This week I will show you the two most likely Elliott Wave Counts on weekly charts, and tell you which I prefer and why. It is no change from the past several weeks.

SP500 Cash Weekly - Primary 5 Impulse Count - ALTERNATE

The first count above is of the potential Primary 5 impulse wave upward. As of this point in time there is nothing that breaks any true Elliott Wave Rule in this count. It would be made up of five Intermediate sized non-overlapping waves higher, and that would normally be fine. Just, if Intermediate (2) were a flat with the higher B wave, as shown, then one would expect Intermediate (4) to be a triangle or a more simple zigzag to provide alternation with the flat wave (2). That could still happen.

But, there are five reasons why I don't like this count that go beyond 'simple' wave analysis, and proceed to more modern wave analysis as it has been further delineated by the likes of Glenn Neely, and Bill Williams.

Fibonacci Five Reasons for Alternate Designation of the Impulse Count
  1. Part of Wave (3) breaks the Primary wave ((4)) to (2) trend line, indicating loss of momentum at that point. Wave 2, as you can see, drops below this lower channel line, and it is more extreme if the real zero-to-(2) trend line were drawn in. 'Usually', 'most-often' all of wave three is above the zero-to-two trend line except in diagonals.
  2. Wave (2) does not have a very deep pull-back - it is only 38.2% at maximum. 'Usually', 'most often' second waves are between 50 - 78% or more except when the first wave is the extended wave in the five-wave sequence. But, the supposed third wave is longer, ruling out wave (1) as the extended wave in the sequence.
  3. Wave (2) is not a sharp wave. 'Usually', 'most often' second waves within true impulses are sharp or zigzag waves, and this one is a flat. While not a deal-breaker, it is a cause for concern.
  4. Sentiment. As you know, we track sentiment weekly. And you can see from the chart below that while sentiment is falling off - as expected - from the high of March 1st, it is nowhere yet near the lows of prior waves. I would expect, before an upward turn, that sentiment will fall lower yet.
  5. Finally, the NYSE Advance-Decline line has recently made new all-time highs. 'Usually', 'most often' the $NYAD will diverge with price before the all time highs are made. That just hasn't happened yet. 
Sentiment Chart

Weekly Bullish Sentiment is Not Yet Near Prior Wave Lows

For the reasons above, and not because of any sacred mystical cycles or hidden proprietary analysis, the count below is currently the preferred one.

SP500 Weekly Cash - Ending Diagonal Count - PREFERRED

As you can see, because there is no Intermediate third wave yet, this count eliminates all of the problems with a third wave breaking the zero-to-two trend line. In fact, in this count you can see from the channel that minute wave ((iii)) exactly lives up to that guideline. Yes, it is not a rule, but it is a guideline we would rather waves live up to than not. Also, there would be four more Intermediate sized waves in the near future to show divergence with that advance-decline line.

So, that's the rationale. We can see both sides of this coin, but clear and unequivocal reasons are provided for preferring one over the other. The real item of interest is that in either the Intermediate Wave (4) wave in the Alternate, or the Intermediate wave (2) in the Preferred count, it is very likely that that down wave could have the form of a zigzag! And that is the amazing part. And that is why the ES 8-hour count published previously may be taking the form of the downward expanding diagonal that we have been showing for weeks now - because in a true zigzag, a real minor A wave is needed - made up somehow of five minute-sized waves.

But, the most important thing is that it is regular Elliott Wave analysis. There is nothing mystical about it. There is nothing mysterious about it. There is nothing here that doesn't already appear in two or three books, and a couple of videos all of which are available for only modest cost or no cost.

Are you learning your Elliott Wave? We hope so. Can we be wrong? You bet. But when we are, you hear it here first - along with why!

Have a great rest of the weekend.

SUPPLEMENTAL: At the request of a comment that asks good questions, I am posting these two one minute charts from the live chat room  that were done in real time.

SP500 1-Minute April 21 1546 PM ET
The first chart shows the detail that is possible today with good measuring tools and speedy equipment. At the time of posting this in the live chat room, I intentionally left all of the wave marker data on the chart, so a person could see that the measurements are exact for an expanding diagonal. Wave ((v)) is longer than wave ((iii)), which is longer than wave ((i)), and wave ((iv)) is longer than wave ((ii)), overlaps wave ((i)) but does not travel beyond the low of wave ((ii)). They are all three wave internal sequences that look like zigzags to me because they don't have "pull backs" in them.

But, this is  the key to me, if it is really a diagonal that I had identified, and not a series of one's and two's higher, then, it should have - as a prediction - at least a deep retrace if it is a Leading Diagonal wave up, or a failure lower if it is an ending diagonal. And here is what the result was ...

SP500 1-Minute Chart April 21 1559 PM ET

Well, there it is! Another Elliott Wave prediction come true. So, please be careful about criticizing building up larger fractals from smaller fractals because they are definitely there! But, stop being amazed there, and think. Look at the very top of the wave in the first chart. You notice that I did not call ((v)) done by placing it on a wave bar? That was because I couldn't find an ending formation yet.

And when you look at the second chart, what do you notice? C'mon now. There is a perfect little ending contracting diagonal for the (c) wave of ((v)). I just didn't draw in the trend lines. Can you?


  1. Thanks for your analysis, Joe. I have a couple of what may seem to be unrelated questions (but they aren't):
    1. What timeframes did Elliott use when counting waves?
    2. Did he use open-high-low bars, or did he use line charts (close only)?

    1. Welcome. Referring to the book, "R.N. Elliott's Masterworks," by Prechter
      1. Primarily hourly, daily and weekly for several reasons. Back then, there were no computers and so obtaining and cataloging anything less than the published hourly data was tortuous. Then, the papers would often summarize the daily and weekly statistics, making them more available.
      2. Elliot's own diagrams in the book showed he personally used close only charts for hourly. And he used "weekly range" charts - that we would call H-L charts for longer term work.
      Remember in Elliott's day, trading ended for the most part on the hour. And there was trading on Saturday. Only in modern times have we added the half-hour increment. A reference to the history of NYSE trading hours can be found here. (copy and paste link to your browser as needed).

    2. Here's why I asked the questions. If the theory and rules behind EW are valid, then it "should" work from tick charts (or 1 minute charts) to monthly charts. The most precise way to count waves would be to build up from the shortest time frame to the largest time frame, and it would never use close only line charts on any time frame. For me to now learn that Elliott used close only hourly charts as his smallest time frame is eye opening. That means that he could label something an impulse wave, when, in fact, it internally had lots of overlaps that he couldn't see. The reason I'm bringing this up is because I haven't been able to count any of the up moves since the 2/11/16 low as impulse waves, because the internal subdivisions had overlaps, or other disqualifying aspects. However, if I turn a blind eye to the internal subdivisions, then it's easy to label "almost" any move as an impulse wave. I'd like to know your thoughts about this. Thanks.

    3. First. Do not misinterpret what I said. Elliott often used close only charts on the hourly time frame. I did not say it was his lowest time frame for analysis. In fact, one day he proclaimed in an article, "decline ended in the last half-hour today", so he was at least aware of such time frames shorter than an hour. What he did with them on a 'regular basis is unknown to me".

      Since then, any scientific-type thinker realizes that hourly closes can miss a lot of variation in between the hour. How you deal with this problem is up to you.

      One thing that might be tripping you up is that if we are indeed in that zigzag upward, as in the Preferred count, then the retraces in the A and C impulses will tend to me short and shallow, and the alternation can vary with wave four being much "shorter" than wave two. That's just how zigzags work, and why they were named that; because they 'zig' and then they 'zag' without as much as a breath between them.

      While I do not like to modify a weekend post I will do it in this one instance to show you how I did in fact count a one minute wave at the end of trading on Friday, and made a correct prediction from it, live and in real time. Please check back after 2 PM for the update. It will just take me time to get it posted.

    4. Please see the supplemental section that was posted.

    5. Joe, thanks for the reply and the supplemental section. I wasn't claiming that EW doesn't work on 1 minute charts, or that you can't build counts from the smallest time frames up to the largest time frames. In fact, that's how I have always counted waves. However, the reason I asked about how Elliott did it, is because I often see people label waves on 5 minute, 15 minute, or hourly charts as ABCs, but then, later on, when the market does something contrary to what their count expected, they then change the count on the daily, weekly, or monthly, from a corrective wave to an impulse wave. In other words, they completely disregard all the work they had previously done in determining the internal subdivisions of the wave, and simply change it to what they "need" it to be to fit their newly adjusted daily, weekly, or monthly count.

      What we know as an absolute is that while "most" waves that meet the requirement of an impulse wave can also be counted as a corrective wave, the opposite is not true: Any wave that does not qualify as an impulse wave, must be counted as a corrective wave, and once that label has been determined, it can NEVER be changed to an impulse wave. So, because of that, I can't label the rally from 1810 to 2401 as any part of an impulse wave. As I posted here right after the market hit 2401 on March 1st, the only way for me to count the 14 month rally is as a WXY, and that is why I started counting 1-2-i-ii down from there. As an alternate count, given the extension in time since the March 1st high, it's also possible to count it as follows:
      abc (zigzag) from 1810-2111
      X (running) to 2084
      abc to 2401
      x to 2355
      abcde (non-limiting contracting triangle) to Thursday's high at 2361.
      If this count is correct, then the market is set up to literally crash from current levels.

    6. Well, then do me a favor. Bring up a daily chart of ES E-Mini futures from 11/01/2016 to 12/16/2016. Label the bottom of 11/09/2016, as 0. Then, label then pennant shape at the top of 2213.75, as wave 1 on 11/30/2016. Then, there are three candles down to 2179 on 12/05/2016 as a wave 2. Then, label the Fibonacci five green candles up to 2255.75 on 12/09/2016 as wave 3. Then there is only one red candle down to 2246.75 on 12/12/2016. And then only green candle up to 2273 on 12/13/2016.

      So, now you have five non-overlapping waves up. Wave 2 is only three candles and wave four only one candle. That is NOT the typical relationship, but they are both red, and neither are they "grossly" out of proportion maintaining a 1:3 relationship or about 33% to each other. Now notice how small the retracement level of wave 2 is. That tells you that the first wave is the 'extended' wave in the sequence, and all of the rest of the waves three and five are smaller than it. So, it all fits.

      Maybe you don't want to listen to this line of reasoning. Maybe you don't want to admit to being 'snookered' by the futures, or by an extended first wave. Oh, the futures? Yea. Just like Elliott didn't have computers, he didn't have to deal with the futures either - a point I have brought up in numerous videos until I'm almost sick of it. I can't change your mind for you. I can only provide my experience. Yep, you said Thursday that we would crash Friday too. We didn't. I'm not saying we can't go lower. The market seems to indicate that at the moment.

  2. TJ, as you know I'm not a true 'expert' on EW but I learned a lot from another site long before you starting publishing this blog (not OEW). While I favor your primary count, I still think Minor C of Int 1 is ongoing with minute 4 underway. Wave 4's are suppose to take up time and move prove price sideways. However, your minute 4 takes up half the time of the sharp minute 2. I expect minute 4 to be at least equal in time and possibly longer to have the 'right look'.

    1. Don't think so, because it is very hard to count the DOW as in anything other than a true Expanding Triangle at the location of minute ((iv)) of minor C, of Intermediate (1). If you find a way to count it as something different, please let me know.

  3. Amazing analysis Joe. Thanks for taking the time out of your busy day to post this..I will be joining the chat room soon

  4. Joe,
    Can you post a longer term chart from 2009 lows? Do you have a public chartlist like the other blogs you mention ?

    1. You can review my long term count in this YouTube video published long before the other sites got on board. (Copy and past link as needed).

    2. If the 1 of 5 ended at SPX 2401, how do you explain Elliott/Awesome oscillator making new highs ? Shouldnt it be lower then 3 of 3 ?

  5. Joe, I strongly appreciate your comments in the opening paragraph of this update. Totally on target. Thanks again for your work and sharing with others.