search elliott

  • Google

Enter your email address:

Delivered by FeedBurner


  • Where From?
    free counters
Related Posts with Thumbnails

« Disney Abandons Silicone Valley | Main | Are Stocks Fairly Valued? »

Tuesday, March 30, 2010


Feed You can follow this conversation by subscribing to the comment feed for this post.

Glenn Loser Neely


Here is a link that might be of your interest:

Please forgive Yelnick for not including anything for you on his post "Disney Abandons Silicone Valley". We all understand that you like "special boys' like Gleen Neely.



Yelnick et al;

I listened to Glenn Neely's 14 minute interview.

First thing that blew me away was the diatribe about how unpredictable the market has been for the last 6 months. What? For the last year the trendlines have been steady up. I see two retracements of limited significance, over that period, each only 38 % of the previous trend up. So I don't see how this holds water.

Then it was how the market is unpredictable, predictable, semi unpredictable, semi predictable, unpredictable to be predictable ...... sounded like a lot of excuses to me.

But what really blew me away was that the BIG TOP was in 2000, per Glenn. Bear market since. And of course there were many references to how accurate his calls were, during the 2000 plus down draft. Reminders I guess, as to how great he has been. See end of note.

If one looks at indices like the Trannies, the Dow Jones Comp, the NYA, the RUT, the Shanghai Comp, various MSCI Global Indices, the XAU, the HUI, the GOX, the XOI, various financials, how can all of this be so? B waves to heavan, everywhere? I do not think so. Things move in unison more often than not.

It makes absolutely no sense said top was a B wave, in all these indices. It makes even less sense the Indu and SPX are for some reason different than nearly all the rest.

Glenn Neely and Precther both in my humble opinion have been having such a hard time here because they seemingly have charted the move down from the late 2007 SPX top as a C wave, in Elliot parlance. Regardless of what Neely says, his background is Elliot wave. Since C waves channel nice and tight and are fear driven waves, they both anticipated a continuous drive to the bottom. It did not happen, not to their initial predictions in any event. Instead it appears imo what we have going on the Dow and the SPX from said top are still in an 'A' wave down, and assuming it is not complete we have a 'c' yet to go.

In brief the BIG TOP was in latter half of 2007.

To move on I personally do not think much if any tuning has to be done to classical Elliot. At one time I did. But I watched and studied all of Neely's chit for a long time, until I decided he was full of chit, or he is conducting a mind fock on all his fans (and subscribers). His Neowave methodology is pure garbage. He himself is consistently non consistent within his own Neowave rules, as to when he uses abc's, abcde's, abcdefg's ad infinitum. What he has done is develop some new sort of code, to code classical Elliot wave correctives in some 'Neely code', that he and only he understands (or thinks he does). All in an attempt to make it look like he stands alone, and above all the rest. Well Mr. Market recently took him down a peg or two.

In all cases that I followed in earnest I was always able to take his patterns and reduce them to a classical and correct Elliot wave structure. With absolutely no exceptions.

It is hard for me to believe anyone could make subscribers pay for something so deceitful. Yet I concluded this was the case, after watching Neely move after Neely move. And if I am so wrong, why is there no new book, yet in his communiques' and writings there is nothing but continued reference to his Elliot wave book, which is essentially classical? Why no new book? Someone as proud of his work as he leads on to be would surely do an update, with some explanation, no?

And all these big words like diametrics, reverse alteration, Neowave rules .... yada, yada, yada .... all classical symptoms of someone trying to buffalo his audience, and make himself sound like he knows all.

I would thus suggest to those that follow Neely (in particular Neely) and Precter ... beware. Wave analysis is a predictor/corrector method of analysis, and if one uses it to predict where the markets are headed, utmost dependency is required as to having an accurate assessment of where we have been. If one gets off a wave (of major degree), it can take a long time to get back in sync. I know from experience. I believe these two yoyo's are currently off, and that more danger lies ahead for those following them. Not immediately perhaps, but beyond.

Being a real LT bear has rarely paid off. Very rarely. Is this the time? What are the odds?

I am IT bearish, yes; but I think in the longer term this market could do a lot more damage to the Elliotician or in otherwords expert crowd. Out in the bush where I reside, street smarts counts for zip. The market imo is the greatest humbler of all, and as it marches forward it could humble all the great smarties, as well as their followers. The Market is the great equilizer.

In market parlance, one is only as good as their last call. No more no less.

The last calls of Neely and Prechter have not been very good.

What do I know? Not much maybe, but then I am not selling newsletters to paying customers, only writing on a free for all blog. Nor am I pumping anyone method(s), except perhaps their own. In this day and age it pays to be vigilant of all possibilities, and do your own homework. With so much information flying around, much of it free, it can be difficult to make proper assessments.



Yeah... to say that market would be unpredictable for the next couple of years is a spin on "My method is not working". It may have worked earlier, but now i guess its time to look at another "Guru".


SPX possibly tracing out a triangle, with more upside to come:


have done a daily update for the spx and asx for those interested... still sticking with my top call.


agree Ashish :-)

Glenn Loser Neely


We need help to understand what your boyfriend Neely tried to say with his Pee Wee voice in the interview.

Was he trying to excuse himself for his erratic, worthless and stupid method?

Why he did not mention anything about his gay relationship with you?

Tell your boyfriend that if he goes out of the closet he will speak better, just as your cousin Ricky Martin is now doing..




Your ability to speak in generalities without anything concrete to say is unparalleled. Your basic point appears to be "I think Neely is full of it" but your rationales are so subjective that they only represent some vague intuition that he is. And you call yourself a scientist?

As for the one semi-substantive point in your post, about the B-waves, it absolutely fits with Neely's long-term count on the US equity market, as detailed in the Appendix to MEW, with its enormous Running Correction wave-II, which set the stage for the strength in the wave-III and IV to follow. In that context, of course B-waves are going to extend over the tops of the A-waves which proceed them.

The Dow is completely useless from a wave standpoint and the S&P was overweight financials and technology during the big bull run-up, hence its lower peak.

In market parlance, one is only as good as their last call. No more no less.

The last calls of Neely and Prechter have not been very good.

Sounds like a recipe for chaos, as you chase "hot hands". Good luck with that.


For those who think that Neely is JUST trying to make excuses for bad calls (I agree there is that), take a look at this analysis of the level of "trending" and, hence, "predictability" between wave-(A) and wave-(B) markets. This is based on Daily data from the start of the bear market in September 2000, as Neely has dated it.

I normalized all of the data with regards to time and wave size, to compare apples to apples. Also, I normalized each wave with regard to direction, because my concern was whether or not there was a difference in the way waves (A) and (B) move from start to finish, not whether the waves were bull moves or bear moves. I then did a linear regression with percentage movement as the Y-axis and time as the X-axis. As anyone can see, wave-(A)s clearly exhibit more "trending" behavior than wave-(B)s, as shown by the ability to fit a straight line to their price movements (a high R-2 score). Even the one wave-(A) with a relatively low R-2 score is higher than the highest of the wave-(B)s. This absolutely provides quantitative and analytical support to Neely's statements about "predictability". That's not to say that Neely has figured out the right trading techniques for these two very different environments, because he obviously hasn't. That's what I've been working on and I think I've got it figured out.

Neely only gets you most of the way to being a good trader. You need to put in your own effort too, like I've done. I also did the same modeling for wave-(C) and they are even more "predictable" than wave-(A), which is why I say that when wave-(C) does start, Neely will once again likely be able to make very accurate trading recommendations. Wave-(E)s are also more "predictable" and wave-(D)s are less. Again, all of this is EXACTLY as Neely describes it, only I've added statistical validation to his point. Since I have a statistics background, I wanted to test his hypothesis. Again, these are not exact sciences so I don't know that these statistics prove the point to the 95th-percentile confidences level, but they provide a reasonable confirmation of Neely's basic point, which is that markets are sometimes more "predictable" and sometimes less.


NS--your post got me thinking. Appreciate that.

DG--I am struggling with my own question about whether too much "rational thought" is crowding out my "highest" level of thought--"rational plus" if you will. All great discoveries seem to go just a little beyond ordinary reason. Maybe great trading needs to as well.

So I observe with genuine interest your uber rationality. Have you really got it figured out? Was it always a question of "better horizontal rationality" or is there merit to my suspicion that some form of "vertical rationality", in the end, is needed. I look forward to your future posts, and finding out.


nspolar, beware you will anger the follower of Neely, the worse is Neely cannot even shut his mouth and he has to.... he really has to find an excuse to cover his inadequacy of his call. As always the one who claims who know a special method so arcane and so complicated only who is the person who understands it.... any challenges or failure in his method, they will put it up as a personal insult and they'll fail to stop justifying and stop self-indulging and keep forcing their count on the market...

If these people are really working in a hedge fund or trading as a pro, over last 6 month they are just counting their wave and still labeling and relabeling their count and cannot realize their call was wrong and have the flexibility to long something, wonder if they could keep their job, let alone if they keep shorting it. People will wonder if they are really trading for the money or indulging themselves that they know something so complex thus make them feel superior? How much S&P move is considered to be a good enough trade to change their mind?

Let they be our lamp post, maybe we can benefit from their dysfunctional inferior complex.


Bird, .... I think DG doesn't think horizontally.... or better... not even vertically.... he is thinking anally!



You know I read and enjoy your posts and I appreciate your efforts to bring some rational analysis to EW. You might want to re-think your A vs B analysis.

Linear regression is not an appropriate tool for time series analysis. Linear regression is used to find relationships between an independent variable (X) and a dependent variable (Y). The idea is to estimate the mean of Y conditioned on X. The pairs of values (X,Y) are supposed to be independent and identically distributed. The pairs are not independent: today's price clearly depends on yesterdays, and today can only follow yesterday, not any other day. The pairs are not identically distributed: the variance of prices that have advanced 30% will be, to a first approximation, 30% larger.

So, how better might your analysis be done? Take a pair of waves, A and B. Calculate the changes in price each day during the A wave and during the B wave. Calculate the and standard deviation for each wave. Your null hypothesis is that the absolute value of the mean daily price change is same for both waves. Then use the standard deviations to see if the difference is significant. You use the absolute values of the means because you are testing magnitude not direction.

Finally, since you have six or eight pairs of (A,B) you could try regressing the absolute values of the means of their daily price moves to see if there is a consistent relationship between the strength of the A wave and the B wave to follw - then you might have a nice predictive or confirmatory tool!

A further analysis might involve looking at the mean and variance of the absolute value of the daily price changes (rather than the absolute value of the mean): are A waves more or less volatile than B waves? Maybe B waves fluctuate more on a daily basis but make less overall progress.

Have fun!


Anal's good. (That's NOT how I meant it. Get your mind out of the gutter.) So long as it works.


Correction: 3rd para should read "Calculate the MEAN and standard ...."



I do think that one needs to completely exhaust the resources and techniques of "horizontal rationality" before any attempt at applying "vertical rationality". Or, perhaps the "vertical rationality" is at the front end of the process, with the formulation of the hypothesis that markets are not chaotic but orderly and that's as far as "vertical rationality" can take you. One then goes about determining exactly what that order consists of using "horizontal rationality". My "intuition" about wave-(B)s was confirmed by the regression analysis. When those are aligned, I am more comfortable moving forward with a technique.

I wish I had more Hourly data to backtest the trading methods I designed for "wave-(B)" markets, but I only have the last year. Had I been using my methods over that time, my gains trading the SPY would have been 37.4% with 1% at-risk on each trade. When you account for actual time in the market (there were times when there was no setup to trade), it works out to a 90% annualized rate, with 1% at-risk. 62% of trades would have been winning trades and the winner-loser ratio would have been 1.9 to 1. I'm also beating the market since I started using it in mid-January.

I've seen the same trading recommendations others have from Neely over the past year. Others look at the outcomes and say it proves that NeoWave doesn't work. I look at the outcomes and say "What if I tweak this a bit and do things this way instead of that way when the market is less predictable?" It's just two different approaches to the same problem.


"If these people are really working in a hedge fund or trading as a pro, over last 6 months they are just counting their waves and still labeling and relabeling their count and cannot realize their call was wrong and have the flexibility to long something, wonder if they could keep their job, let alone if they keep shorting it. People will wonder if they are really trading for the money or indulging themselves that they know something so complex thus make them feel superior? How much S&P move is considered to be a good enough trade to change their mind?"


These are ALL excellent points.

To try and claim (as Neely has) that the market has been UNPREDICTABLE over the last 6 month's is pure denial of just how strong the upward trend of the market has been and what the equity market has accomplished in the face of so many naysayers.

The Neely's of the World are in total denial of reality. They cling ever so stubbornly to their methodology, and then when the MARKET has proven that such methodology has gone terribly wrong, they come back with "secret" rules and marketing schemes like Neely's REVOLUTIONARY "River Technology".

Let's face it.
This guy (and others like him such as Prechter) couldn't trade their way out of a paper bag. They are not Traders. They are newsletter writers.
Plain and simple.



The pairs are not independent: today's price clearly depends on yesterdays, and today can only follow yesterday, not any other day.

This is why I converted price to percentage movements, to normalize. That's why the Y-axis always starts at 0% and goes to 100%, except some waves where the movement between the start and the end of the wave went beyond the price extremes at the start and the end. I don't show the X-axis to reduce clutter, but it would also run from 0% to 100%, so that regardelss of whether a wave took 25 trading days or 250 trading days, I'm regressing on the percentage of time taken to complete the wave and how linearly the wave progresses through time. The point I was getting at is that wave-(A) markets push ahead without "backing and filling" whereas wave-(B) markets require a significant amount of "backing and filling". This squares with Neely's high-level observation that wave-(B) is less predictable on a day to day basis.

Today's price is dependent on yesterday's price, yes, but today's percentage movement is not, nor is the propensity of counter-trend movements to persist dependent on yesterday's price direction, so I don't think your criticism of using a linear model holds. The hypothesis is that "mass psychology" is more "uncertain" during wave-(B) markets and hence the ability of the market to progress in a straight line through time is diminished relative to wave-(A) markets.

At least you are bringing something substantive, though, and I appreciate that.


This guy (and others like him such as Prechter) couldn't trade their way out of a paper bag. They are not Traders. They are newsletter writers.
Plain and simple.

And yet, as I pointed out the other day, Neely still has a better long-term trading track record than CARL FUTIA.

I'm still laughing over that one.

Trading success is a marathon, not a sprint.


The point I was getting at is that wave-(A) markets push ahead without "backing and filling" whereas wave-(B) markets require a significant amount of "backing and filling". This squares with Neely's high-level observation that wave-(B) is less predictable on a day to day basis.

Another way to say it is that wave-(A) oscillates much more tightly around the fitted line derived from the regression equation, while wave-(B) oscillates in a much wider range around that line.


nspolar, you raise an interesting question of whether 2007 is an expanded (irregular) B wave or a wave 5 top. The Dow certainly went higher, although not outside what an expanded flat does; the S&P barely crested 2000; and the Naz remained 50% below 2000. In Euro terms, the Dow did not get to a new high; in gold terms it also did not. So it is not as if there was an across the board confirmation of a top. Instead it is arguable that what really happened was a drop in the USD due to the credit bubble, not a rise of assets in Dollar terms.

There has been a 17 +/- cycle of tops/bottoms since 1949 in constant dollar terms: 49/66/82/00. It targets 2017 +/- as the bottom. Now, cycles for the most part come and go and are less reliable than other technical indicators, other than the 4-yr Presidential cycle which is based on real intervention the the economy. Some people argue 99 as the real top, not 2000, and target 2014 as the next low, also within the four-yr presidential cycle. Whatever. But this 17 yr cycle makes 2000 more compelling than 2007. Not convincing, just another point in favor.

I ponder whether we have things off because we see 1929 in isolation. There was an auto bubble from 1915-19 that was strikingly like the Internet bubble of 1995-00. The Dow fell 72% off the top. Then we had the roaring twenties and the huge credit bubble into 1929. This would support a different analogy: 2000=1919 and 2008 = 1929. Auto Bubble = Dot-Com Bubble. Housing Bubble = Roaring 20s bubble. Your 2007 becomes THE top.

In a linear (arithmetic scale), 1929 looks so huge, as does 2000, that this view seems off. But it is better over longer time scales to use log/percent scales to see the market. Going from 50-100 is the same as 5K to 10K, and looks scaled the same in a semilog chart, but looks hugely different in arithmetic scale. This 1919=2000 and 1929=2008 (when the crash occurred) looks much closer in semilog scale.

Whichever it is will be unveiled over the next few years, say into 2014.

I think the Hope Rally was predictable to at least go 50-62% up, especially as the 2008 drop fulfilled five waves down, hence wasn't a wave 4 blip. So now past 50% and approaching 62%, the interesting question is what is up if we pass 62%. Waves 2 can do that but seldom do.

If 2000 was the top, and 2008 a C leg of an expanded flat, we could be done and in a new bull marekt (Tony C's view); or could have more to go in the C (Prechter's P2 view); or could be in a complex correction (my view). This Hope Rally may be an X instead of P2, or a smaller degree B of a zigzag wave. We don't know yet.

At the scale of a 50%+ retrace, it was quite predictable. Going to 62%, no surprise. If it passes that level (say Sp1250 for a little slack), I would say we are at a point of maximum entropy, meaning we could no longer predict whether it is a new bull market, a P2, or X or B of a triangle leg 3. At that scale the market is unpredictable.



I am not sure you are getting my point. By using the 0 to 100% scale you are not addressing any of the issues I raise - you have simply scaled the price move to 100 units and shifted the start price to zero. i.e. Pt%=(P0-Pt)/(P0-PT) where Pt%= price in % of move at time t, Pt= price at time t, P0 is start price and PT is end price.

The idea of "independent and identically distributed" is that today's price change can be approximated as a stochastic process with a given mean and standard deviation. i.e. if the mean is +0.1% then todays price change is just as likely to be 1.1% as it is to be -0.9%, and this will be true for every data point in the series from start to end.

Today's price cannot be so approximated. Today's price depends on yesterday's price (serial dependence) and the confidence interval also depends upon yesterday's price (heteroscedasticity).

I think your intuition is correct - your charts show it very nicely. The analysis I suggested will support your hypothesis and avoids the statistical glitches I have pointed out.

I would also add that I have had a very hard time getting a definitive read on whether linear regression has a valid place in time series analysis.


Yelnick, yes there are many possibilities.

A subtle point here is linear scales or semi-log (log price scale with lineat time scale).

At one point in time I was a linear fellow, rarely used anything else. I have changed my view. Am I now a semi-log fellow? No. I now use both. Why?

I use both because wave analysis is so subjective. Impulse waves for example are supposed to show power. It is very informative then to look at impulse waves in both methods, and subjectively view the difference. It makes you think about it more, which is good. The semi-log approach will inform one quicker I think as to when the impulse is changing, and losing power.

My suggestion to you is to look at thinks like the RUT, the DOW, the SPX, the Trannies, Gold, the HUI, the XOI, the NYA composite, Shanghai Comp, MSCI global, etc ... and flip back and forth as you do from linear to semi-log. It helps to have a good system to do this in, to minimize the time between views.

You should as well go back to the base of the big move up.

Now the other thing here that comes to mind is a unified wave theory of sorts. Is it possible for all these different indices to be on different waves, or must there be some unification?

After you do this, come back and let me know if you have seen anything that now looks a bit different than it perhaps did before.

Thanks if you would, I would appreciate it.



I think your intuition is correct - your charts show it very nicely. The analysis I suggested will support your hypothesis and avoids the statistical glitches I have pointed out.

If this had not been the case, I might have tried other ideas.

I just went an looked at the mean and standard deviation. For (A)s it's 0.28% and 1.54%(using the scaled 0%-100% range) and for (B)s it's 0.18% and 1.47%, so the S.D. for (A)s is ~5.5X the mean and for (B)s it's ~8.1X the mean, which says to me there is much more "noise" around the central directional tendency in a wave-(B) than in a wave-(A).


By the way, as I delved more into this, I began to think of this type of analysis as the application of the idea of "market regimes" to Elliott Wave and the need to tailor wave-based trading strategies to the two identifiable regimes ("predictable" vs. "less-predictable"). Ultimately what came out of this was a variation on Neely's trading methods which is aligned better with the characteristics of "less-predictable" markets.

One of the blogs I discovered during this process gave me a lot of conceptual "food for thought" in this area.

There are probably optimizations I will find over time, but for now I am using these concepts to guide my selection of trading strategy.


nspolar, good chart of the gold-dow over at Prgmatic Capitlaist.


JT...., there are lots of EW and MEW enthusiast here.... and some are on the edge to defend Neely or Prechter at all cost. Yet doesn't matter how complex and simple the trading method is, they eventually fail, like trend following method, breakout, counter trend, buy the dips... so why not MEW?

The worst part of Neely is as a seasoned and well experience analyst (and someone here thinks Neely is an experience market participant as well), I have to question why he needs to justify his failure so publicly. Maybe he tries to save his newly launched fund and hopefully his bad call won't scare off any more investors? Or maybe the feeder fund is already disappointed enough to cut him lose...

The problem I have with lots of these statistician or MEW expert here is they kept on and on about showing their craft and its complexity, yet no one comes out and say... I have made 100 million last 5 years... or a few millions over last year using the method (employ by any institution).... its that simple, results count in the money management business, not just methods.

As I said before I have question if these people is working or have worked in any financial institution before or they just trade for themselves with minimum capital (less than a few million dollars)... yet they way they talk in such authority and thought the method has led them to certainty and think by looking backward (data fitting) will allow them to predict anything... like Neely's denial... reflect their inadequacy and ignorance.

For one thing, I'll remind my hedge fund manager friends beware not to hire self-proclaim MEW expert as they cannot flip out from their obsession in counting or show-off, so don't count them to deliver any return by each month end.



Are you saying that I did "data fitting" to come to my conclusions?

vipul garg

how much money do your hedge fund managers manage that they will accept all and such opinions of yours.


Vipul... most of my friends have upward 500 million to 1 billion in size, consider to be a small potato in the field.

Yet, with all their quantitative guys (yes they have people know elliot stuff), its not like they'll keep justifying and backtesting and still justifying and backtesting an opinion or theory.

In reality they don't need me to remind them anything who to hire, I don't think they will anyways.


DG... I lost appetite to speak with genius.... after all I am just a low level farmer..... btw some farmer in china who made few millions dollars farming garlic last year, if the news is correct 500% gain in one year.

So how much dough you pull in last 12 months? 5 million? 10 million?



Are you accusing me of "data fitting" or not? It's a simple "yes or no" question. If you are, what is your proof?

The comments to this entry are closed.