Topic: Elliott Wave Analysis - page 4. (Read 20044 times)

Hmm... So, while valid Elliott waves, my count present invalid "Swannell waves". What could that mean? Well if I respectfully favour the late Swannell (September 2011) over the late Elliott (January 1948), I'd have to squeeze the first two weeks of November into a fourth wave. Some of my previous counts actually do that: Remove black I-II; Move magenta iv to where black II was; Replace black III with magenta v, completing the first black I at $2.1 on 15 November.




I'd probably change the colours (black becomes cyan, magenta becomes black) to stay consistent with higher scale charts. I'd also probably favour a final zig-zag pattern cyan 5-3-5 until the very bottom of major blue wave II (seen far above in this thread).

What difference would that make? Not a whole lot, except more extreme price ranges. Old black IV (now 'a') would be part of a new bigger second wave, that could find its peak in the price range of the previous grey iv triangle, $2.8.

Is $2.8+ likely? So far it doesn't look likely (I still have my stops [between $2.6 and] above $2.80). The price movement since 15 November looks fairly sideways. In fact, the drop two days later slightly below $2 seems to cancel out any potential sharp or zig-zag corrective patterns in the near future.


I'd like to stay on the "Elliott Education" topic, as there are already numerous threads discussing other technical, fundamental and seat-of-the-pants nonalysis. I'll respond from my/an Elliott perspective.

"Wave two corrects wave one... Typically, the news is still bad. As prices retest the prior low, bearish sentiment quickly builds, and "the crowd" haughtily reminds all that the bear market is still deeply ensconced. Still, some positive signs appear for those who are looking: volume should be lower during wave two than during wave one, prices usually do not retrace more than 61.8%... of the wave one gains, and prices should fall in a three wave pattern. ... As wave three starts, the news is probably still bearish, and most market players remain negative; but by wave three's midpoint, "the crowd" will often join the new bullish trend. Wave three often extends wave one by a ratio of 1.618:1." (Wikipedia)

Since S3052's double three suggestion, and my later triple three combined correction, I've held the belief that any indication of a 'quadruple' combination signals the death of bitcoin. I personally am not so morbidly bearish. I would buy like mad at $0.6.

Yes, further than 0.6.

The reason being the diminishing incentive to use it as a currency the less stable the price is.
I know these arguments: If it drops to (insert any low price) someone could buy up all the bitcoins in existence, etc..

My point being: They wouldn't want to. At a price of 0.1 - 0.2 USD all bitcoins ever to exist would still be worth 2.1 - 4.2 million USD. (less half on that with the ones currently issued)
And while there are individuals which could buy alot of coins (some might) at this point it will be questionable if this is a wise investment. (While in fact it is not, it is just that bubbles can occur in 2 directions)
Furthermore if you view bitcoin as a currency there would have to be much much wider use and a much bigger economy behind it. Statements that bitcoin could be just "an investment" because it is "scarce" are delusional IMO. Bitcoin needs the economy behind it which was the point to speculate which it in the first place, and I think to value this economy around above figure right now would be a pretty good estimate.

I even think that the economy behind bitcoin is smaller than that yet, so that figure factors in a (healthy) amount of speculation.


There are 2 different Schools of Thought at work here, lets see which one will turn out to be correct.

Your subjective objective was obvious. Do not confuse objectivity with falsifiability nor subjectivity with lack thereof. Just because something is difficult to statistically control does not make it any less valuable or predictive, it only means you have to try harder if you're trying to publish a peer reviewed methodology. I've found Elliott waves to be enormously valuable, both predicting price movements and also for immediate in/validation which has allowed me to terminate positions with minimal loss.

It is my subjective opinion that your count is inferior to mine and I believe you would agree. Your subjective point was made, but you will find no argument. All technical analysis is subjective, even if that subjectivity is programmed. There were obvious price channels you arbitrarily ignored. I could program (objective) software to make a best fit in retrospect. It would not come up with your chart. What is more difficult for (eyeballs or software written by) humans is to count Elliott waves in real time. This is an art form. But past performance can be objectively measured. Some humans are better at it than others.

Computers can not yet play a decent game of go. While difficult to quantify (write an AI algorithm), it is no less possible to objectively claim that a professional has great skill, or race car driver, ballet dancer, football player, METAR/TAF weather forecaster...



Bitcoin has consistently expressed currency and commodity like properties rather than behaving like a stock chart. For example, the triangle from early this month took me off guard, because they've been rare ("Ending Diagonals appear in Wave 5 about three times as often in stock markets than commodities markets." Swannell, p59, 2003). If fact! I've just come across this again now, "Wave 2 can be any corrective wave except a Triangle" (Swannell, p63, 2003, but not Elliott).

Just because commodity charts often exhibit longer fifth waves on up trends (whereas stock charts tend to exhibit longer third waves) does not at all mean that a fifth wave MUST be longest. Bitcoin's extended fifth wave April-June should not surprise us and lends weight to the assertion that bitcoins behave like currency and commodities but not stock.

That does not mean that any 'ole longest five wave count is ideal. There are many other factors of which you are ignorant or have intensionally ignored. Most significantly is that every Elliott wave consists of smaller Elliott waves (and by implication, are members of larger Elliott waves). There are also suggestive price targets within each subwave which lend strength to a particular count. To count a single series on a single time scale is as good as rolling dice.

"Wave three is usually the largest and most powerful wave in a trend (although some research suggests that in commodity markets, wave five is the largest)." (Wikipedia)

"Wave 3 must never be the shortest by price when compared to Waves 1 & 5" (Swannell p37, 2003)



Drop further below $0.6 ?!

$0.6 to $1.1 is suggested by the previous fourth wave. It is not guaranteed, but I give it about 70% likelihood.


June was a bubble (which is typical of fifth waves), however bubbles collapse violently or symmetrically at best. The bubble and pop is over. Bitcoins are expressing a classic corrective pattern thus far. It could drop to zero, but I do not think so out of fundamental (not technical) conviction.


I'm not sure you can say with a straight face that, if I am proven correct in a couple of months, then my prediction would have been best. My predictions HAVE BEEN accurate in the past and MIGHT continue to be proven correct in the future.

The reason why I ignore the peak in February is just to demonstrate this subjectiveness.

Also since you quoted it yourself: In commodities wave 5 is usually the largest so I guess the whole point in whenever my count would turn out to be true is whenever bitcoin turns out to be mainly a commodity not a currency.

So to say it in other words: If the bitcoin economy remains less relevant it could be possible that the price crashes much further than you would initially expect. There are people who are convinced that bitcoin is a classical speculative bubble and if this turns out to be true the wave count I presented here could match that conclusion.

As for myself admit that I just don't know whenever bitcoin is a bubble or not and I think it is too early to say this for certain. There is a significant economic development taking place and if we are growing in the next couple of month I strongly believe that your wave count will be the best prediction.

Its also some "gut feeling" I have that with current sediment the price could drop much further.  The main reason for this I don't see a development into the direction of maturity taking place yet, there was a glimpse of hope when the "Bitcoin Foundation" was proposed but people weren't supportive enough. I mainly view the fundamentals to predict any long term price movements.
It's the Gartner Hype Cycle I'm talking about. To be able to say that the price will be raising again would mean that we are at the slope of enlightenment. That means developments which will shape the future outlook for bitcoin in a permanent way will become apparant.

So in short: If I see such a development during the next 2 months I'll support your thesis.

You present a valid count. I question why you ignore the all too obvious peak in February, March correction and impulse beginning early April. I don't deny, as you attempt to show, that Elliott waves have an aesthetic/subjective component. However, the market is reacting to this same publicly available aesthetic picture. The entire bitcoin community has agreed that there were doldrums during March and a huge impulse (some say bubble) April-June. Your Elliott waves should reflect and predict sentiment.

As a side note, Swannell's work in the past decade includes a deterministic Elliott wave algorithm, real-time computation, and statistics with thousands of price histories across multiple asset classes. For example, his work questions Elliott's earlier assertion that within an impulse wave, subwaves 2 and 4 will often take alternate forms. His statistics find that wave 4 is most often sideways nearly irrespective of wave 2's form. And, as I mentioned in PM, the fifth wave bias of commodities are less pronounced during downward trends, in which case third waves are likely to be longest regardless of asset class.

Ok, you are right there is something wrong if I interpret absolute peaks as endpoints of the waves.

However:
If I use the "typical price" I get a wave which fully conforms to the requirements!
The indicator doesn't make much sense here but I would still draw them this way if confronted with this chart.

While Aroon, RSI, or other indicators might hint at peaks and troughs, I would not expect Aroon to be a reliable determinant. Even so, there are multiple significant Aroon peaks you overlook, such as January, several in April, May, and many others.

Looking at your count specifically: Wave two (orange) can not drop lower than wave one (red). Though if we, as you suggest, assume wave one starts at $0, then wave three is the shortest wave of one, three, and five. So depending on your starting point, either rule #1 or #2 is violated. Rule #3 looks kosher: Wave four (purple) must not enter the price territory of either one or two.

My long term chart specifically cuts out earlier Mt. Gox history (about two months later than your start date). Perhaps I'm guilty of selective data, but I simply find the early data unreliable and also avoid accounting for the 'missing' data since Satoshi's gensis block January 2009.

No opinion on my interpretation?

The first ten days of November were clearly corrective, but I was blind to the triangle that now seems obvious. The 3-3-3-3-3 justifies the long sideways second wave. The dramatic fall since the triangle (III), though clearly impulsive, does not count five valid waves. The subsequent correction could not have reached above $2.8 before invalidating the entire black count above. The price could rise another 10% but without a clear previous fourth wave, $2.6 looks as good as any retracement from where to execute shorts.

As for the yearly chart, my bet is on another zig-zag similar to mid August through mid October (x-a=zig, b-II=zag). It is unlikely (impossible?) that we'll see a fourth combined correction down (else we might just call bitcoins dead). Despite the massive bid wall at $2.1, I believe the price will continue into the $1's in the short term. However, no matter where V bottoms (even well below $2), the correction (cyan b below) will likely retrace back into the large price territory of wave IV $2.1-$2.6 before sliding back into the $1's not to return for a long time.

Ok because of the invitation you'd have to endure my interpretation of Elliott Waves on the bitcoin price.

Is there something fundamental wrong about it? (I made it according to your explanation.)

I used the Aron Up Down indicator to get the most significant highs and lows to date. There are 5 waves and all end up on their respective most signifcant Aron values.

(About the first wave, it could actually start as zero since bitcoin wasn't initially traded for currency and given away for free.)

Above, I have two counts. The count on the prices (above) are valid. Those below the prices (identical to the 30 October count) are not valid, but look more realistic as long as we don't look too closely at the hourly sub-waves.

Magenta a above or black IV below (2 Nov. $3.45) display picture perfect three waves. Afterwards, it is not possible to count five waves, so corrective magenta b is better than impulsive 1.V (5 Nov. $2.85) though the cyan count looks better in my opinion.

Wave W or 2 (8 Nov. $3.2) is definitely corrective. Wave 2 (below) returns dead center into the suggestive fourth wave (in the below count, bigger than the magenta bar, III-IV or $3.1-$3.4).

The next magenta 'a' surprised me (today. $3.1). I expected to see at least one more wave down (there is no way to count five valid waves). The lower count black II looks pretty good (as long as the price stays below $3.2). Either way, a or II, it's corrective. I predict we'll drop down below $3 very soon, but what's next? I expect much lower, but I might exit my shorts just below $2.7.

For giggles, here is 30 Oct's count with eleven more days of data:



Something like this (as long as prices remain lower than $3.2, otherwise II could be 2, stop at $3.45):

I agree that the short count on this sideways correction is difficult, but I believe the larger trends are unambiguously pointing down.

29-31 October is a five wave downward impulse with an extended third after mid October's textbook zig-zag. I see no other viable October count. Whether November has drawn a complex second wave or a fourth wave after a pathetic third, I am not sure. But the larger message seems clear. Gravity is a powerful force and bitcoins are heavy.

"Paddle out and watch the first wave, dive under the second, ride the third, exit the fourth, ride the fifth."
      -- wave surfer, Elliott Island

The drop from the II.c $3.2 peak early 8 Nov looks impulsive, though I'm not ready to label even the first wave on any scale.

There are multiple options in the short term and this leads to a lot of uncertainty in the outlook.
This is why it is best to step back and let the market make a decision.
Long and mid term charts are still all pointing down .
A prudent investor would either wait until btcusd rises above 5$ or wait until the pattern becomes clearer.

"Netrin, you've presented three different variations, but all of them result in a downward arrow. Why are you so bearish?"

Great question. Glad you asked. Bitcoin could easily fall below $3 this minute, it could jump up between $3.2-3.31 as suggested by I.iv (30 October), or even retrace up to $3.8 at the beginning of wave I (29 October). Would a price above $3.8 make me bullish? Not necessarily. Certainly a move above $3.8 would invalidate my daily count (black I, II, III above in earlier posts) and force me to pull out bigger charts and longer scales (including October, perhaps back to August). A move into the $4's would only demonstrate a typical ending diagonal triangle wedge (left). Only an impressive third wave extension would make me bullish. What would that look like?

Notice the magenta bars mid September and most of October. Those represent the suggested corrective retracements ("A correction usually finds its [end] point in the area of the fourth wave of the preceding impulse wave of the same degree." -Robert Prechter 1988, if not earlier). It was hit dead center on 20 September and no less predictively on 29 October.

Notice the grey bar 22 October through 8 November. This represents the 30 October failure of the hopeful fourth wave since the rally cum correction beginning 19 October. The fourth wave can not retrace into the price range of wave one nor two. That's not a suggestion, that's a rule. I and II must not be labeled 3 and 4.

We should now expect, if we optimistically hope for a bullish rally, to see the third wave (cyan) manifest as a five wave (black) extension, labeled I,II,III,IV and 3.V on the chart above. Note that III must not be shorter than I (and likely longer), nor may IV retrace into the price range of I and II (cyan bar in the chart). 3.V should extend above III and given that bitcoin acts as a futures commodity we would expect prices ending well above $6, pehaps the double digits.

It could happen. I'd love it. But I'm not holding my breath.

Truth be told, I posted those only for the presentation. The above count has just been invalidated, now for two reasons. (1) Wave III was not itself five waves and (2) wave IV entered the price territory of wave one.

Earlier, I had rejected a count similar to that below only because sideways twos are half as likely as sharps (Swannell, 2003, p38). However, after this $3.2 pop up, not much else fits. Looking now at the smallest scales, I'm surprisingly happy with this count (stretching II from late October through this week). If correct, nothing prevents the price from rising above $3.5 or even up to $3.8 (beyond which the entire correction from 29 October comes into question). The dark bar demonstrates the fourth wave's violation of the first wave's price territory. The pink bar represents the expected price range of the retracement -- into the area of the fourth wave of the previous wave of the same degree.



Wednesday's count still looks pretty good, but the red i,ii,iii count fails on smaller scales (though bitcoin's low volume produces mostly noise on less than daily scales). We could be in any type of correction now, but if we allow for a very simple zig-zag or 3-3-5 flat for the previous and next few days, then we can expect a small impulse into the magenta iv range, around $3.3. We could just as easily fall from here. While the fourth wave of the previous downward impulse suggests a retracement between $3.2-3.3 (pink price band), it does not leave much room for a final five wave impulse for II.c. I have removed the count between magenta b and black II because I really don't know how it will play out. However, aside from negotiating pennies this week, none of these counts changes meaning significantly. We're in a small daily correction in a downward trend in all higher scales.

   Good to see more of this from you, m8.  Your labeling presentation and zooming has improved a lot, in the sense of someone like me being able to follow along easier.  Thank you for sharing.

  Very unique person that Swannell, from what bit I have looked through so far on the link you posted. Humans have such short life spans, it makes it near impossible for those that wish to spread knowledge, love and good will to be able to do so as much as the world could surely use....


    Cheers

I'm playing with an alternate count based on S3052's comments. It does not necessarily change message, but does open up the possibility of further ugly corrective patterns (triangles, zig-zags, flats, oh my!), whereas the fifth impulsive wave of the final c should have been definitive.

Swannell (RIP September) tells us that whereas upward fifth waves tend to be longest in commodities it is less true on down trends, where third waves tend to be longest for all asset classes. We'll have a better picture as we move out of the late October bear correction. Watch one, surf two and three, exit four, ride five.





Note: III is not five waves. The past week seems to have been and continues to be one big sideways corrections. Wednesday's count still looks as good as any.

While this (past?) week like most corrections has befuddled me, the downward meaning remains the same. Likewise, I have at least two alternative counts for the past week (no, BTCurious, none of them include a continuation of the corrective rally since 19 October). The hourly count (green) is typical bitcoin noise. Since Wednesday, I've recounted red iv as magenta ii, and the previous black 2 is now magenta iv. Neither count changes very much. We're trending down in all scales (seasonal II, monthly c, weekly 5, daily I, and magenta as either iii or v are both down).



I'd be surprised to see a jump above $3 again for weeks. However, if that were to happen in the next few days, I'd say we're still in magenta ii (or iv Wednesday count) but the chart remains the same, just shifted forward for a longer round of bouncing.

I believe S3052 is correct. My long term chart is technically invalid on the seasonal (II.a) scale. I considered the Mt. Gox hack to be invalid data. But I suppose that is not honest. It was a real event with major price implications. I'm still working on an alternate count. My eyeballs would make my old red b wave, S3052's X wave, and the implications could be more miserable than I have projected. We've got weeks before that would come into play before which I'll have a new yearly chart. Stock up on some rotten tomatoes. Thanks S3052 for the feedback. I appreciate it.

Interesting. Do you think you could line up some of those points A, B, W, X with dates? The 12 June hack and sell off certainly messed up the trend

As promised, some feedback on your EW count.

Overall, I see the longterm wave count the same (I: big rise from Aug '10 to Jun '11. II: correction Jun '11 to date, III: huge rally once II is over).

The only small issue is that the wave II correction can not be counted exactly that way.

Wave a can't be counted in 5 waves and this is why the current c wave will also not be a 5 wave move. Wave II is not a simple a-b-c zigzag, but rather a more complex correction, i.e. a "double three".