rpietila Wall Observer - the Quality TA Thread ;)

[Queeq]

Imagine you were back in the 90s, the WWW is starting to gain traction, you're convinced of its golden future, and you are able to buy shares of the HTTP. You don't even need a brokerage account or anything, you can do it in any amount, by cash, anonymously, on the street. That's pretty amazing.

Great point. Makes for a good sales pitch.

I love how these types of simplistic, short sighted, narrow, etc. comments (following):

All of the above arguments could also be applied to dot-coms when they were in their early stages. I'm sure early adopters were as ecstatic as BTC early adopters. So it must not be a surprise when bubble crashes sooner or later, before cryptocurrencies become as usual as Internet.

Bring out TRUTH that not only squashes the OP's comment but does so in a multi-faceted way.
Not only do we now see the diamond that is Bitcoin (and cryptocurrency like tech) but we get to see more of her facets thanks to the bright comments shared here.

Its About Sharing and thanks for doing so  

Sorry, I haven't exactly caught the tone of your message, either it's positive or negative. I assume that's due to English not being my native language. Anyway, I'll expand the idea behind comparison with dot-coms.

WWW and HTTP is very good example. Considering that blockchain is underlying technology like TCP, HTTP, etc., let's imagine we could had bought protocol shares back then, before the boom of dot-coms, just like we now buy Bitcoin and alts. Then, during the dot-com bubble, the price of these protocol shares would likely have risen along with dot-coms and crashed later. Just because underlying technology is not as demanded as before and initial euphoria is gone. That's what I meant by speaking about bubble crashing. Even though I wouldn't mind it for growing indefinitely, as I suppose the majority on this forum do, I don't believe it could. At one point it will fall heavily and it's better to be ready for it. I'm not saying it's going to happen tomorrow or happening already, just at some point in future.

[1714485860]

1714485860

[1714485860]

1714485860

[Its About Sharing]

Imagine you were back in the 90s, the WWW is starting to gain traction, you're convinced of its golden future, and you are able to buy shares of the HTTP. You don't even need a brokerage account or anything, you can do it in any amount, by cash, anonymously, on the street. That's pretty amazing.

Great point. Makes for a good sales pitch.

I love how these types of simplistic, short sighted, narrow, etc. comments (following):

All of the above arguments could also be applied to dot-coms when they were in their early stages. I'm sure early adopters were as ecstatic as BTC early adopters. So it must not be a surprise when bubble crashes sooner or later, before cryptocurrencies become as usual as Internet.

Bring out TRUTH that not only squashes the OP's comment but does so in a multi-faceted way.
Not only do we now see the diamond that is Bitcoin (and cryptocurrency like tech) but we get to see more of her facets thanks to the bright comments shared here.

Its About Sharing and thanks for doing so  

Sorry, I haven't exactly caught the tone of your message, either it's positive or negative. I assume that's due to English not being my native language. Anyway, I'll expand the idea behind comparison with dot-coms.

WWW and HTTP is very good example. Considering that blockchain is underlying technology like TCP, HTTP, etc., let's imagine we could had bought protocol shares back then, before the boom of dot-coms, just like we now buy Bitcoin and alts. Then, during the dot-com bubble, the price of these protocol shares would likely have risen along with dot-coms and crashed later. Just because underlying technology is not as demanded as before and initial euphoria is gone. That's what I meant by speaking about bubble crashing. Even though I wouldn't mind it for growing indefinitely, as I suppose the majority on this forum do, I don't believe it could. At one point it will fall heavily and it's better to be ready for it. I'm not saying it's going to happen tomorrow or happening already, just at some point in future.

Basically, I was saying that your statement was based on an analogy. This is very simplistic. Not that simple isn't sometimes right, and not that your analogy was completely false, but the replies to your message, showed much deeper and well thought out possibilities.

The analogy you pose, loses traction (or grip) as we start to understand that the protocol in this situation is money (for starters, and when you look at the further uses of it, more value is added). So, not only is it money, but it is a backing for other technologies which will be built on top of and along side it and all that they entail. (e.g. Counterparty, NXT, Ethereum, etc.) So, where as we can't exactly buy "http" we can by "BTC". The companies that bubbled and crashed are not exactly analogous to BTC itself crashing, where as those companies were built on top of the internet, they were not the internet (http). We might say any of Counterparty, NXT, Ehtereum, etc. might fail, but that is not the same as BTC failing. BTC (http) is not going anywhere anytime soon (unless there is a catastrophic event), so companies such as Pets.com, etc. might have failed, but HTTP never bubbled and failed. The analogy you pose is incomplete and has to be, as the current technology of Bitcoin (and it's blockchain) is not just a protocol. It also contains an intrinsic value in a sense.

Its about sharing

[Queeq]

Basically, I was saying that your statement was based on an analogy. This is very simplistic. Not that simple isn't sometimes right, and not that your analogy was completely false, but the replies to your message, showed much deeper and well thought out possibilities.

The analogy you pose, loses traction (or grip) as we start to understand that the protocol in this situation is money (for starters, and when you look at the further uses of it, more value is added). So, not only is it money, but it is a backing for other technologies which will be built on top of and along side it and all that they entail. (e.g. Counterparty, NXT, Ethereum, etc.) So, where as we can't exactly buy "http" we can by "BTC". The companies that bubbled and crashed are not exactly analogous to BTC itself crashing, where as those companies were built on top of the internet, they were not the internet (http). We might say any of Counterparty, NXT, Ehtereum, etc. might fail, but that is not the same as BTC failing. BTC (http) is not going anywhere anytime soon (unless there is a catastrophic event), so companies such as Pets.com, etc. might have failed, but HTTP never bubbled and failed. The analogy you pose is incomplete and has to be, as the current technology of Bitcoin (and it's blockchain) is not just a protocol. It also contains an intrinsic value in a sense.

Its about sharing

Thank you for clarification.

Protocol is money, yes. I would say that all the cryptocurrencies are some kind of protocols in their kind. We can say that if BTC is "kinda TCP/IP", then other alts are other protocols (either higher level like HTTP and SMTP or same level, like IPX, AppleTalk, etc.). Some of them will fail being a standard (IPX, AppleTalk), some won't.

But that was not the main idea. I am trying to convey that there will be the boom of companies that use cryptocurrencies as the main underlying resource, just how Pets.com et. al. used TCP/IP and HTTP. Let's call them cryptofiners (operating with cryptographic finances). This must lead to de-facto standard cryptocurrencies rise drastically as they get higher and higher demand. These cryptofiners will appear and appear causing a bubble similar to dot-com bubble, consequently, leading to cryptocurrencies' value bubble as well. Then the majority of them will fail due to the same reasons the majority of dot-coms failed. Some may survive but that won't stop the whole cryptofiners' bubble to burst. Along with it the value of underlying cryptos will fall. That's right that protocol itself haven't failed, however its value will decrease significantly.

This new bottom might be either at this level in case we are on the verge of cryptofiners boom, or at several K$ in case we see that boom after next rally, or elsewhere, but I see it very likely for dot-com bubble history to repeat.

[wachtwoord]

But between protocols there were wide differences. If protocols competed for the same functionality: one won. Most altcoins aren't different from Bitcoin in their application. You fill in the blanks.

[Its About Sharing]

Basically, I was saying that your statement was based on an analogy. This is very simplistic. Not that simple isn't sometimes right, and not that your analogy was completely false, but the replies to your message, showed much deeper and well thought out possibilities.

The analogy you pose, loses traction (or grip) as we start to understand that the protocol in this situation is money (for starters, and when you look at the further uses of it, more value is added). So, not only is it money, but it is a backing for other technologies which will be built on top of and along side it and all that they entail. (e.g. Counterparty, NXT, Ethereum, etc.) So, where as we can't exactly buy "http" we can by "BTC". The companies that bubbled and crashed are not exactly analogous to BTC itself crashing, where as those companies were built on top of the internet, they were not the internet (http). We might say any of Counterparty, NXT, Ehtereum, etc. might fail, but that is not the same as BTC failing. BTC (http) is not going anywhere anytime soon (unless there is a catastrophic event), so companies such as Pets.com, etc. might have failed, but HTTP never bubbled and failed. The analogy you pose is incomplete and has to be, as the current technology of Bitcoin (and it's blockchain) is not just a protocol. It also contains an intrinsic value in a sense.

Its about sharing

Thank you for clarification.

Protocol is money, yes. I would say that all the cryptocurrencies are some kind of protocols in their kind. We can say that if BTC is "kinda TCP/IP", then other alts are other protocols (either higher level like HTTP and SMTP or same level, like IPX, AppleTalk, etc.). Some of them will fail being a standard (IPX, AppleTalk), some won't.

But that was not the main idea. I am trying to convey that there will be the boom of companies that use cryptocurrencies as the main underlying resource, just how Pets.com et. al. used TCP/IP and HTTP. Let's call them cryptofiners (operating with cryptographic finances). This must lead to de-facto standard cryptocurrencies rise drastically as they get higher and higher demand. These cryptofiners will appear and appear causing a bubble similar to dot-com bubble, consequently, leading to cryptocurrencies' value bubble as well. Then the majority of them will fail due to the same reasons the majority of dot-coms failed. Some may survive but that won't stop the whole cryptofiners' bubble to burst. Along with it the value of underlying cryptos will fall. That's right that protocol itself haven't failed, however its value will decrease significantly.

This new bottom might be either at this level in case we are on the verge of cryptofiners boom, or at several K$ in case we see that boom after next rally, or elsewhere, but I see it very likely for dot-com bubble history to repeat.

I see your point and I think most of us here don't really claim to know what eventually is going to happen to BTC. But it does appear that the first mover advantage is a very very difficult thing to overcome. And now we can see BIG money (e.g. Wall St., Venture Capital funds, etc.) jumping on board with BTC. They will try to protect their investment at all "costs". Throw on top of that necessity, like remittances and person to person value transfers. The list goes on.

So, let's say there is a bubble in BTC. Where are we on that timeline? I bet you most would say we are at the very beginning. When the internet bubble happened (2002? ish) we all knew what the internet was. It had gained mass adoption due to AOL (those CD's all over the place), masses of websites, multiple ISP providers, etc.

At this point in time, it appears very premature to say we are anywhere near THE BUBBLE popping. I mean most people don't have a clue about Bitcoin. When the internet bubble popped money was beyond flowing into it, it was crazy. That is JUST starting here.

Here is a nice calculator to look at all of the variables involved in predicting Bitcoins price. We are just getting started (and that is NOT to say we don't pop/adjust in the interim).
http://worldbitcoinnetwork.com/BitcoinPriceModel-Alpha.html

[sidhujag]

Wow that was "gold" sorry for the pun. Maybe one day we will say wow that was bitcoin instead? lol

[holomen]

Everyone is on vacation while we reached -0.5 Log from the S.reeds trendline

It's already 30 weeks after last ATH (last one ,,consolidation period'' took 27 weeks)
 
If market would repeat itself identically we would need one more high volume sell-of to trigger run-up, if I made right assumptions?

[rpietila]

Everyone is on vacation while we reached -0.5 Log from the S.reeds trendline

It's already 30 weeks after last ATH (last one ,,consolidation period'' took 27 weeks)
 
If market would repeat itself identically we would need one more high volume sell-of to trigger run-up, if I made right assumptions?

Shameless repost of my new research, originally in the Trendline thread:

I just updated my personal sheet tracking these things. The parameters of that one:
- flat price 0.005 assumed until Mt.Gox opens
- for 7/2010-2012, Mt.Gox prices used
- for 2013, three exchanges
- for 2014, Bitstamp only
- daily VWAP
- first day 2009-1-3, the first day of Bitcoin; last day 2014-7-9, the 2014th day of Bitcoin

The exponential trendline for this time:
USD/BTC = exp(-2.869800 + 0.003012 * D), D being the number of days

R^2 = 0.934183

The trendline is at $1583.

At $623, we are currently -0.405 units below the trend.

For comparison, the max we ever went below the trend prior to the first 2013 runup, was -0.573.
Prior to the second 2013 runup, -0.341.


The maximum downside (greatest drop in price experienced ever after) after a day when price has already been at a low of -0.4 or lower, is -19%, experienced in 2012-10-5 to 2012-10-26, when the price dropped from $12.75 to $10.31.

This is comparable to the price going from $623 to $504 this month. This would put us on par with the maximum historical downside from already such a low point as $623 represents in the trendline.

I will not go to the upside calculations now, just leave the closing remark:


TL;DR: Historically, buying at a price such as this has represented a very low risk entry point with very limited downside in dollar terms.

Historically, selling at a price such as this (for any other purpose than covering immediate needs) has been counterproductive, as it is very likely that you will get a better or at least not much worse a deal; 1, 2, 3, 6, or any number of months in the future.

[rpietila]

Mr. Reed is not updating his trendline, which has become a problem after a few months below the trend. Actually it should be now at about -0.4 instead of -0.5. The difference is 30%, which is not insignificant

My latest one shows that there is no way to predict the start of the uptrend based on the relative price in the trendline only.

There IS a way to rather accurately predict the end of the uptrend, especially coupled with other indicators such as the doubling-time introduced by Mr. Reed.

There IS also a way to tell that -0.4 (or of course -0.5 if we ever hit it) is a very-low risk entry point, with a low chance of even moderate decline of the investment's value historically evidenced.

We'll just have to live within what is possible to predict, and with what probability. What I would not do at this point is to strategically sell bitcoins (even a fraction of your holdings), based on "it's going nowhere". If you don't need the money, just keep the coins. Their price is not going down!

I know it is hard to get anyone to buy them at this exact low-point of risk, which is the high-point of reward. But that is just the way of markets. The technology is silently getting better and more approachable, wall st is joining, etc etc., we'll just have to wait.

In my opinion, based on rather successful investing since 1996 (bitcoins 2011), we don't need to wait too long for the next runup in price.

[holomen]

Yes I have seen that before.appreciate.

 9:14 to 09:39 almost perfect timing ))  two same thought's and actions almost identical time on earth.

I Update his spreadsheet on my pc on a daily basis, But as I understand I was making a wrong assumption on trendline. As it should be constantly recalculating itself to current cycle.
I should work on spreadsheet to correct the matters regards to your last mentioned points.

I'll also need to find that doubling time indicator. But somehow more and more I look at the data I feel that this cycle could be a hyper-cycle.

I definitely don't think about selling any BTC , Crypto infrastructure and legal basis are starting to make ground.
I have heard of bitcoin since 2013-08 first invested since 2013-10 never sold any. Will be buying it till it reaches ATH. As I'm in mid 20's, after that there will be no way to increase the holdings by significant amount. Income will be  spent more on business and quality of life improvement.
 
As per legal issues on Bitcoin, Would you agree that Eastern financial centers- Singapore and Hong Kong will adopt Bitcoin way faster than Wall st. ? Taking in mind the position of China as master economy. Or you think that It will be faster lobbied by US top 10 families?    

 

[LOADING.READY.RUN]

I just updated my personal sheet tracking these things. The parameters of that one:
- flat price 0.005 assumed until Mt.Gox opens
- for 7/2010-2012, Mt.Gox prices used
- for 2013, three exchanges
- for 2014, Bitstamp only
- daily VWAP
- first day 2009-1-3, the first day of Bitcoin; last day 2014-7-9, the 2014th day of Bitcoin

The exponential trendline for this time:
USD/BTC = exp(-2.869800 + 0.003012 * D), D being the number of days

R^2 = 0.934183

The trendline is at $1583.

Thanks for the update!

This version of the formula directly works on Wolfram Alpha (copy/paste):

10^(-2.869800 + 0.003012 * ((number of days since 2009 Jan 03)/days))

or just follow this link:

http://www.wolframalpha.com/input/?i=10%5E%28-2.869800+%2B+0.003012+*+%28%28number+of+days+since+2009+Jan+03%29%2Fdays%29%29

[Driv3n]

Everyone is on vacation while we reached -0.5 Log from the S.reeds trendline

It's already 30 weeks after last ATH (last one ,,consolidation period'' took 27 weeks)
 
If market would repeat itself identically we would need one more high volume sell-of to trigger run-up, if I made right assumptions?

Shameless repost of my new research, originally in the Trendline thread:

I just updated my personal sheet tracking these things. The parameters of that one:
- flat price 0.005 assumed until Mt.Gox opens
- for 7/2010-2012, Mt.Gox prices used
- for 2013, three exchanges
- for 2014, Bitstamp only
- daily VWAP
- first day 2009-1-3, the first day of Bitcoin; last day 2014-7-9, the 2014th day of Bitcoin

The exponential trendline for this time:
USD/BTC = exp(-2.869800 + 0.003012 * D), D being the number of days

R^2 = 0.934183

The trendline is at $1583.

I put this info in Excel just to take a look at what the trendline gives for some future dates... lets just say next year looks amazing.  

2016 just looks like a dream world... but the fact that it could be reality blows my mind.  I wish I had known about Bitcoin earlier, but damn am I happy I found it when I did.

[rpietila]

I'll be organizing a summer retreat in Malla castle in August. The place is nice in summer. We had an event there already last month with 23 people attending from at least 11 countries.

Show it to me!

[dnaleor]

I'll repost my updated trendline here also:

I have updated my log chart, maybe interesting to share
used data: bitcoinaverage (starting at 17/07/2010)

Log chart

Trendline: P = 0,1711645509 * exp(0,0059413348*x) (with x = number of days since 17/07/2010)
R² = 0.8952

Log deviation from trendline (log(trendline)-log(exchange rate)):


And finally... The exponential trend on a linear chart. Always great to see this kind of charts  

I have a few thoughts about the much quoted "logistic model" (can be found here)
I think the logistic model is the correct one, but I tried to redo the work but it seems the author just used eyeballing to decide where the trendline "fits" based on some points and not on regression analysis. I know it's difficult, I tried (and failed) to do it myself based on regression analysis (because you need to do some assumptions about "saturation price" and the time needed to reach 50% of the market cap) but I'll give it another try.

So far, I think just an exponential model gives the best results. I also have some pre gox data, but making a regression analysis (like risto does) on asumptions starting in 2009, is very tricky: you give an overwhelming influence to the estimated data and data with very low traded volume. In fact, it would be more logical to assign weights based on traded USD volume. To simplify: trends based on the last year or 2 years data, I think will give better results than using as much data as possible.

thoughts?

(PS: I did not meant to offend someone.
I just want a small debate on what is the best approach for our models and maybe we can learn something from eachother )


edit: when we significantly deny the exponential trend, it could mean that we are in the second half of the adoption curve. So an exponential model is only valid until it's broken. on the other hand, the logistic mdoel takes a guess on when the market reached 50% of the saturation level...

[wachtwoord]

Is that the updated trendline? It's only 1000 now? Then it really got dragged down a lot   

[dnaleor]

Is that the updated trendline? It's only 1000 now? Then it really got dragged down a lot   

that is my updated trendline yes, but my trendline was always lower than the one of risto and SlipperySlope
(that's why I want to start a little discussion about the method)

[Wary]

Is that the updated trendline? It's only 1000 now? Then it really got dragged down a lot   

that is my updated trendline yes, but my trendline was always lower than the one of risto and SlipperySlope
(that's why I want to start a little discussion about the method)

Speaking of the method. Somebody once suggested using BTC capitalization, rather than price. IMO it's very good idea. 

[dnaleor]

Is that the updated trendline? It's only 1000 now? Then it really got dragged down a lot  

that is my updated trendline yes, but my trendline was always lower than the one of risto and SlipperySlope
(that's why I want to start a little discussion about the method)

Speaking of the method. Somebody once suggested using BTC capitalization, rather than price. IMO it's very good idea.  

Market cap is good for testing the "network effect", but I wouldn't use it for Technical Analysis.
(I don't know people using TA on the EUR or USD market cap )

[sidhujag]

Is that the updated trendline? It's only 1000 now? Then it really got dragged down a lot  

that is my updated trendline yes, but my trendline was always lower than the one of risto and SlipperySlope
(that's why I want to start a little discussion about the method)

Speaking of the method. Somebody once suggested using BTC capitalization, rather than price. IMO it's very good idea.  

Market cap is good for testing the "network effect", but I wouldn't use it for Technical Analysis.
(I don't know people using TA on the EUR or USD market cap )

Do we know usd or eur market cap? MAybe because with a printing press it tends to 0 therefor value tends to 0

[Wary]

Is that the updated trendline? It's only 1000 now? Then it really got dragged down a lot  

that is my updated trendline yes, but my trendline was always lower than the one of risto and SlipperySlope
(that's why I want to start a little discussion about the method)

Speaking of the method. Somebody once suggested using BTC capitalization, rather than price. IMO it's very good idea.  

Market cap is good for testing the "network effect", but I wouldn't use it for Technical Analysis.
(I don't know people using TA on the EUR or USD market cap )

The exponential trendline is not quite TA, because it is not derived from sentiments, mass psychology or whatever TA predictions are derived from. The trendline is result of technological revolution, growth of new technology and it's natural to measure and forecast this growth in % of market share, rather than on price of new gadget. Growth of internet wasn't measured in price of routers, and growth of mobiles wasn't measured in price of mobile phones. 

So the exponential is not derived using TA methods, but it can be used as input in TA analysis. We can extrapolate growth of bitcoin market share, from this share we calculate "real" price of bitcoin and that price we can use as fundamentals, to estimate how much bitcoin is overprices or underpriced.

Does it make any sense?