Bitcoin price cycles

[hacknoid]

(Note: This is a follow-on thread to my original one about price cycles; however, as I have re-analyzed things, I think at this point it deserves a new starting point.  See that thread for more background.)

I have been looking at trends in the Bitcoin price for a while now, and I've been noticing something, which I will detail below.  I think what I am seeing is the result of a couple of things:

• Bitcoin is a unique ecosystem.  It has arguably been allowed to grow organically, and as such I think the price is exhibiting some natural cycles, not unlike what you would find in nature in cycles of sunspots, plant shapes, sea shells, tree rings, etc.
• Price is, of course, driven ultimately by people.  Now, individually people's behavior is almost impossible to predict, but in large groups, certain fundamental patterns will exist.  Basic human emotions such as fear and greed are behind a lot of it, but regardless of exactly why it is, people can be somewhat predictable in large groups, not unlike animal herds or insect swarms.

With that in mind, I looked again at the price of Bitcoin over time, and tried aligning some of the peaks in price.  Previously this wasn't overly successful (although some general similarities showed up), but this time i though I would see what would happen if I looked at a full cycle between the peak in June 2011 and that in November 2013.  It turns out that there is about 903 days between those peaks, so I am going to round this to 900 and plot the data.  Observe:



All of a sudden there is clearly some repetition to the price cycles.  The scale is different, of course, but the trend is up, and the major events correlate quite well.  Know what else?  That 900 cycle divides neatly into 3 phases of 300 days each:



Each of those phases is marked by some clear trends, which has repeated each time.  (And yes, I know we would only be in the third cycles, but at least now, if the concept holds, behavior can be seen in the previous 2 cycles).  I'll break each down below.  Note that a 300 day phase is about 10 months, which is quite a long period of time in Bitcoin.  Here are the same graphs in linear scale, each on their own graph:







Phase 1: Initial runup
This phase is marked by a generally level or rising trend in the price, with one or two distinct jumps in price.  Note that the price run ups, as has happened with every major price increase, are followed by a fibonacci retracement, but in each case ends up higher than before the runup began.  In each cycle, Phase 1 ends with price being about an order of magnitude higher than at the start of the phase.  (Cycle 1 Phase 1 I didn't get the data for the first 40 days or so; doesn't really have much impact, and with such as low price and relatively low number of people trading, I wouldn't bet a lot on it's worth.)

Phase 2: Long phase exponential increase with more volatility
Each of the major runups in the two previous cycles (one in each) has been another order of magnitude increase or more, followed by a retracement valley and then up again before the phase ended.  The end of the cycle has been between about 6x and 10x higher than the start.

Phase 3: Volatility and instability
The least predictable of the 3 phases.  Cycle one showed a general increase over time (~50% increase), while Cycle 2 had a drop-off of about half.  Note again that this is over 10 months, so while a long slow decline is painful, it's not that bad in comparison to the overall trend.  Interestingly, there is a "hiccup" in Phase 3 of each of the two cycles, occuring at almost the same time (about 2/3 of the way through the phase).

What drives all this?  I'd characterize Phase 1 as showing pent-up demand (and maybe fading memories) about the last major increase, which, by this time, would have been well over a year before (previous Phase 2).  Phase 2 shows confidence gained from Phase 1 resulting in increased demand.  Phase 3 shows some shakiness in the confidence and basically is a "rebuilding" phase.  All these things are probably affected greatly by media attention, which also (I suspect) goes in cycles. 

As I said above, I think people's behavior as a group follows trends, even if individuals' behaviors are not predictable.  It is the individual volatility, such as the implosion of Mt. Gox and the clamp down by China (both back in 2014) that is trying to break trends and cannot be predicted.

What does it mean going forward?  Maybe nothing.  I'll say again, as I did before, that past history does not predict future performance.  However, look at the points and trends which correlate between those two previous cycles.  That's a lot of coincidences.  The cycling also predicts that before March of 2016 we should have expected at least one or two major price spikes.  We had one this month; if the cycle holds, there will be at least one more before March 2016.  Note also that the halving occurs part way through Phase 2 of this current cycle, which is widely expected to drive price up. 

So maybe there is something.  Maybe not.  But I have been getting tired of all the meaningless posts lately and wanted to post something new.

Oh, and if you are into numerology, there are a lot of "3"s in this analysis!

Disclaimer: I'm tired, and am not going to proofread this post now.  Have a good night, all.

[1714983744]

1714983744

[1714983744]

1714983744

[1714983744]

1714983744

[Alley]

Nice analysis.

[btcbug]

Interesting thanks!

[Harmonica]

Another attempt to predict a major bull run?

It just isn't happening in a cyclical nature for bitcoin, that is not how technology develops.

[Anemed]

The price cannot be predicted. Only the major adoption news of bitcoin will increase the price. Nothing else.

[NorrisK]

If it will follow this pattern, it will take around half a year for the next big run to happen.

However, it can also be that we are still in the initial pump, a big spike might obliterate the current peak in the graph and make it look like part of the rise.

[hacknoid]

The price cannot be predicted. Only the major adoption news of bitcoin will increase the price. Nothing else.

I think all the people that do the TA on this forum might not agree with your first statement...

It just isn't happening in a cyclical nature for bitcoin, that is not how technology develops.

We're not talking about the technology here... the tech is developing on its own; it's the price that we're talking about, and that is completely driven by other factors (media coverage, rates of adoption, greed, fear, panic...)  In that sense, and as I said in the opening post, Bitcoin is behaving more "organically".  We really have not seen anything like this before.  Was it possible to buy a "piece of the internet" or part of a torrent network?  Bitcoin is the first technology with completely digital units that you can buy.  And yet it's not like other currencies or commodities, since we always know the rate at which new bitcoins are produced, exactly how many there are now, and exactly how many there will be at the end.  This has never happened before.

In terms of predicting price or noticing cycles, take a look at this image highlighting what I would say are correlating events in the cycles as I described them:



Tell me if you don't agree that at least some of those areas look pretty similar?  That's a lot of coincidences, even if you don't agree with all 12.  And what does it mean?  Well, maybe nothing, but as I say, it's a lot of coincidences.  So maybe there will be some trend this time around.  Don't underestimate the power of human nature; people tend me to be influences by what others due, so price run-ups generally lead to others jumping in.  Similarly sell-offs and profit taking can drive others to do the same.

So there are a lot of factors that affect price.  If it were only good news that drove the price we should have had lots of increase since mid-2014 with all the positive news that came out.  But we didn't.  Why?  My thought is we were still too soon after the 2013 price peak, and the momentum was for price to push down.  General group sentiment by all Bitcoin traders.  That falls in Phase 3 of my cycle above.  Now?  In Phase 1, that sentiment would have run it's course and the time is ripe for the group mentality to change.  I don't know if it will (nobody does), but I think the timing is right for it.

[RyNinDaCleM]

In terms of predicting price or noticing cycles, take a look at this image highlighting what I would say are correlating events in the cycles as I described them:

Yes! Some of those areas look similar, but a few to note are:
(D) In 2011, D was already making higher highs (HH's) above the B. In Elliott wave, that was a subwave of the whole rise to $31.9099. In this case, it was a 3. In 2013, it was still part of the correction, or in the best supporting argument, a subwave of a subwave, 1 of 5. Visually, it makes no HH, so it really doesn't look similar at all.

Next, compare price of everything after the $1200 peak with the 2011/12 data you have there. Your corresponding points hit the bottom and began rising in the latter. In 2014, it continued down. This tells me that we are in a much larger cycle here that isn't really comparable to the 2010-2012 data you have without first stretching that data about 4x. I would argue that your F, G and H are more appropriately set up like this


Where F and G were all of 2014 and we are still somewhere in the H. Or at best, in the I. There is also the possibility that F was the entirety of 2014 and we are in G since January this year. I don't really have an opinion on which is more likely.

[Zaun]

Oh waw! This are some good research.

Thank you for sharing this and taking your time doing this.

[hacknoid]

Yes! Some of those areas look similar, but a few to note are:
(D) In 2011, D was already making higher highs (HH's) above the B. In Elliott wave, that was a subwave of the whole rise to $31.9099. In this case, it was a 3. In 2013, it was still part of the correction, or in the best supporting argument, a subwave of a subwave, 1 of 5. Visually, it makes no HH, so it really doesn't look similar at all.

Next, compare price of everything after the $1200 peak with the 2011/12 data you have there. Your corresponding points hit the bottom and began rising in the latter. In 2014, it continued down. This tells me that we are in a much larger cycle here that isn't really comparable to the 2010-2012 data you have without first stretching that data about 4x. I would argue that your F, G and H are more appropriately set up like this


Where F and G were all of 2014 and we are still somewhere in the H. Or at best, in the I. There is also the possibility that F was the entirety of 2014 and we are in G since January this year. I don't really have an opinion on which is more likely.

Hmm.. those are some very interesting points.  In terms of the period since $1200 peak, I had actually looked at compressing the comparative timeframes to see if they would better match up, but then I ran into two problems: 1) this no longer fit in with the concept of human "recovery times" that I was sort of basing this on, and 2) when (or does) the "compressed" period end?  The latter means its harder to judge when something might happen (though it would give a better idea of what would happen next: for example, if we are indeed in H or I, then we know a rise would still be coming, it would just be harder to judge when).  I'll have to go back an analyze those results again taking that into view (and I'll have to read more about Elliott waves).

Given the model I have so far, I could not help but notice the correlation between G and H, which are two period of relatively higher prices (though still on a downtrend) after the volatility ("F") that follows the peak price.  The main disparity is in the period between J and K, where in one cycle it is up, and the other it is down.  I could attribute the down trend to extenuating factors (China ban and Gox), but that's a guess.  My overall thought is this is a period of "light winds", as it were, where the price could easily be shifted in either direction, or lay flat.

Thanks for the input... I'll go back when I have more time and see about changing the time scale to see how the trends work out.

[mixan]

If it will follow this pattern, it will take around half a year for the next big run to happen.

However, it can also be that we are still in the initial pump, a big spike might obliterate the current peak in the graph and make it look like part of the rise.

A half year is where we will land when the halving comes then all the charts that is shown now will mean nothing.

[dothebeats]

Another attempt to predict a major bull run?

It just isn't happening in a cyclical nature for bitcoin, that is not how technology develops.

Price analysis != prediction of tech development.

There had been too many TA recently and in the past, but I don't think most of them had even come close to what have actually happened to bitcoin's price. :/

[hacknoid]

If it will follow this pattern, it will take around half a year for the next big run to happen.

However, it can also be that we are still in the initial pump, a big spike might obliterate the current peak in the graph and make it look like part of the rise.

Agreed.  Those are the two options that I see at the moment too.  In either case, a big run up looks to be in the cards sometime between March and December, 2016.  Both possibilities would be a rosy outcome, with the latter being even more rosy!

[hacknoid]

If it will follow this pattern, it will take around half a year for the next big run to happen.

However, it can also be that we are still in the initial pump, a big spike might obliterate the current peak in the graph and make it look like part of the rise.

A half year is where we will land when the halving comes then all the charts that is shown now will mean nothing.

The current estimate for block halving next year is around July 23.  That would put it about 4 months into the second phase, which extends to January 2017.  I think it's quite likely that the halving will drive price up sometime during that period, but exactly when and by how much nobody knows.  Last halving there was little effect at the time, but it followed about 2 months later by the rise to $266.

[Wilhelm]

The price cannot be predicted. Only the major adoption news of bitcoin will increase the price. Nothing else.

Yes the price cannot be predicted otherwise we would all be billionaires and a billion would be worth nothing .

TA can be done to determine the general direction of the market, it's no more than an educated guess that will give you a few percent better chances.

News is most likely not the main driver of increasing prices. First of all bitcoin news happens after something happens with bitcoin not before. So news will only act as an accelerator.

[Amph]

The price cannot be predicted. Only the major adoption news of bitcoin will increase the price. Nothing else.

Yes the price cannot be predicted otherwise we would all be billionaires and a billion would be worth nothing .

TA can be done to determine the general direction of the market, it's no more than an educated guess that will give you a few percent better chances.

News is most likely not the main driver of increasing prices. First of all bitcoin news happens after something happens with bitcoin not before. So news will only act as an accelerator.

actually if everyone had the chance to predict the price, nobody would gain anything, so no billionaires  at all

we would end with a sum zero game

[Wilhelm]

The price cannot be predicted. Only the major adoption news of bitcoin will increase the price. Nothing else.

Yes the price cannot be predicted otherwise we would all be billionaires and a billion would be worth nothing .

TA can be done to determine the general direction of the market, it's no more than an educated guess that will give you a few percent better chances.

News is most likely not the main driver of increasing prices. First of all bitcoin news happens after something happens with bitcoin not before. So news will only act as an accelerator.

actually if everyone had the chance to predict the price, nobody would gain anything, so no billionaires  at all

we would end with a sum zero game

Yep

and a billion would be worth nothing

[uki]

hacknoid, thanks for an interesting topic.
I am not a big fan of cycles in TA, but there is one thing that actually is very interesting in your chart. If you look at the ratio between the max and min price in each cycle, you will observe very interesting story, namely, for the first cycle (market introduction) we made more than 1000-fold increase ($0.01 to $30) in price, for the second cycle (initial adoption) we made about 100-fold increase ($10 to $1200). That leads me to a very interesting conclusion, which actually matches my expectations, in the third cycle (I would call it broad adoption) the price will go up ten(teens)-fold maximum. That would mean $2-3k as the realistic price target for the next 900 days.

[Slark]

The price cannot be predicted. Only the major adoption news of bitcoin will increase the price. Nothing else.

But you gotta admit that these bitcoin price graphs looks amazingly similar. I know it proves nothing certain as we have not enough repetitive cycles to study, but still.
Pure coincidence?

[hacknoid]

hacknoid, thanks for an interesting topic.

You're welcome! 

I am not a big fan of cycles in TA, but there is one thing that actually is very interesting in your chart. If you look at the ratio between the max and min price in each cycle, you will observe very interesting story, namely, for the first cycle (market introduction) we made more than 1000-fold increase ($0.01 to $30) in price, for the second cycle (initial adoption) we made about 100-fold increase ($10 to $1200). That leads me to a very interesting conclusion, which actually matches my expectations, in the third cycle (I would call it broad adoption) the price will go up ten(teens)-fold maximum. That would mean $2-3k as the realistic price target for the next 900 days.

I noticed something similar, although I don't put too much weight in the initial value of the first cycle; it was so early, before the first major public exposure and subsequent runup to $32.  I definitely agree that the cycles so far have tended to decrease in the amount of gain, and I am thinking the third would also be a somewhat smaller increase than the last. My personal feeling is that we would look for a peak price somewhere between 10x and 100x the starting price, or around $2300-$23,000.  Note that the peak doesn't last though, and tends to drop back to about the midpoint between the start of the runup and the peak, so that would mean a stable price more like $1500-$15,000.  On or about October, 2017.