Something, something, something, technical analysis

[zimmah]

not to mention candlesticks need a market open/close to even exist. you can make them by picking some arbitrary time as your cutoff, but depending on which time you pick, you get different patterns.

Doesn't matter.

A 1 day or 4 hour candlestick or candlestick pattern still gives plenty clues about market psychology over that time period.

I agree, it doesn't really matter if there's an open and close or not. Charts can still be made and are still relevant.

Not a single chart or tool can perfectly predict the future, but some can give a very good indication. And whether or not a market closes does not matter for charts

[oda.krell]

not to mention candlesticks need a market open/close to even exist. you can make them by picking some arbitrary time as your cutoff, but depending on which time you pick, you get different patterns.

Others said it already: you have a point in that the exact partitioning of the price data (which is what candles are) is arbitrary (though well defined ahead of time, so you cannot just "pick" the ideal boundaries you like best... however, you can pick another time scale, which can hide or show patterns that don't exist on another time scale), but it turns out that, on average, quite a few usable patterns emerge even after that arbitrary choice. Thomas Bulkowski's "Pattern Site" is probably the closest we have (or at least: that I know of) to provide some empirical foundation for TA as she is practiced, so with the caveat that this is not peer reviewed quality stuff, I'd cite him as evidence that (candle based) patterns tend to work -- don't know if he has data for shooting star though...

[Broseph Stalin]

My sloppily drawn chart. Two more legs to go?

It would seem like I was correct in my analysis. Sorry for not including the price to the right, but the C at the end was $620.. which we passed a bit down to $613 I think, but whatever. Close enough.

This would now put us in the fourth wave of this rally. The question is; how high will we go? Unless these two remaining waves are steeper than the previous ones, I don't see us hitting a new ATH. Maybe $850-900.

If they are steeper though.. all bets are off.

[sgbett]

I accept it's midnight utc to *draw* the chart but what I'm getting at is that midnight utc doesn't actually have the same effect on trading that an actual market open/close might have.

I can see you are all very keen for that to not matter though. I'm a chart noob anyway, I probably haven't got a clue what I'm talking about. Probably best to ignore everything I say

[sgbett]

Just to further discredit myself I would add that I would put the moon phase indicator as a much more reliable indicator than candlesticks. No homo. Or whatever it is you crazy kids say when you aren't joking.

[Gingermod]

Just to further discredit myself I would add that I would put the moon phase indicator as a much more reliable indicator than candlesticks. No homo. Or whatever it is you crazy kids say when you aren't joking.

Stop posting

[zimmah]

pretty interesting how the price seems to bounce around the first fib line (which dates back several weeks) then, after a drop recovers to find a new fib line that is climbing even faster, then it overshoots both lines just before both lines meet, and after a correction it seems to try to follow the old line again.

Pretty curious behavior if you ask me, not sure if it means anything but thought i'd just share.

by the way when you zoom in on the marked points, most of them have like 3 or 4 points of contact. Even at the last point it has 3 points of contact, which drew my attention because i didn't expect a line that old to actually be that significant even now.

[sgbett]

Stop posting

At least the other people that disagree had their big boy pants on.  What was your point again? You know this chart stuff is an art right? You know some people like Rembrandt and some people like Picasso? Tell me about your favourite indicator and why it's so good. Be sure not to say anything that everyone disagrees with though, the consequences could be unthinkable. You'll never be in the cool gang unless you say the right words.  So tell me to stop posting again. See if it works as well as the first time.

[thezerg]

Stop posting

At least the other people that disagree had their big boy pants on.  What was your point again? You know this chart stuff is an art right? You know some people like Rembrandt and some people like Picasso? Tell me about your favourite indicator and why it's so good. Be sure not to say anything that everyone disagrees with though, the consequences could be unthinkable. You'll never be in the cool gang unless you say the right words.  So tell me to stop posting again. See if it works as well as the first time.

You're right.  Someplace in the pacific would be better.  I'd be interested to see if it matters?  Do the candlesticks change completely if centered around different TZs?  If yes a lot of math needs to be applied to integrate an average worldview.

[oda.krell]

Just to further discredit myself I would add that I would put the moon phase indicator as a much more reliable indicator than candlesticks. No homo. Or whatever it is you crazy kids say when you aren't joking.

I get the impression you're half-serious, half-trolling... but your comments are always fun to read, so I'm not taking offense. Anyway. I'll try to parse your earlier points in two ways, then give an answer:

Questioning TA's validity in general: no really satisfactory way to answer to that. True, it's not properly tested, and quite possibly untestable for now (no proper model yet of the human mind, and only incomplete models of the human mind in groups), but some of those who use TA produce results that are statistically unlikely. I'd put TA in the same class of "intuitionist" frameworks as chess opening theory or Go strategies: they work, but can't be formalized to the point where they become testable under academic standards.

Questioning in particular candle based patterns: arbitrarily defined (but fixed) opening and closing times are de facto not a problem if the patterns that result from them produce statistically relevant effects. Like I said above, Bulkowski's methodology is not beyond reproach, but I wouldn't dismiss it out of hand either. Candle based patterns seem to work, on average, so whatever detrimental effect the arbitrary choice of data partitioning has, it doesn't ruin candle's usefulness entirely.


That said: turns out, that was not a shooting star (under the stricter rules of the pattern). Under my own (laxer) interpretation of looking at the June 1 candle as an inverted hammer, it starts to look like it didn't really mark a significant top (though we have to break 684 yet)

[sgbett]

I'm always at least half joking, I'm afraid its a condition I've developed whereby I find it best not to take anything too seriously, least of all myself! (and magnified on the internet)

I do accept there are patterns, I certainly couldn't say with any conviction there aren't. I've become more open minded to TA as practised by people like yourself, who do seem to understand how nuanced it is - an art form. As opposed to those that think its science, and seemingly refuse to accept that any given interpretation could be entirely wrong.

I think on a broader scale all patterns represent depictions of crowd psychology, and to this end their certainly is probably some information in them. That's why as outlandish as it sounds I think the moon indicator belongs in their. Human psychology and full moons seem to have some relationship, even if its along the same lines as people believe in chart patterns therefor they come true line of reasoning.

My reasoning behind the candlestick patterns being less relevant was due to that psychological effect of market open/close and how the intraday price in relation to the open/close price seemed to form some of the reasoning for certain candlestick patterns being described as they are.

I think the dailies (and weekly etc) are a different kind of beast to the intraday, 5 min, 10 min etc candles. Whilst, as you say, they both likely contain statistically relevant information, I suspect the kind of information they provide or the behaviour from which the two groups arise is likely different. Human psychology from one minute to the next is more a continuum, whereas the overall behaviour in a defined period with deadlines would be more discrete, and the time during which the market is closed provides quite a long period for significant changes to occur - a fairly simple example of this is the gap up/down one sees typically following some development out of hours. As BTC doesn't have out of hours we often get much smoother (although steep) moves on significant news, and these moves sometimes echo as other markets around the world enter their peak hours. In some way it may be that the always open nature of the bitcoin market somewhat dampens the volatility. In markets which gap up or down, they have to open somewhere and I think that initial jump may be much bigger due to people wanting to ensure they get in on the action (something like how an auction with sealed bids produces different results to one without). Again all of this is just me pondering the possible effects, not stating they are fact.

I suppose if I had to sum up what I am thinking, if one can intuit the difference in a regular market between the intra and inter day candles and what they show, then I would be more likely to treat all bitcoin candles regardless of period as 'intra' style. Given the market is continuos.

As I said before, I don't know really anything about it in practice and I'm just thinking out loud about the theory. Maybe it might be useful to someone, even if its only a starting point for a much deeper investigation.

(and definitely no offence intended)

[zimmah]

pretty interesting how the price seems to bounce around the first fib line (which dates back several weeks) then, after a drop recovers to find a new fib line that is climbing even faster, then it overshoots both lines just before both lines meet, and after a correction it seems to try to follow the old line again.

Pretty curious behavior if you ask me, not sure if it means anything but thought i'd just share.

by the way when you zoom in on the marked points, most of them have like 3 or 4 points of contact. Even at the last point it has 3 points of contact, which drew my attention because i didn't expect a line that old to actually be that significant even now.

happened again today, the top was exactly at the point where it hit the old line.

[Bitcoin_is_here_to_stay]

Nice to see this thread still alive, there is almost no TA left anywhere in this forum

Now we consolidating in 630-640 area on Stamp. Up or down from here?

The local bottom at ~$620 on Stamp was exactly the target of triple top formation, so we could try to go up from here. On the other hand, since $450 breakout, we went almost always up, without any substantial correction. Moreover, after at least 3 failed attempts to go above $680, even if we go North of $640, panic selling may begin even earlier this time? Yet on third hand  , some longs closed and new shorts opened on 'Finex. These shorts may be squeezed/scared by a move toward $680?

Thoughts?

[oda.krell]

We're getting closer to a decision point, I believe: Bollinger band width is near a point where the market usually starts moving again. The (hidden) bullish divergence on the 2h MFI (works on 6h as well) makes me lean towards upwards breakout, but daily MACD is about to go negative which means there's also reason to go down, short term.

Whatever way it moves at first though, I see a lot more potential to the upside than the downside currently: slow momentum is there (3d MACD positive, 1w MACD about to go positive), we're just banging our head against the daily SMA200 ~= 38% fib of entire downtrend ~= $650 resistance for now.

One way I could see it resolve is a dip at first, down to around daily SMA20 (~$612), then gathering the momentum to break the $650 resistance on the rebound.

[JustAnotherSheep]

It's also possible we'll continue to consolidate here for another week or two. My longest triangle count ends at the 22nd.



There also seems to be hidden bearish div on 30m MACD, though at that timescale I'm not sure what significance it has, if any. Yet that, and price being so close to the support lines, does make me a bit wary. It could further be argued that the real triangle (red support line in chart above) was already broken down on the 9th, although I'm hesitant to go with that idea and would rather await confirmation of the short-term bear scenario by the breaking of lower supports.

As for daily MACD about to go negative, that could mean two things in my view. It could "bounce", as it did it November 2013 where, just as it was about to cross down, the market resolved to shoot up instead. But that was within the bubble, and a fairer comparison could be September 2013 where it did cross down, which was followed by a month long consolidation period.

Also, the "magic" pre-bubble RSI level I noticed in my previous post has again seemingly been met (84.5), and now it's crossed down from being in overbought territory.

Though, looking at historic timescale on Stamp, I just noticed an interesting pattern: For every case of (daily) RSI being overbought outside of bubbles, the RSI top has been extremely consistent before price went down/consolidated: 83-85 (what do you even call that, RSI level?). Currently we're on 79, which according to this would give us some more room to go before the hypothesized correction (also assuming this is not the start of a bubble, which I doubt). Curious to see if this pattern holds again.

If that historic pattern continues to hold, it would support the longer consolidation/sideways movement scenario before the big rally can begin (which could be achieved by either a short-term price drop and formation of a much larger triangle pattern, a short spike (to $750-ish?) and the same, or just continuing within the longer triangle formation(s) at this level).

But then again lucif seems confident that it will (eventually?) go up from here, and I make it a habit to preemptively pack my bags when the clown prophet says we're going to the moon

[MatTheCat]

We're getting closer to a decision point, I believe: Bollinger band width is near a point where the market usually starts moving again. The (hidden) bullish divergence on the 2h MFI (works on 6h as well) makes me lean towards upwards breakout, but daily MACD is about to go negative which means there's also reason to go down, short term.

Whatever way it moves at first though, I see a lot more potential to the upside than the downside currently: slow momentum is there (3d MACD positive, 1w MACD about to go positive), we're just banging our head against the daily SMA200 ~= 38% fib of entire downtrend ~= $650 resistance for now.

One way I could see it resolve is a dip at first, down to around daily SMA20 (~$612), then gathering the momentum to break the $650 resistance on the rebound.

My thinking precisely!

Except the times in the past when we have actually agreed with each other, we have normally turned out to be wrong.

[cbeast]

So, what's urinalysis?

[sgbett]

My analysis (short to medium term) based on nothing but a gut feeling is the current rise is a tiny bit overextended, which is perfect for an engineered smack down designed to shake out a few loose coins to be snapped up cheap, before we resume a gentle uptrend.

Long term, I think this is it for super-exponential rises followed by blowoffs.

I think the 4-5 figure target is almost inevitable in the 2-5 year timeframe, but I think the route by which we get there is going to be very different to what people are used to and fraught with peril.

Forgive me for the lack of chart to back up my rambling

[Gingermod]

My analysis (short to medium term) based on nothing but a gut feeling is the current rise is a tiny bit overextended, which is perfect for an engineered smack down designed to shake out a few loose coins to be snapped up cheap, before we resume a gentle uptrend.

Long term, I think this is it for super-exponential rises followed by blowoffs.

I think the 4-5 figure target is almost inevitable in the 2-5 year timeframe, but I think the route by which we get there is going to be very different to what people are used to and fraught with peril.

Forgive me for the lack of chart to back up my rambling

Your rambling is unsubstantiated. There has yet to be a break in the overall trend line. We're going up:
https://www.tradingview.com/v/A1qSApSj/

Expect a couple weeks of this stability.

[sgbett]

If something hasn't broken it can never break?

I see.