It was brought up recently in this thread about the previous ATH being different in relative terms, due to the total number of coins in circulation at the time vs now.  I think it's possible that there may be a way to model expectations by trying to take into account the factors you mentioned above.  Certainly those are valid points in both of those cases, but that's not what I am really trying to show in my graphs.  As I said earlier, I think what the 900-day cycle shows is a reflection of sentiment towards bitcoin, as reflected in the price, but modulated by a whole bunch of external factors, including what happens in other markets (FOREX, stocks and bonds, etc.), news in the Bitcoin/crypto world, focus/attention on BTC or competing cryptos, etc.

At least to this point, there is no perfect model.  I have seen some really good short term TA (Klee's PnF seems to be working remarkably well), and that can focus on price, but longer term the price becomes difficult, if not impossible, to predict.  I am pointing out the trends that have occurred and seem to be repeating, but don't expect relative highs like previous cycles - they could be higher or lower, relatively speaking.  Look at phase 1: cycle 3 repeated the trend of cycle 2 remarkably well, but we didn't reach a new ATH.  However, the graph shares the same shape.

I read through your analysis, though I admit I don't follow it all; it certainly seems to have reasonable conclusions, but remember, ultimately what we and all TA is trying to do is predict the future.  Humans haven't been able to do this in any arena so far, although Nostradamus arguably has a pretty good track record, and Moore's law continues to hold up well.  But everything is subject to change at some point, based on whatever unexpected events may occur in the future.

I think looking forward to prices, the best we can guess is a range, as you have done, and I applaud that.  Hopefully nobody bets too much based on anybody's models. 



I started looking at the trends back in 2014 when I was kinda bummed out by the price; at the time, I came up with various shorter-term models (~670 days) which matched well to that point, but began to break down as time went on.  Since that time, a 900-day cycle has matched (and continues to match) better.  It may not indeed be 900 days exactly, and it may vary from cycle to cycle.  A while ago rpetiela postulated a multi-year trend (2 years, IIRC), which does fall along my line of thinking as well.

The question about where to start a cycle is interesting, and one I struggled with.  I have already had to change the exchange I based my numbers on once (Gox was significant earlier, and since then Bitstamp is my source of numbers, and that may change again).  But there was value assigned before that... thing is, if the cycle length is correct, when exactly it started doesn't matter.  The up- and down-trends still match up, and the events occur ~900 days apart.  The earlier you start taking data from, of course, the smaller the sample size, so I wouldn't weight it too heavily.  I already don't assign too much value to cycle 1/phase 1, as the price was much more volatile and the market so much smaller.


As I say, I think the field is open to anyone's opinion; after all, this is the speculation forum.  I'm sure the Winklevii coins' status, as well as that of other large holders (Tim Draper, for example, or the BIT) have an impact on price, as they effectively affect market liquidity.  But even more so are the halvings.  They certainly affect price, but not instantaneously.  For example, there was no sudden doubling of price in Dec 2012, and there won't be next month.  However, as the number of new coins available to drop into the market decreases, there will be an effect - of that I am certain.  However, next month it could be less.  Dec 2012 we decreased the number of new coins per day by ~3600, next month we will decrease by ~1800.  In a market with so many coins trading each day, the effect will likely be smaller.  However, in a perfectly balanced market (reflected by a stable price), this could be enough to tip the scales up to a new balance point.  In a bull market the effect could be even more.

At the end of the day, the price reflects what someone is willing to pay for one Bitcoin.  Thus I am choosing not to try to account for these other things, but rather just watch how they affect the overall number.



Again, be careful - I am not trying to predict what the price will be, just pointing out the trend in price.  I have personal feelings on what the numbers may be, but that is just based on gut feeling and being generally hopeful and bullish about Bitcoin's future, rather than any sort of analysis of the graphs.

Thanks a lot for contributing to the discussion... certainly some interesting points you raised.  Let's hope the high ranges do pan out!