Fixed :D

Not attacking you, by the way, I appreciate your input.  Just that it's a pet peeve of mine: pointing out that despite all the other assumptions (remove outliers or not?, which period to use for input?, exponential function? double exponential?), the assumption that it is exactly one function we're looking for is perhaps the biggest (simplifying) assumption of them all.

Haha, nice.

I suspect that, even though the "official" accounts (hazek, for example) aren't posting much anymore, they still *read* the forum...

Bitstamp is pretty okay in my book, especially considering what the alternatives are

Why deprive more than a billion English speaking people of my brilliant ideas by limiting myself to a language only spoken by about ~100 million?

(I'm kidding, I'm kidding. Don't have time to post in both, so I'd rather post in the international parts of the forum. Exception is the stuff that is strictly of interest to Germans, like tax questions etc)

I keep forgetting, you're holding an order of magnitude or two more coins than me. Of course you already have a Mercedes

Anyway, glad you work the PR machinery the way you do. Not all large BTC holder understand the market like you do, IMO.

Let me be clear: there's absolutely nothing wrong with being bearish. I'm not so delusionally positive about BTC that I don't see the potential for a medium or longer term downtrend.

But there *is* something wrong with being bearish when the situation changed, and the outlook is much more balanced between continued correction and trend reversal upwards. Which is, IMO, where we are now.

Semantics, though. Wasn't my intention to attack you.

EDIT: this whole discussion is a bit tricky without price points, though. If "capitulation event" means revisiting 850, sure, I can see that happen. But right now there's still a significant number of people hoping for sub-500 coins, including those poor sods who "follow" rpietila and his log chart linear regression ideas... *That* would be a real capitulation, and it starts looking more and more unlikely we'll go back there, in my view.

I never quite got the excitement over Lamborghini, Ferrari or Bugatti. Maybe it's cause I'm German and I appreciate rock solid engineering and incremental technological innovation, but to me the pinnacle of a car is whatever is the latest Mercedes S-Klasse. Possibly tuned for an extra bit of power by their in-house tuning guys

I see.

Yes, I agree, they should handle it "lenient", plus they're active on here, so they can see you're a respected member of the community.

Still, your case is a bit of a special case -- you're making heavy use of the API, and you "lost" money more than really losing it, no? Don't get me wrong, I'm with you that they should unban your IP, but it's a bit harsh to put "they don't unban my IP fast enough" in the same category as "withdrawal backlog reaching as far back as July", just to name an example...

You forgot the mention of the cigar brand he was about to smoke. tsk tsk.

Care to explain the details?

Also, which alternative do you suggest? Certainly not mtgox or btc-e, right? And as much as I like bitfinex, they can't handle bigger volume right now :/

Anyway, my experience with bitstamp hasn't always been ideal, but their support always answered within reasonable time in my experience...

Well, the conclusion then should be that a log linear regression model cannot be used for daily trading advice, no?

It's useful to get an idea of the order of magnitude that can be expected at some time x, but for the nitty gritty details of trading decisions ("should I buy back in at 700?"), it sucks.

Problem is, he simply doesn't get that.

I don't know who "trashed" you for critizising rpietila, but the only reason why I recently decided to stop calling out his bullshit "predictions" as a "pro trader" is because he called me gay last time.

No really, he did.

That hurt man, that really really hurt

yeah, I'm going to half-defend you and half-ridicule you

Lot's of respect for your TA in general (told you that already earlier), but in the past week or so, you've been a bit too focused on the "duty" of the market to continue the correction -- in reality, even if the likelihood of a longer post-ATH correction was rather high, and still is, perhaps, it's important to keep your eyes open for the possibility that this time it *is* indeed somewhat different.

EDIT: though for fairness sake, you have been pretty objective IMO. didn't make a dumb price call, like e.g. rpietla with his "buying at prices above 600 is not worth it" mantra, based on his misunderstanding of a linear regression model.

You're based in Amsterdam? Awesome. We should have a drink some time. (Not gay.) (Just a bit gay.) (Only the tip.) (Never mind, I'm an idiot spending too much time on the sleazy part of the Internet)


Aaaanyway...

Like I said a page ago, I'd consider myself "agnostic" re: likelihood of further correction, despite the current breakout (or breakout attempt).

However, as I posted earlier in this thread, or in the Wall Observer thread a few days ago, in my daily trading decisions I've been following my experimental 'vwap/vwma median' method closely, and I'm pretty pleased with the result so far. I described the idea behind the method earlier in here, so go dig up my post history in case you care. So here's the most up-to-date view:





The first important step was staying above 770 USD (mtgox price), which we did, except for a brief while around Decemer 29. I posted about that result shortly before New Year's Eve.

The next median level (the one that corresponds to the first peak/bottom cycle) is at 915 USD (mtgox). We approached it yesterday, and now, instead of bouncing off of it it, we shot right through it.

So I admit, my "agnostic" stance is slightly leaning towards optimism right now. However, it continues to be conditional on the events of the next week or two:

If we stay above 915 for most of the time, and the other indicators play along (money flow, bid/ask ratio, shorter term trends), I'll seriously consider the possibility that the December correction has ended, much earlier than expected. Whether that means that we'll immediately see another big rally, or whether we first go into a longer consolidation/slow uptrend period is another question, but I do think that if we stay above the level mentioned above, another sharp drop -- i.e. the often demanded "final capitulation" -- is becoming rather unlikely.

Are (some of) you guys still trying to figure out which exact wave count to use to make sense of luc's "very long bear market" prediction?

Did it occur to you that, even under the assumption that TA works (and I *do* assume it does), sometimes the best (in the sense of: accurate and certain) result your TA can give you is "outcome unclear, try again tomorrow"?

What I mean is that to me it seems foolish to try to fit the current situation into a rigid mould. Yes, I am also aware of the historic precedent that ATH corrections simply don't finish that quickly. At the same time, you'll have a hard time explaining the current uptrend away. Maybe it won't last. But don't dismiss what happens because it doesn't fit what you expect.




Good analysis, in principle. Thanks for posting.

I see two problems with it though: (1) the "corrective" part is by now substantially outlasting the "trending" part. At some point one will have to concede that the "corrective" part becomes the trend. (2) Granted, volume is low, but even a cursory glance reveals that, in the past 10 days, volume has been substantially higher on the upward movements than on the downward movements (2h and 6h views on mtgox, and to a lesser degree, bitstamp, show what I mean quite nicely).

In summary: I remain "agnostic" about this correction. A lot of historical data suggests we'll be going down again in the near future, or at least won't continue going up until a final capitulation, but right now, there's quite a bit of evidence the current (upwards) trend has some staying power, and that we'll see a historic first of an extremely shortened post-ATH correction (i.e. much shorter correction than run-up).

A question for you, OP.

(By the way, don't take it personal, but I won't read through an auto-translated technical text, that's really not going to help me much.)

Since e is a constant, essentially your BTC price formula is an (exponential) function of the total number of BTC addresses. My question then is: how do you see it as significantly different form the countless attempts to model BTC price as an (exponential) function of time, the most well known one on this forum the linear regression of BTC price mapped to a log chart (recently rpietla made a big thread about this, IMO, rather uninformative method).

Looking at the graph in your Italian post (the blue line in the last graph, right?), it looks to me like your "formula" suffers from the same problem as the purely time dependent (linear regression) model: it gives a decent idea of the order of magnitude of BTC price at any time, but it is significantly off at times, and therefore more or less useless for trading -- more precisely: it is good to know that we can expect BTC price to be in the range of 1000 in mid 2014, but for example right now, your formula is "wrong", in an empirical sense, and it was even more off target a month ago.

@bitcoinwisdom:

any chance you could add bid/ask ratios? As in:


For each exchange X, for each time point Y: total bid sum on X divided by total ask sum on X, normalized for price at time Y.


It's an extremely useful indicator, especially when watched over time, in my experience

If you implement it, it could be added to the 'indicator' list (the one that contains now StochRSI, MACD)

Really?

If you would have taken the 5 minutes to actually read this thread and some of the replies by forum members that tested it, you would have noticed their newsletter/subscription service is very much *not* a scam.

I'm happily (mostly) subscribed for about a month now -- regular updates, competent TA, very friendly and quick responses to all questions I had.

In my experiences, well worth the price.

OP, sorry, really, but that's some of the sloppiest TA I've seen in here for a while. And that's saying something, considering that drawing a line through a log price chart is considered sophisticated around here...

Your point is basically that we have a price/average crossover similar to April/May, and a vague resemblance of the price chart, which you get by eyeballing... yeah, that's not good enough I'm afraid.

Just from the top of my head: the post crash consolidation phase is already lasting 50% longer than in April/May (then: ~2 weeks before the decline, now ~3 weeks). We have 3 green weekly candles atm, in fact. Another obvious difference: speed of the "recovery" -- significantly more controlled/slower right now, which I take as a good sign.

In any case, I'm personally undecided whether I think the correction is over or not, but your argument above is comparably weak.

I kind of  like the picture...  Assuming no panic kicks in later today, I take the measured response of the market to that red candle as a very positive sign. In fact, no sign of panic selling on bitstamp at all so far.

On the other hand, bitstamp has been acting as dampener to a rapid price increase during the last days (example: night of 30th to 31st). We continue staying just above the monthly ema, and while I know it can't last forever, that at some point it'll resolve either into a real rally, or another downtrend, for now it's all we ever hope for: very controlled growth, without all the excessive exuberance we love/hate so much.

EDIT too early though to say the dump had no effect. I'd keep my eyes open for panic selling later today.

Woohoo. Completely missed the spectacle....  Nice dump.

30d ema is support though it seems, and not for the first time during this Correction. Looks good to me.



Anyway


Happy New Year, everyone