Topic: Are we parabolic yet, it certainly feels/looks that way? also spikes closer (Read 2583 times)
November 12, 2013, 03:07:49 AM
You ain't seen nothing yet. Four digits BTC will create parabolics, many of them.
November 11, 2013, 11:53:26 PM
As for those that use "hyperbolic" as I've heard, it's nonsense when talking about stocks (or actually any real-life situation) : it would imply that the value of the stock becomes infinite, in a finite timeframe (singularity):
Actually, it doesn't imply that at all. All it implies is that the value of the counter currency will drop to zero, which is certainly not nonsense when you're talking about fiat currency. 
November 11, 2013, 03:02:17 AM
That is precisely why it makes sense to use it, at least for BTC. As you move towards the singularity, the doubling time halves. In spring, with BTC, it took 1 month to double the price, then 15 days, then 8 days, which effectively meant there would be a day 8 days further in the future when $/BTC would be infinite.
Hence, Bubble Pop.
Hence, Bubble Pop.
It literally is an asymptotic division by zero.
http://demonstrations.wolfram.com/TheHazardsOfProppingUpBubblesAndChaos/
November 11, 2013, 02:55:27 AM
Double exponential means that the curve is exponential on the log graph.
Last week it was looking parabolic on the log graph.
It will probably take at least a week to tell, but that curve may have been the growth rate jumping. If the price was already climbing at more than 1000% per year, what is it doing now?
Last week it was looking parabolic on the log graph.
It will probably take at least a week to tell, but that curve may have been the growth rate jumping. If the price was already climbing at more than 1000% per year, what is it doing now?
November 10, 2013, 07:15:26 AM
As for those that use "hyperbolic" as I've heard, it's nonsense when talking about stocks (or actually any real-life situation) : it would imply that the value of the stock becomes infinite, in a finite timeframe (singularity):
That is precisely why it makes sense to use it, at least for BTC. As you move towards the singularity, the doubling time halves. In spring, with BTC, it took 1 month to double the price, then 15 days, then 8 days, which effectively meant there would be a day 8 days further in the future when $/BTC would be infinite.Hence, Bubble Pop.
November 09, 2013, 07:23:26 PM
I'm thinking of the very long run when (if) it's a major global currency. Then it will follow the shape of the economy. In an ideal world the economy would slowly deflate to something more sustainable than now, but I guess we'll just get crashes instead.
November 09, 2013, 07:17:10 PM
Oh, here's one that just occurred to me: a perfectly priced bond should follow a hyperbolic curve with the par value as the asymptote.
I think
I think
November 09, 2013, 06:56:02 PM
Since bitcoin is inherently deflationary in the long run, I'm not sure it will follow a logistic function. Theoretically it should keep rising, to some extent, forever...
Logistic functions are probably not so good for stocks and such. They are good for describing things that have built-in limits, like population size. For example, the number of smartphones may follow a logistic curve until everyone in the world has one.
Logistic functions are probably not so good for stocks and such. They are good for describing things that have built-in limits, like population size. For example, the number of smartphones may follow a logistic curve until everyone in the world has one.
November 09, 2013, 06:38:33 PM
Oops, my bad, you're right of course, you can have horizontal asymptotes...
One would think it doesn't make sense in markets, having a top price, but now that I think of it, you can have markets following a logistic function, and bitcoin might be one of them (of course we're still full into exponential territory).
Logistic function :
One would think it doesn't make sense in markets, having a top price, but now that I think of it, you can have markets following a logistic function, and bitcoin might be one of them (of course we're still full into exponential territory).
Logistic function :
November 09, 2013, 06:17:51 PM
I don't think so. When you get infinities in physics, it just means that your model is not adapted to the singular point and its neighborhood.
(Of course you have hyperbolic trajectories, in orbital mechanics for instance, but I'm not using this concept in this sense.)
(Of course you have hyperbolic trajectories, in orbital mechanics for instance, but I'm not using this concept in this sense.)
You are right, regarding infinities. All I'm saying is hyperbolas don't have to have singularities at zero like the one you plotted. You can have rectangular hyperbolas approaching a horizontal asymptote.
The first example that comes to my mind is classical enzyme kinetics.
EDIT: Here, I googled something up:
http://web.anglia.ac.uk/numbers/functions_and_models/hyperbolic/hyperbolic_models.pdf
November 09, 2013, 06:00:37 PM
I don't think so. When you get infinities in physics, it just means that your model is not adapted to the singular point and its neighborhood.
(Of course you have hyperbolic trajectories, in orbital mechanics for instance, but I'm not using this concept in this sense.)
(Of course you have hyperbolic trajectories, in orbital mechanics for instance, but I'm not using this concept in this sense.)
November 09, 2013, 05:53:07 PM
As for those that use "hyperbolic" as I've heard, it's nonsense when talking about stocks (or actually any real-life situation) : it would imply that the value of the stock becomes infinite, in a finite timeframe (singularity):
Not necessarily, it depends on where you plot the asymptotes...
Hyperbolas have a lot of use in science, and are relevant to many real-life situations. I can't think of any relevance to stocks, though.
November 09, 2013, 05:32:58 PM
Maybe, we'd need a double-log chart to see that...
November 09, 2013, 05:30:06 PM
Secretly, unbeknownst to everyone, we are actually growing at a TRIPLE exponential rate.
November 09, 2013, 05:19:21 PM
Parabolic (x^2+1 in blue) is "slower" than exponential (e^x in red) :
In general, x^a ends up slower than a^x, because the growth rate of the second one keeps increasing : its value doubles at fixed intervals :
As for those that use "hyperbolic" as I've heard, it's nonsense when talking about stocks (or actually any real-life situation) : it would imply that the value of the stock becomes infinite, in a finite timeframe (singularity):
EDIT : The current increase does look double-exponential :
http://bitcoincharts.com/charts/mtgoxUSD#rg120zigDailyztgSzm1g10zm2g25zvzl
In general, x^a ends up slower than a^x, because the growth rate of the second one keeps increasing : its value doubles at fixed intervals :
As for those that use "hyperbolic" as I've heard, it's nonsense when talking about stocks (or actually any real-life situation) : it would imply that the value of the stock becomes infinite, in a finite timeframe (singularity):
EDIT : The current increase does look double-exponential :
http://bitcoincharts.com/charts/mtgoxUSD#rg120zigDailyztgSzm1g10zm2g25zvzl
November 09, 2013, 04:35:40 PM
I've always found it weird when people say "going parabolic"... is parabolic short for double-exponential (which makes more sense)? Actually growing parabolically would be very slow...
Good point. I guess people that say "parabolic" do in fact think of the parabolic shape on a log chart, which would make the function super-exponential. Then again, even the periods that are *not* called parabolic could be super-exponential in fact, just with small enough parameters that they appear near straight on the log chart.
November 09, 2013, 04:18:01 PM
Log chart. Logarithmic fucking chart. Always!
Loga-fucking-rithmic chart. Def worth repeating. And understanding why.
November 09, 2013, 04:01:10 PM
I've always found it weird when people say "going parabolic"... is parabolic short for double-exponential (which makes more sense)? Actually growing parabolically would be very slow...
well it's in the shape of a parabola, i think that is what people are thinking...... especially when it crashes..... like April run-up for instance November 09, 2013, 03:49:24 PM
I've always found it weird when people say "going parabolic"... is parabolic short for double-exponential (which makes more sense)? Actually growing parabolically would be very slow...
November 09, 2013, 03:23:48 PM
Yes, we (especially China) were going double-exponential, so it's a decent time to take a few % profits if you're overweighted in BTC, but not to sell the whole stash ARE U INSANE? 
We could very easily see a scary correction any day now, but the trend should continue much higher before the big bubble peak (in the thousands I'm thinking).
NOTE: Never use a linear chart when considering things like this. Log chart. Logarithmic fucking chart. Always!
We could very easily see a scary correction any day now, but the trend should continue much higher before the big bubble peak (in the thousands I'm thinking).
NOTE: Never use a linear chart when considering things like this. Log chart. Logarithmic fucking chart. Always!
Jump to: