Stephen Reed's Million Dollar Logistic Model

[SlipperySlope]

Your assumption is plausible but its still speculative.   The other plausible specualtive assumption is that bitcoin would go to zero.

Agreed, Satoshi himself had the same viewpoint - which I share. Bitcoin most probably either takes over the world or becomes worthless.

1.  You are looking at  bitcoin adoption only.   Not considering its competitor, the fiat.  Its maybe growing exponentially relative to bitcoin.   But has very far for market cap to catch up w fiat.  That is,  if it doesnt go bust first.   Bitcoin is fiat except the govt part.   Its fiat in the sense that it has no intrinsic value.   Except w fiat the demand is guaranteed because we are legally obligated to pay taxes w fiat.   

Yes I am looking at Bitcoin adoption only. In particular I am modelling the adoption of Bitcoin purchases by the population of all those entities which are now or in the future are capable of purchasing. I ignore fiat in this analysis. Bitcoin is unlike fiat in that its value is determined by its users, not by a government. Demand for Bitcoin is related to its utility, and I believe that governments will eventually accept tax and fee payments in Bitcoin. I suppose that central banks may get the notion that Bitcoin is better than, or at least a useful alternative to gold when accumulating reserves.

2.  Deflationary and decentralized currency is inferior to centrally planned currency.   See the history of money.   Banks used to be decentralized and currency was backed by gold standard.   That didn't work well so that's why we have what exists today.   Bitcoin is not an evolution its devolving.

I have a different view of the history of money. Governments have a history of debasing their fiat currencies. Inflation is a horrible menace to savings and rational allocation of resources. Bitcoin fixes that problem.

3.  Unless it becomes legal tender its is just a commodity.   But unlike other commodities its value is pure speculation.

Agreed.

4.  First mover advantage is a myth in tech.   Most widely adopted tech come after 2nd or 3rd iteration.  As long as something better comes along and no cost to switch itll happen.  Currently there is a big cost to switch from dollars to bitcoin.   

Wikipedia has an excellent article on First-mover advantage that convinces me that the concept is not a myth. Furthermore, Bitcoin was not the first digital virtual currency - rather Satoshi solved all the problems with the prior iterations to produce Bitcoin.

And I give you the last word on this day.

[1713983465]

1713983465

[1713983465]

1713983465

[1713983465]

1713983465

[1713983465]

1713983465

[podyx]

@SlipperySlope do you believe we are near the bottom based on your charts??

[SlipperySlope]

@SlipperySlope do you believe we are near the bottom based on your charts??

Yes, I share the feeling that we are near the bottom. I believe the resolution of the November 2013 bubble is remarkably similar to the resolution of the April 2013 bubble. Others have observed that twice prices dipped down briefly to around $400.

However I would not be surprised if prices fell briefly to an even lower level as $400 is about 34% of the peak at $1163. Back in the collapse of the April 2013 bubble, prices fell from a peak of $259 to $63 which was about 24% of the corresponding peak.

Sorry I cannot be more precise. The logistic model offers little in the way of short term price prediction aside from the price direction bias towards the trendline.

[podyx]

Thanks

[Bit_Happy]

so the end of 2014 looks to be around $6k and end of 2015 looks to be about $100k. Not a bad little chart. Im guessing 2014 will be higher than the chart has it, and 2015 will be a little lower

The rate of increase is about 10x annually. My model assumes that about 10x more speculators are buying bitcoin each year. In three years that is about 1000x more speculators buying bitcoin than now.  Perhaps in three years we will see participation by large companies, commercial banks, or central banks as speculators.

I think BTC will eventually be up to $100,000+ but it could take many years.
In the early years BTC went up many thousands of %.
Doesn't long-term growth require more "consolidation" first, before more "insane" Bull runs?

If we go up too fast it will be a slippery slope, Slippery Slope. 

[rpietila]

@SlipperySlope do you believe we are near the bottom based on your charts??

Yes, I share the feeling that we are near the bottom. I believe the resolution of the November 2013 bubble is remarkably similar to the resolution of the April 2013 bubble. Others have observed that twice prices dipped down briefly to around $400.

However I would not be surprised if prices fell briefly to an even lower level as $400 is about 34% of the peak at $1163. Back in the collapse of the April 2013 bubble, prices fell from a peak of $259 to $63 which was about 24% of the corresponding peak.

Sorry I cannot be more precise. The logistic model offers little in the way of short term price prediction aside from the price direction bias towards the trendline.

The comparison of post-bubble crash low to pre-bubble ATH will show that:

$2 to $1.05 => 1.90
$52 to $32 => 1.63
$384 to $266 => 1.44

I don't think we have any reason to go lower, and what is more, it suggests that going lower is quite improbable. Buying right now is definitely the thing. Very little potential gains if waiting (esp. offset against the heightened risk of keeping money in an exchange), but if the price starts slowly going up, you might end up chasing the slow-moving train and lose significantly.

[BitOnyx]

Well in theory bitcoin will hit 100k in 2100 soooo....

[Bit_Happy]

@SlipperySlope do you believe we are near the bottom based on your charts??

Yes, I share the feeling that we are near the bottom. I believe the resolution of the November 2013 bubble is remarkably similar to the resolution of the April 2013 bubble. Others have observed that twice prices dipped down briefly to around $400.

However I would not be surprised if prices fell briefly to an even lower level as $400 is about 34% of the peak at $1163. Back in the collapse of the April 2013 bubble, prices fell from a peak of $259 to $63 which was about 24% of the corresponding peak.

Sorry I cannot be more precise. The logistic model offers little in the way of short term price prediction aside from the price direction bias towards the trendline.

The comparison of post-bubble crash low to pre-bubble ATH will show that:

$2 to $1.05 => 1.90
$52 to $32 => 1.63
$384 to $266 => 1.44

I don't think we have any reason to go lower, and what is more, it suggests that going lower is quite improbable. Buying right now is definitely the thing. Very little potential gains if waiting (esp. offset against the heightened risk of keeping money in an exchange), but if the price starts slowly going up, you might end up chasing the slow-moving train and lose significantly.

Good rallies take time.
It does seem highly likely that this is a great time to buy.

[SlipperySlope]

so the end of 2014 looks to be around $6k and end of 2015 looks to be about $100k. Not a bad little chart. Im guessing 2014 will be higher than the chart has it, and 2015 will be a little lower

The rate of increase is about 10x annually. My model assumes that about 10x more speculators are buying bitcoin each year. In three years that is about 1000x more speculators buying bitcoin than now.  Perhaps in three years we will see participation by large companies, commercial banks, or central banks as speculators.

I think BTC will eventually be up to $100,000+ but it could take many years.
In the early years BTC went up many thousands of %.
Doesn't long-term growth require more "consolidation" first, before more "insane" Bull runs?

If we go up too fast it will be a slippery slope, Slippery Slope. 

Before understanding the logistic model, I too assumed that the rapid growth rate of bitcoin prices would decline. I accepted then, as I do now, plausible arguments that bitcoin prices could be 1 million USD or higher - provided that the Bitcoin economy eventually replaces the existing financial infrastructure.

But after fitting the logistic model to the historical bitcoin price series, it is a clear result of the model that the average exponential growth we have witnessed in the past four years will continue until we approach the half way point of speculator adoption. So I believe we will continue to see an average bitcoin growth rate of 10x annually for a few more years - given the tiny fraction of possible bitcoin purchasers that have actually purchased bitcoin.

Regarding consolidation, we can see from the price history that the consolidation periods follow bubbles, with the longest consolidation period of approximately 18 months following the greatest bubble in June 2011. The consolidation period following the April 2013 bubble in contrast was only about 5 months. Most observers here believe that the next bubble will begin in a few more months - assuming a consolidation period of say 5 to 8 months from the collapse of the November 2013 bubble.

There is little we can do about prices getting ahead of fundamentals, e.g. the steady 3.2x annual Bitcoin transaction growth, because that is human speculative nature.

[SlipperySlope]

The comparison of post-bubble crash low to pre-bubble ATH will show that:

$2 to $1.05 => 1.90
$52 to $32 => 1.63
$384 to $266 => 1.44

I am intrigued by your analysis, but cannot position the figures on my chart. Can you provide the dates for the figures so that I can see them on my Bitstamp daily price chart? For 2010 and 2011 I can use the Mt.Gox price series.

[SlipperySlope]

Metcalfe's Law Explains 10x Price Growth Vs. 3.2x Transaction Quantity Growth

At least two posters, gbianchi https://bitcointalk.org/index.php?topic=441336.0
and Peter R https://bitcointalk.org/index.php?topic=400235.msg5877592#msg5877592
have presented a theory on the relationship between Bitcoin transactions and bitcoin prices.

The consensus result is that prices are growing in proportion to the square of the transaction quantity and also in proportion to the number of addresses. Peter R uses the Blockchain.info data series that excludes the 100 most popular bitcoin addresses from the recorded transaction quantities.

Here is the outstanding chart provided by Peter R that best illustrates the application of Metcalfe's Law to bitcoin prices . . .



And here is the example device network from the Wikipedia article on Metcalfe's Law. The insight of the law is that the utility of the network, e.g. bitcoin users, is directly related to the number of reachable nodes. The number of connections is the approximately the square of the number of nodes.



Supposing that Bitcoin completely replaces the current payment infrastructure entails a transaction rate that surpasses the current quantity of bank card transactions. At a growth rate of 3.2x annually, Bitcoin daily transaction quantity will exceed 350 thousand in 8 more years. But applying Metcalfe's law predicts a bitcoin price of 58 billion USD per coin at that point - which is implausible. Therefore it is reasonable to suppose that small-world effects will disconnect the bitcoin price relationship from transaction volume, in that the future whole Bitcoin network will not be uniformly well connected but rather be clusters of well connected nodes. Here is the example graph from the Wikipedia article on small world networks . . .

[ThirdRenaissance]

Therefore it is reasonable to suppose that small-world effects will disconnect the bitcoin price relationship from transaction volume, in that the future whole Bitcoin network will not be uniformly well connected but rather be clusters of well connected nodes.

I don't see how this applies to Bitcoin. One of its fundamental properties is the decentralized consensus, as long as this mechanism works all nodes must be in the same state. By definition there can't be any clusters that behave differently and connectivity is not an issue as long as there is a connection to the network. Unless you are assuming that forks could exist alongside each other in completely disconnected networks, but then Bitcoin has failed as a whole.

[SlipperySlope]

Therefore it is reasonable to suppose that small-world effects will disconnect the bitcoin price relationship from transaction volume, in that the future whole Bitcoin network will not be uniformly well connected but rather be clusters of well connected nodes.

I don't see how this applies to Bitcoin. One of its fundamental properties is the decentralized consensus, as long as this mechanism works all nodes must be in the same state. By definition there can't be any clusters that behave differently and connectivity is not an issue as long as there is a connection to the network. Unless you are assuming that forks could exist alongside each other in completely disconnected networks, but then Bitcoin has failed as a whole.

Sorry that I was not sufficiently clear. By way of example lets look at the worldwide email system. Every email client can reach any other email client by email address - just like Bitcoin. But if China was suddenly subtracted from the email system, it would not make my email service worth a lot less because I cannot communicate in Chinese. The worldwide email system has small world effects in that well connected nodes, i.e. nodes that might send messages to one another, are partitioned by natural languages that the users understand.

By well connected, I mean that a node has a reasonable probability of reaching another node in the network - not that it merely possible to reach that node.

[kdrop22]

Nice chart, very insightful. Thanks for bringing it to our attention.
Slippery, please continue to post anything you might find interesting from the other threads here.

[marcus_of_augustus]

Therefore it is reasonable to suppose that small-world effects will disconnect the bitcoin price relationship from transaction volume, in that the future whole Bitcoin network will not be uniformly well connected but rather be clusters of well connected nodes.

I don't see how this applies to Bitcoin. One of its fundamental properties is the decentralized consensus, as long as this mechanism works all nodes must be in the same state. By definition there can't be any clusters that behave differently and connectivity is not an issue as long as there is a connection to the network. Unless you are assuming that forks could exist alongside each other in completely disconnected networks, but then Bitcoin has failed as a whole.

Sorry that I was not sufficiently clear. By way of example lets look at the worldwide email system. Every email client can reach any other email client by email address - just like Bitcoin. But if China was suddenly subtracted from the email system, it would not make my email service worth a lot less because I cannot communicate in Chinese. The worldwide email system has small world effects in that well connected nodes, i.e. nodes that might send messages to one another, are partitioned by natural languages that the users understand.

By well connected, I mean that a node has a reasonable probability of reaching another node in the network - not that it merely possible to reach that node.

But if everybody on the network can speak Bitcoin ....

[SlipperySlope]

Nice chart, very insightful. Thanks for bringing it to our attention.
Slippery, please continue to post anything you might find interesting from the other threads here.

Thanks! I received permission from Peter R to post his graph. And I will abide by your wishes regarding on-topic relevant material from other threads.

[twiifm]

Metcalfe's Law Explains 10x Price Growth Vs. 3.2x Transaction Quantity Growth

At least two posters, gbianchi https://bitcointalk.org/index.php?topic=441336.0
and Peter R https://bitcointalk.org/index.php?topic=400235.msg5877592#msg5877592
have presented a theory on the relationship between Bitcoin transactions and bitcoin prices.

The consensus result is that prices are growing in proportion to the square of the transaction quantity and also in proportion to the number of addresses. Peter R uses the Blockchain.info data series that excludes the 100 most popular bitcoin addresses from the recorded transaction quantities.

Here is the outstanding chart provided by Peter R that best illustrates the application of Metcalfe's Law to bitcoin prices . . .



And here is the example device network from the Wikipedia article on Metcalfe's Law. The insight of the law is that the utility of the network, e.g. bitcoin users, is directly related to the number of reachable nodes. The number of connections is the approximately the square of the number of nodes.



Supposing that Bitcoin completely replaces the current payment infrastructure entails a transaction rate that surpasses the current quantity of bank card transactions. At a growth rate of 3.2x annually, Bitcoin daily transaction quantity will exceed 350 thousand in 8 more years. But applying Metcalfe's law predicts a bitcoin price of 58 billion USD per coin at that point - which is implausible. Therefore it is reasonable to suppose that small-world effects will disconnect the bitcoin price relationship from transaction volume, in that the future whole Bitcoin network will not be uniformly well connected but rather be clusters of well connected nodes. Here is the example graph from the Wikipedia article on small world networks . . .

Are you sure Metcalfes law can used to price an asset class?   Isn't it used to describe network effects?

[SlipperySlope]

Therefore it is reasonable to suppose that small-world effects will disconnect the bitcoin price relationship from transaction volume, in that the future whole Bitcoin network will not be uniformly well connected but rather be clusters of well connected nodes.

I don't see how this applies to Bitcoin. One of its fundamental properties is the decentralized consensus, as long as this mechanism works all nodes must be in the same state. By definition there can't be any clusters that behave differently and connectivity is not an issue as long as there is a connection to the network. Unless you are assuming that forks could exist alongside each other in completely disconnected networks, but then Bitcoin has failed as a whole.

Sorry that I was not sufficiently clear. By way of example lets look at the worldwide email system. Every email client can reach any other email client by email address - just like Bitcoin. But if China was suddenly subtracted from the email system, it would not make my email service worth a lot less because I cannot communicate in Chinese. The worldwide email system has small world effects in that well connected nodes, i.e. nodes that might send messages to one another, are partitioned by natural languages that the users understand.

By well connected, I mean that a node has a reasonable probability of reaching another node in the network - not that it merely possible to reach that node.

But if everybody on the network can speak Bitcoin ....

Well yes, all peers on the network can speak Bitcoin. My hypothesis is that when considering the likelihood of connections between peers, i.e. the likelihood of a transaction between two addresses, the entire future Bitcoin network will be partitioned somewhat into clusters of addresses more likely to participate in transactions. For example, walk-in local merchants will have transactions with local walk-in customers, but are much less likely to participate in transactions with far away customers.

If someone can convince me that within 7 years Bitcoin entails the economic version of the Singularity, please do so. Otherwise I believe that Metcalfe's Law must eventually disconnect from explaining bitcoin prices given daily transaction quantity and addresses quantity.

[SlipperySlope]

[Are you sure Metcalfes law can used to price an asset class?   Isn't it used to describe network effects?

I am not completely sure that Metcalfes law can used to price an asset class.

However, the stunning results posted by gbianchi and Peter R are sufficiently convincing, e.g. the close alignment of the data series in Peter R's graph, make me very interested to see just how the law can indeed apply. Especially considering much headroom the Bitcoin economy has with regard to transaction quantity growth over the next seven years.

You might get a stronger explanation by posting your issue in gbianchi's thread linked earlier.

[twiifm]

[Are you sure Metcalfes law can used to price an asset class?   Isn't it used to describe network effects?

I am not completely sure that Metcalfes law can used to price an asset class.

However, the stunning results posted by gbianchi and Peter R are sufficiently convincing, e.g. the close alignment of the data series in Peter R's graph, make me very interested to see just how the law can indeed apply. Especially considering much headroom the Bitcoin economy has with regard to transaction quantity growth over the next seven years.

You might get a stronger explanation by posting your issue in gbianchi's thread linked earlier.

Yeah I tried reading that thread but the idea of using a formula to price a future is laughable.   

BTW I trade options so I know about how options price works and I know quite a bit about technical analysis.

One thing I can say is that most bubble charts have the same shape.    If you look at AAPL in 2012 compare it to FB,  TSLA,  AMZN today.      You'll find an uncanny similarity in there shapes.   With only a variance of 2-4 weeks.   Ive looked at past bubbles like MSFT,  YHOO,  NOK, CSCO too.   They have some resemblance to curent bubbles but slightly different.

Anyways by studying shapes of bubbles I predicted the top of TSLA,  FB,  NFLX they all crashed this week.   I believe AZMN is showing signs of bubble but not the top yet.   Same as GOOG

My theory is that bubbles are more technical than fundamental but the similarity is they disconnect from fundamentals and then eventually fundamentals pull them back into alignment.

I don't trade bitcoin so I don't know its chart.   But a sign of bubbles is when speculators invest on future market domination.   I believe they do this to justify their investment as well as recruiting greater fools so they can exit.  To me bitcoin smells exactly like a pump and dump.   

IMHO