search elliott

  • Google

Enter your email address:

Delivered by FeedBurner


  • Where From?
    free counters
Related Posts with Thumbnails

« Greece Going Lehman On Us UPDATED | Main | Lazy Trader Market Playbook »

Thursday, April 08, 2010


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


Great post Yelnick - very helpful. PS! - has anyone ever told you that you sound like Dustin Hoffman? :)


Given that this rally has been rising on lackluster volume, such a strong wave is less likely than the correction.

Careful, Yelnick, the "volume is not predictive" bulls will flog you for this statement. Although, in this case, I have to agree. I don't think anyone has a clear handle on what "real" volume is in the absence of the kinds of leverage that were standard during the run up from 2003 to 2007. I do think that the increases in volume during declines is "suggestive" of the notion that we're in a bear market rally, rather than a new bull market.


We are in a big A-B-C wave rebound since 2007. The complete of wave 5 now (or anytime soon) is the complete of wave A. Then we should enter a wave B correction (or consolidation). After that, it will be another melt up for wave C to be finished in late 2011 / early 2012, just in time for the 2012 US election and the new leadership in China. Also, just in time for the too-big-to-fall US to default its loans.

It does not matter whether the peak of wave C will surpass the peak of 2007. There is enough money to be made during the fluctuations.


Great post indeed! It's very helpful to analyse different opinions and highlighting the most probable outcome. Both Caldaro and Prechter are highly respectable. Yelnick adds the "finishing touch". Thanks. Your posts are very helpfull and interesting even here in Belgium. Good luck!

Mamma Boom Boom

>The broader question is whether we are in large corrective pattern, .....or a new bull market that will shock & awe the bears.<

Depends on definitions: Everyone thought 2003-2007 was a bull market, but it slammed the doggie do-do right out of them.

"Ha shut up, silly woman."
Said that reptile with a grin.
"You knew darn well I was a snake before you took me in."


Will be interesting to see if the market players are as keen to fill the gaps below the market as they are to fill any of these opening gaps. If so, the March 2009 lows won't hold or will barely hold because there's a gap on the 60 minute all the way down there.

vipul garg

if bull mkt means a new all time high , then its not one.
if bull mkt means not breaking the lows , then it is one.

in neelys defense i would say that he is almost alone ,in the e wave community in public domain, when it comes to figuring out the long term position of the market.
the current structure he has on spx makes a whole lot of sense and with maybe one final revision he will get it spot on.

trading is a different beast altogether though and this is where neely like almost every trader out there is trying to do on continous basis: improve his skill set.

Mamma Boom Boom

>he will get it spot on<

I'm sure he can use the publicity. He has had to of lost a lot of his base, this past year. A blunder like that can just about wipe out your business.

"Time is on his side"


So who is going to buy NBG?


"We are in a big A-B-C wave rebound since 2007. The complete of wave 5 now (or anytime soon) is the complete of wave A. "

I think we are seeing the same thing.


What a crock of chit from the boys at citi today.

What a complete waste of time.

No one could see that lending 500 k$ to someone that couldn't speak English, let alone write it with no full time job and nothing down was a bad bet.

Think about that when someone holds their nose and tells you they have no choice but to means test SS.

You are what you accept.



A corrective pattern insinuates a resumption of the downtrend that started in Summer of 2007. This puts objective of making a new low which is unfathomable for many reasons. Albeit the March 2009 low undercutting 2003's low, we are in a secular sidways market that started in 2000. The decade long top and bottom ranges very evident. We are in a cyclical bull market which we can set an objective that secular forces will change the tide. EW analysis might be good to establish where this roadblock will happen. I am probably saying the same thing as other but just with different perspective.


Mark, a major down move does not require breaking the Mar09 lows. The wave pattern lends it self to three macro level interpretations:

New Bull: 2000-2009 was an ABC flat correction of 1982-2000, and we have begun a new wave up to higher highs

Sideways Bear: The Hope Rally is a large X wave connecting us to a second flat correction soon. Alternatively, it could be the B of a large 3-wave off 2007 as part of a big triangle. In both cases it need not break Mar09 levels

Big Bad Bear: We are nearing the end of P2 and are about to commence P3 (with P4 and P5 to follow) to break below the March lows. Note that we cannot fall the same distance again or we go below 0, but the better practice is to use semilog scales and talk of percent drops. P3 is expected to exceed P1 in percentage down from wherever P2 ends.

Calling P2 or X a cyclical bull makes sense.

If we are in the sideways secular bear, it likely is a big wave 4 with a nice wave 5 to follow. One issue with the Hope Rally as that wave 5 is it has come pretty fast given the prior waves 123 off 1949 all went around 17 years.


Cloudslicer, if you label the Hope Rally as a big X I am with you that this is a very plausible option. Your A broke as a big flat, ending with a five-wave move to Mar09. If what follows is a big B of a larger flat, it usually would last longer than A, putting your whole pattern out a bit farther, and it would make the whole flat very irregular, in that the A wave broke above the start, not the B wave as in an expanded flat. I would suggest you consider an alternative structure, if you keep to a B not an X wave, of a large triangle.


Wrong wrong wrong. The comments made in this video said that the waves did NOT alternate, THEY DID, the first was a double zigzag, the second was a single zigzag....all you need to do is look at the internals to see this!!!!


Dingo, a double vs a single zigzag is not enough for alternation. Both are sharp corrections. Both took about the same time an went about the same distance. You need a larger difference, such as one being slower, or breaking sideways.

Leslie Martin

Does anyone know of a statistical analysis that would have a bearing on whether four-two alternation has any more than the authority of reason behind it? Notice how that authority is continually nibbled away at its fractal edge by the tides of empiricism, as in "It is reasonable to assume that the earth is flat." Hasn't EWT now reached the stage of claiming too much on the back of some rather large untested assumptions? I mean, it's fun, but who really cares how many angels can dance on the head of a pin?



Great question! I've been testing of some of Neely's claims regarding "predictable" and "unpredictable" market environments because I think that's something a bit more straightforward to test than some of the other claims, plus it has been front and center in regards to trading performance. Some of the more abstract claims like alternation, which could encompass any number of types of alternation, would be harder to test, but may be less important for trading than the "predictable"/"unpredictable" dichotomy.

I don't know if you've read the book "Evidence-Based TA", but that's what they try to do. In fact, they don't even test Elliott Wave precisely because it has traditionally been so loosely-defined:

"EBTA rejects all subjective, interpretive methods of Technical Analysis as worse than wrong, because they are untestable. Thus classical chart patterns, Fibonacci based analysis, Elliott Waves and a host of other ill defined methods are rejected by EBTA. Yet there are numerous practitioners who believe strongly that these methods are not only real but effective. How can this be? Here, EBTA relies on the findings of cognitive psychology to explain how erroneous beliefs arise and thrive despite the lack of valid evidence or even in the face of contrary evidence. Cognitive psychologists have identified various illusions and biases, such as the confirmation bias, illusory correlations, hindsight bias, etc. that explain these erroneous beliefs."

When people say that Neely's rules are "too rigid", I am reminded of this paragraph because it makes clear that without something as rigid or more as Neely's rules are, Elliott Wave won't even be considered anything but "subjective" and "interpretive".

Leslie Martin

Hi DG -
Thanks for the link. I had come across this approach. Your reminder has dislodged it from the sediment :)
As regards Neely, I would say:
1)that predictability" v. "unpredictability" in the market is fundamentally a matter of degree: someone working on a very short timescale will be satisfied with a degree of predictability absent from the focus of someone who is looking longer term;
2) it seems to me that the approach he takes in MEW is to set up the rules prior to examining the data, and therefore he remains tied to the unexamined cognitive biases to which you refer, because it's the actual seeing of the patterns that is subject to cognitive bias, and to attempt to explain cognitive bias by using the patterns (a la Prechter) is simply putting the cart before the horse.
So, if there is an objectve way of deriving the patterns (IS that taken care of by the data mining approach?)... well, that would be a whole different ball game.



1) There is the matter of timescale, but I've also found evidence that Elliott Wave Progress Labels exhibit measurable differences of "predictability", with predictability defined as a the probability that time t+1 to t+2 unit of time measurement will exhibit the same directional movement as time t+0 to t+1. The directions are more likely to be the same during specific Elliott Wave labels.

2) In my opinion, ALL approaches to the financial markets remain mired in epistemological darkness, yet presume that their initial assumptions are correct without empirical proof. So, I would say that just as one may not be able to derive Elliott Wave patterns from the data, nor can one derive a "random walk" from the data. One of the studies I read claimed that if "random walk" were true, we'd see at least one equity market reach a price of infinity, for all practical purposes. By deduction, since we don't see prices reach infinity, we can claim that the evidence against the "random walk" is more compelling than the evidence for it. What then remains is to determine the exact nature of the order which informs markets, as well as to determine if that order is of a constant nature or if it itself has some measure of variability.

Also, what Neely has done which is different from other wave theorists is that he has required that "post-pattern behavior" comport with the logical implications of the pattern one has identified. This "confirmatory" stage is completely absent in original Elliott Wave theory. Thus, yes, our cognitive bias can lead us to see a pattern which "isn't there" in reality, but if the behavior of the market after that pattern doesn't comply with expectations, at least we have a mechanism to now rethink our original interpretation. Neely calls this phenomenon of patterns that aren't "real" but seem to be "Emulation" and it is part and parcel of the uncertainty inherent in any market analysis. We are most definitely NOT in the realm of classical physics here!

Leslie Martin

"What then remains is to determine the exact nature of the order which informs markets, as well as to determine if that order is of a constant nature or if it itself has some measure of variability."
I agree completely, DG.
As to the epistemology - I think we share a desire to clarify the basis on which it is claimed to have knowledge about market patterns. There is a book title that seems to be haunting me lately: in translation, "Knowledge and Human Interest" (Jurgen Habermas).
Maybe I ought to read it now :)
Re timescale: I think you have a better understanding of Neely's work than I do. I don't quite get the exact sense of your first point. But regarding what I was trying to convey in my previous post, I should really have said something about the way that predictability waxes and wanes on a timescale several degrees smaller than the fractal "context" within which events in the short timescale happen. Somehow that "pinch point" (itself a moveable feast) has to feature in a formal account.



In the fundamental dualism of a deterministic universe and a probabilistic universe, put me down for a probabilistic universe, even if the shape of the distribution is unknown. I agree with the "long tails" thesis. For example, my trading method has an average holding time of just over three days with a standard deviation of about 2.5 days, but I'm in a trade right now that's been open for 20 days with no real end in sight, making that about an ~7-sigma event.

My point on the variation in predictability is that Neely seems to make it solely a function of time-frame reduction (e.g., trying to predict the next hour is less fruitful than trying to predict the next month), whereas my analysis preliminarily shows that the ability to predict the next month is also subject to fluctuations, depending on which "kind" of wave you are in and that the "kind" of wave you're in is directly correlated to the Elliott Wave label (assuming your label is correctly applied). More simply, a prediction about the direction for the next month is more accurate when you are in wave-(A) than when you are in wave-(B). Neely has been saying this for a while but I tried to go the next step and actually show it analytically.

Of course, the assumption underlying Elliott Wave is that the cause of all of these things is mass psychology. Either you agree with that causality or you don't.

Leslie Martin

Thanks for that clarification, DG.
It does make intuitive sense to suppose that the predictive nature of one's interpretation is as fractally patterned as the data from which one derives the Elliott waves.
This analogy springs to mind - that the pattern is to the predictability as the first derivative is to the second derivative in calculus. But that's just an analogy and, respecting your observation re physics, most likely a mightily misleading one. After all, chaos theory is supposed to get us beyond that bind.
Have you had a look at Rich Swannell's attempts to quantify the process of interpreting market action in EW terms?
Invoking "the psychology of the markets" on the basis of the fact that they are fractally patterned is, for a purist like me, a bit like invoking God on the basis that the universe seems to have been "designed": for me it is enough to posit random fluctuations, a selection process, persistence, and iteration. The exact same process in fact as Darwin conceived of as explaining biological evolution will suffice to account not only for trading recursions but also the very psychological states with which they correlate.


"This analogy springs to mind - that the pattern is to the predictability as the first derivative is to the second derivative in calculus. But that's just an analogy and, respecting your observation re physics, most likely a mightily misleading one. After all, chaos theory is supposed to get us beyond that bind."

Another way to analogize it is to think of a sentence which is complete in itself, but allows for additional words to be added, which could shift the meaning slightly or substantially.

E.g. "The ball was in the yard" is perfectly acceptable on its own, but compared to "The ball was in the yard until someone came and stole it", the meaning is very different. When we first think a pattern is complete, it is like reading the first sentence. Then, as the market continues and doesn't react the way we expect due to the logical implications of the pattern we thought was complete, we have to admit that the second sentence is the more accurate.

"Have you had a look at Rich Swannell's attempts to quantify the process of interpreting market action in EW terms?"

I hadn't until now. A quick Google search turns up some "interesting" stories. Still, I think that is a promising approach, if executed correctly. If Neely is right, software and automation don't get you all the way there and there is always a need for human judgment.

"The exact same process in fact as Darwin conceived of as explaining biological evolution will suffice to account not only for trading recursions but also the very psychological states with which they correlate."

I agree that the markets exhibit the same characteristics as Darwinian evolution. This is one of the reasons why I am not as offended by Neely's idea that new pattern structures can emerge which were never seen by Elliott when he came up with the originals. With one caveat, which is that I don't think that any of the new patterns which could emerge would be Impulse patterns, they would all have to be Corrective patterns. The rules of Impulse construction have to be the "unmoved mover" in the entire system or it breaks down into chaos and back into subjectivity, in my opinion.

Leslie Martin

Hey, DG -
Been away and just caught up with your latest post here.
"...or it breaks down into chaos and back into subjectivity..."
I agree completely, and I think that's exactly where we are. I have seen uncorrected fives entirely retraced, and once you've seen that I think that you have to conclude that the "rules" simply cannot be formulated as though someone had access to the pattern emanating from the beat of the butterfly's wing in such a way as to permit classical mathematics to trump chaos theory.



One of the difficult things to do when counting waves is to let go of an interpretation which has served you well, whether it has done so for an hour, a day, a week, etc.

A :5 which is not the end of a pattern should never get fully-retraced. What needs to happen if it does is that the wave counter needs to reassess what the pattern which looked like a :5 really was.

The rules are just the abstract logic of wave theory and they can't correct the perception of the wave analyst nor overcome the analyst's bias, nor can they provide anything more than a snapshot perspective of what should happen next.

Leslie Martin

Hi, DG -
Wrt my own attitude: I've got no probs whatsoever in jettisoning a pattern: NONE of them serve me well; it's just the way I am. So we can discount that as a possible source of bias.
My point is that you can't set rules for a contingent phenomenon which are not themselves subject to contingency, and therefore, you have to conclude that no such rules (as Neely attempts to describe them) exist.
This, incidentally, is the source of all these so-called "new patterns". A robust fractal (Prechter's term), in my opinion, defies such description. This is one of the fundamental tenets of chaos theory as I understand it. It is what makes "Fooled by Randomness" (Taleb) the antidote to such prescriptive approaches to the market's patterns.
To do anything else is (as our friend Yelnick might say) simply ptolemaic. Strangely, Neely does not seem to appreciate this. But then both he and Prechter, along with many others, have commercial axes to grind.



There was recently an experiment at MIT in pattern recognition using stock charts. Some of the charts were randomly generated and some were actual stock charts. Turns out that those who were actually good at pattern recognition could distinguish between the real charts and the randomly generated ones with some degree of statistical significance. Does that "prove" that patterns in charts are real? No, but it does suggest it and it suggests that recognizing them is a skill one either has or doesn't have, to some extent. Perhaps Taleb is just the world's worst chart reader?

As for the emergence of "new patterns", it's not simply a matter of the failure of contingent rules for contingent phenomena, it's an emergence of new behaviors to fill niches in price/time behavior which were previously unfilled, as Neely explained in this week's "Question of the Week".


In past updates I've seen you reference waves-F and G. Is there a limit to this wave "alphabet"? Why not just begin anew with a-b-c?


Unlike orthodox Elliott Wave, which allows great variance in the behavior, relationships and time consumption of specific patterns, NEoWave imposes strict limits. Accurate forecasting can only be accomplished when precise rules relating to price, time, complexity, behavior and relationships are required of all waves in all patterns.

An important part of the construction of 3-legged patterns (Flats and Zigzags) is that clear alternation exists between the 3 segments on a time, complexity and behavior basis. Since larger degree patterns (under NEoWave) impose maximum time restrictions on smaller degree patterns, a developing (smaller degree) correction only has so much time to unfold. If a correction is allowed no more than 10 days to form (based on the next larger degree) and the market wants to form a Flat, there can be great variance between waves A, B and C in TIME during period. For example, wave-A might take 1 day while wave-B takes 6 days, leaving 3 days for wave-C. This also follows the necessary TIME alternation required in Flats.

On the other hand, under the same 10-day time restriction, if a market wants to form a 7-legged formation (such as a NEoWave Diametric), that leaves little room for time variance between waves. Waves-A and B might take 1 day, while C, D and E take 2-days, which only leaves 1-day for waves-F and G. As a result, the more waves a pattern contains, the more similar each must be to fit within the time limits imposed by the larger degree pattern. From personal experience, the limit to the wave "alphabet" is A-B-C-D-E-F-G-H-I, which forms a NEoWave Symmetrical.

The reason waves-F and G (and sometimes H and I) must be used to label some wave patterns is because specific time and behavior characteristics are required of Flats and Zigzags that are not present in NEoWave Diametrics and Symmetricals. Time relationships during corrections are your primary tool for determining the kind of correction unfolding. If the first 3 waves of a correction are very different in time, a Flat or Zigzag is forming. As time between the waves becomes slightly more similar, a Triangle is the next best choice. As the time between waves becomes very similar (but price differences still exist), a Diametric is your best option. If the waves are very similar in time and also very similar in price, a NEoWave Symmetrical is your only choice.

Because all corrections possess a specific "time behavior stamp," that is the reason you cannot simply start a new ABC when you become uncomfortable with the continuation of the correction. A completed ABC Flat or Zigzag must possess clear alternation in time between all three waves. If that time alternation is not present, a Flat or Zigzag is NOT forming, but a more complex formation that requires the use of D and E, sometimes F and G and in rare cases, H and I.

Leslie Martin

Interested to get the reference for that MIT study if you can lay your hands on it.
As I said, I defer to your superior knowledge of Neely's work. But (but, but, but, but...) the account he gives does sound rather ptolemaic to me.
I don't question that regularities exist: what I question is the way they are accounted for. Many accounts may exist of the same phenomenon (especially when that phenomenon is a social phenomenon), but there can be only one scientific account, and that does not necessarily lend itself to making money in the markets.


There is always going to be something "heuristic" about any trading method. Even professional money managers only get measured to their benchmark index, not to "perfect" trading and capturing each tick of the entire market fractal during the relevant time period.

Also, since risk management is where most of the variation in trader returns comes from, according to what I've read, the "ideal" trading method is the one with the best risk management features, perhaps a better way of comparing methods is in how much of the initial risk of a position the system is able to eliminate for a given amount of time and then how that risk elimination relates to subsequent price movements, i.e. did you get a "false positive" signal to move your stop causing your overall profits to be non-optimal. I would assume that every system does this to some extent, but an objective measure of the actual extent could be a useful comparison metric. Perhaps one system's risk management techniques enables it to capture 90% of the relevant profit opportunities which ensue from its entry points, but another only enables 30% of the opportunities.

I think it's important to think creatively about what the right ways to distinguish skill from luck are in trading, precisely because so much of what markets do seems random to very intelligent people.

Leslie Martin

" much of what markets do seems random..."
But isn't the definition of a fractal that it is randomly patterned?
This isn't just semantics and sophistry: it's the difference between science and "Ok, I can live with that".
I take your point about trading though: money management is probably 90% of the game: "Way of the Turtle" by Curtis Faith illustrates this really well.


"This isn't just semantics and sophistry: it's the difference between science and "Ok, I can live with that"."

My perspective is that we will never know. Have you ever heard the term "mutually exclusive, collectively exhaustive"? Neely's rules come closest to that ideal, in my experience, so that's what I like about them.

Of course, the great line in Reminiscences is "You can't know until you bet!", right?

Leslie Martin

Yes, that has to be taken into consideration in a properly scientific account of social interaction. In other words, it does not represent a full stop but rather a datum.

The comments to this entry are closed.