Stephen Reed's Million Dollar Logistic Model

[Adrian-x]

Here is a great reference to the infrastructure growth and price (hint we in for some change.)
See: https://bitcointalk.org/index.php?topic=518111.msg8633552#msg8633552

The Jevons_paradox concisely explained.

Total network power consumption is a function of price.

Mining difficulty and hash rate are a function of mining efficiency.

BurtW also outlines the concept of price eras which is a function of difficulty that ultimately drive the total energy consumption of the Bitcoin network.

These price eras when viewed in conjunction with network growth, efficiency growth, available energy and price could in my view plot probable deviations from the Million Dollar Logistic Model mean.

We are at a juncture points now, but just off the top of my head this model may need to be refined to fit the price era data.

[arthur_excalibur]

You guys might find this clever research also interesting, I'm sure. Volume vs price over several years, indicating we are at the brink of something.

http://www.reddit.com/r/BitcoinMarkets/comments/2h0b48/daily_discussion_sunday_september_21_2014/cko9o1n

[ssmc2]

This thread is in my top five 

[Biodom]

At what point south of -0.8 log delta does the model need to be adjusted?

I would wait until the next bubble peak to adjust the model, or 2015 whichever is first. The left end of the logistic function is constrained to fit through the first recorded price. The right end of the model, I try to fit so that the recent variation is balanced above and below the line. If I were to fit the line today, I would balance the current price against the peak last November, and the trend line would likely pass through the low price in April 2014.

Thanks to all of you for watching this model unfold.

below -0.9 log delta and the line has gone down vertically in the last few days.
from the stock market-for reversal you typically need a capitulation (1-3 days) with an extremely high volume.
I posit that it will occur before the beginning of November (similarly to 2011 bottom). We seem to be in a purely technical decline right now.
Priceline was in such decline from $900 (split adjusted from 150) to $6 (split adjusted from 1) before it reversed itself and is currently $1186.

[SlipperySlope]

According to this chart of the smoothed adjusted number of bitcoin transactions, the uptrend of the past few months could, if continued at the same pace, surpass the previous all time high 0f 76394 within two to three more months. In the history of this data series, a new all time high has always been associated with the inflation stage of a price bubble.

This is my falsifiable hypothesis: if the 7-day average of the number of transactions excluding popular addresses as calculated by Blockchain.info exceeds 76394, e.g. reaches 80000, then bitcoin price will be rallying into a new bubble.

Among the ways this notion could be wrong is that transaction volume is now including more bitcoin tumblers to provide anonymity, or some other sort of non-commerce transactions, not filtered by Blockchain.info's simple algorithm.

[zimmah]

According to this chart of the smoothed adjusted number of bitcoin transactions, the uptrend of the past few months could, if continued at the same pace, surpass the previous all time high 0f 76394 within two to three more months. In the history of this data series, a new all time high has always been associated with the inflation stage of a price bubble.

This is my falsifiable hypothesis: if the 7-day average of the number of transactions excluding popular addresses as calculated by Blockchain.info exceeds 76394, e.g. reaches 80000, then bitcoin price will be rallying into a new bubble.

Among the ways this notion could be wrong is that transaction volume is now including more bitcoin tumblers to provide anonymity, or some other sort of non-commerce transactions, not filtered by Blockchain.info's simple algorithm.

i have been following those graphs closely, and i believe they are a good indicator. They also explain why we haven't seen a new ATH yet. Because we are not ready for it. But, we are getting near the point where we are ready.

[marcus_of_augustus]

This is my falsifiable hypothesis: if the 7-day average of the number of transactions excluding popular addresses as calculated by Blockchain.info exceeds 76394, e.g. reaches 80000, then bitcoin price will be rallying into a new bubble.

Obviously you are aware of Peter R's work re: transaction volume, Metcalfe's law, network value and market cap.

I'd be interested to know how far off the log trend of capital value growth, i.e. not price growth, that we are at present ... and what the previous largest deviation was using the capital value metric (not price).

[SlipperySlope]

This is my falsifiable hypothesis: if the 7-day average of the number of transactions excluding popular addresses as calculated by Blockchain.info exceeds 76394, e.g. reaches 80000, then bitcoin price will be rallying into a new bubble.

Obviously you are aware of Peter R's work re: transaction volume, Metcalfe's law, network value and market cap.

I'd be interested to know how far off the log trend of capital value growth, i.e. not price growth, that we are at present ... and what the previous largest deviation was using the capital value metric (not price).

Sorry, that I am very busy with my altcoin development. I will add the data series from Blockchain.info on bitcoins in circulation ...

Total Bitcoins in Circulation (CSV)

[oda.krell]

This is my falsifiable hypothesis: if the 7-day average of the number of transactions excluding popular addresses as calculated by Blockchain.info exceeds 76394, e.g. reaches 80000, then bitcoin price will be rallying into a new bubble.

Among the ways this notion could be wrong is that transaction volume is now including more bitcoin tumblers to provide anonymity, or some other sort of non-commerce transactions, not filtered by Blockchain.info's simple algorithm.

[...]

Or, alternatively, Metcalfe's law doesn't hold for Bitcoin (it's "Metcalfe's conjecture" in reality anyway, as an economic statement), or, alternatively alternatively, Metcalfe's law does hold for the Bitcoin network on a long enough time line, but is a poor predictor for shorter term phenomena such as a the start of a rally.

[arthur_excalibur]

My short-term analysis for what it’s worth… We saw some crazy trading and random madness at the end of last week, with a drop to $378. Volumes and volatility picked up nicely, but could and should be way better I think. Now it looks ‘business as usual’ again, but somehow I think we didn’t see the capitulation point yet. If we did, I’m not complaining of course lol! But looking back at late 2013, we saw a rather stable price around $125 for some time after some turmoil early September, then a sudden drop to $85 (-/- 33%) on October 2nd. This moment ignited the famous bull run which brought us to $1163 in November 2013. Somehow I’m thinking this type of moment is getting nearer and nearer again. And the drop from $400s towards $378 was probably not enough yet, in my opinion.

So I think we might see the following happening in the upcoming weeks. Price picks up slowly towards $450 (or lower, I’m quite doubtful about higher levels), then we will have another 1 or 2 days of craziness with another firm drop which will definately not be 33% like last year. Because all those % of jumps and drops seem to be getting less extreme over the years. So at $450 we will have a drop of ~25% which will bring us to $340, which happens to be exactly the lowest low we have seen so far after the $1163 top last year. Then, let the sprint begin…

[zimmah]

This is my falsifiable hypothesis: if the 7-day average of the number of transactions excluding popular addresses as calculated by Blockchain.info exceeds 76394, e.g. reaches 80000, then bitcoin price will be rallying into a new bubble.

Among the ways this notion could be wrong is that transaction volume is now including more bitcoin tumblers to provide anonymity, or some other sort of non-commerce transactions, not filtered by Blockchain.info's simple algorithm.

[...]

Or, alternatively, Metcalfe's law doesn't hold for Bitcoin (it's "Metcalfe's conjecture" in reality anyway, as an economic statement), or, alternatively alternatively, Metcalfe's law does hold for the Bitcoin network on a long enough time line, but is a poor predictor for shorter term phenomena such as a the start of a rally.

I think it the second one, metcalfes seems to hold, but since it's not instantaneous, it's not a great indicator for rallies.

[arthur_excalibur]

i have been following those graphs closely, and i believe they are a good indicator. They also explain why we haven't seen a new ATH yet. Because we are not ready for it. But, we are getting near the point where we are ready.

Check the 1w KDJ indicator, for example on bitcoinwisdom, and see how ready we are, indeed. However, this indicator says also that for a sprint and following ATH we have to wait a little more. Check the history at what KDJ levels sprints started. But the turnaround is near, that's for sure.

[SlipperySlope]

Here is the 15-minute chart of Bitstamp prices showing what I believe to be a short-covering rally. It will be interesting to see if this continues past a few more days....

[SlipperySlope]

And here is the daily gold/USD price chart which shows a small rally coincident in time with today's bitcoin price rally ...

[SmoothCurves]

And here is the daily gold/USD price chart which shows a small rally coincident in time with today's bitcoin price rally ...

That's an interesting coincidence. Or is it?

[BitDreams]

I would like to see bitcoin decouple from gold. Both bitcoin and gold formed higher lows in May/June and after both bitcoin and gold failed to extend those bullish patterns thru attaining a new higher low, bitcoin has broken above May's 421 mark while gold is still about 15 points below its May Low of 1238 - so gold is lagging bitcoin on this rally.

[marcus_of_augustus]

using total BTC (network) value not price (which is only proxy for total network value) ... and correlates with txs^2, (completely unadjusted)

[SlipperySlope]

using total BTC (network) value not price (which is only proxy for total network value) ... and correlates with txs^2, (completely unadjusted)

Awesome.

This is a nice complement to what Peter_R calculated. I hope that transaction quantity growth continues for the next few months. My hypothesis is weakened a bit with this chart which shows transaction quantity relatively surpassing market cap in the summer of 2012, without a major new bubble.

[BitDreams]

Beautiful chart, you can really see the divergence and I don't think it's transaction rate that is going down either.

[marcus_of_augustus]

My hypothesis is weakened a bit with this chart which shows transaction quantity relatively surpassing market cap in the summer of 2012, without a major new bubble.

That was Satoshi Dice running it's network transaction quantity "stress test" I think you'll find ...