first post here after lurking for a while..this very interesting thread and interesting BTC times (current bubble) have compelled me to post though
..
A thought: perhaps the underlying assumption of a log-linear relationship (log price, linear time) is incorrect and the time axis is also participating in a non-linear manner. For example, as BTC network/ecosystem gets larger and larger, the frequency of the swings and moves will also get farther and farther apart (because of that relatively increasing inertia as we move forward here.) Also the moves themselves (including the bubbles) will get drawn out over a larger time frame. If the previous bubble went through its parabolic rise in a week, for example, perhaps the next one would take all of a month (again as an example). Of course I do agree that absolute prediction is a ridiculous prospect - all kinds of random events can influence and there are different aggregates of rational and irrational behavior in the market as a whole at any given time.
Another thought: considering that the entrance of alternative cryptocurrencies is dissipating and perhaps will increasingly dissipate some of the energy in bitcoins successive bubble phases (such that the peaks of this and subsequent bubbles will be at lower and lower BETI values as a result of some of that cap being eaten up by the growing ETH's, BCH, LTC's and the like...i'm not saying zero sum game, all boats are going to continue rising here..but perhaps there is a resulting effect of dampening the acceleration somewhat..
Finally in concert with the thoughts above...I would argue that the periodic irrational 'bubble' phases that occur diverge wildly from the "true value" of bitcoin at that given point in time to a drastic extent (not a controversial point.) When it then eventually corrects back to the "boring lows", it is reasonable to assume that it is gravitating back to a more reasonable estimate of the "true value" growth of bitcoin at that point in time. If you just take these intervening periods when the world at large forgets about bitcoin for a while, there is still good growth, but the slope of this trend is very definitely NOT linear...as time goes on the trend of the 'lows', or what I am saying is an analog of a market assessed 'book value' of bitcoin, more resembles a sideways parabolic trend as shown in the image below.
http://i68.tinypic.com/29wo8s4.pngIf this updated assumption is accepted the previous bubbles then get corrected BETI values in the range of 2.4 to 3.2 (difference of Ln price and the parabolic). Using these values in the current bubble phase we are in right now gives estimates of:
BETI ATL (2.4) = approx $19,800
BETI ATH (3.2) = approx $47,900
So in this model, we could already have hit the current bubble peak (presuming we top out at BETI ATL) and we are now beginning our protracted correction back down to meet the 'book value' growth curve (grey sideways parabola in image). OR, we could be hanging out for a spell and will be at ~50k in the coming weeks/months (before the then inevitable protracted correction back to 'book value' growth again). Of course we could also land somewhere in between as well.
It also implies that whether 20k or 50k or somewhere in between, the next bubble after this one could potentially be 5 years or more down the road, and when it does hit, would likely not be the one that takes it over a million (presuming nothing else gets in the way to derail the BTC manifest destiny before that
)
Presuming we can get another couple of these bubbles, I think we still get to 1 million / BTC (or greater)...it just might take a little longer than what the original model prescribes.
Sorry for the book. Thoughts?