Wall Observer BTC/USD - Bitcoin price movement tracking & discussion

[JorgeStolfi]

China's slice of the total volume increased slightly from 88% to 89%.

Their zero fee exchanges can churn thousands more btc than normal fee charging exchanges. [ ... ]

Of course.  But they do have that much volume, just as an exchange with higher fees or other handicaps may have less volume.

Obviously there are other useful measures, like book depth order, etc.  It would beinteresting to see them tabulated and compared.

Nevertheless, Huobi and OKCoin do seem to have greater liquidity and stability, a narrower spread, etc..  They do seem to lead the market most of the time.  So volume, independently of the reason, may be an important variable after all..

[1714916451]

1714916451

[1714916451]

1714916451

[lyth0s]

China's slice of the total volume increased slightly from 88% to 89%.

Their zero fee exchanges can churn thousands more btc than normal fee charging exchanges. [ ... ]

Of course.  But they do have that much volume, just as an exchange with higher fees or other handicaps may have less volume.

Obviously there are other useful measures, like book depth order, etc.  It would beinteresting to see them tabulated and compared.

Nevertheless, Huobi and OKCoin do seem to have greater liquidity and stability, a narrower spread, etc..  They do seem to lead the market most of the time.  So volume, independently of the reason, may be an important variable after all..

That volume is most likely just bots exchanging and re-exchanging the same coins even for a possibility of a 0.00001% gain since there is no fee. Does high volume of exchanges keep the price more steady? I think so because it creates a more baseline value of each coin at that moment (multiple "people" buying/selling around the same price). Are they the ones that lead bitcoin price changes? I don't think so, I really think they are reactive to other bitcoin markets which involve more human transactions, even if it means that their bots have large and quick responses to subtle changes in the other exchanges making it appear although they lead the exchange price. Just IMO.

[JayJuanGee]

That was nice dump.  

caught me by complete surprise, i didn't think the first small move down to $620 was significant at all

Did you put your money where your mouth is?

[JayJuanGee]

To be honest the only thing that can collapse the price IMO are the week hands, (non believers) stolen Bitcoin and whales redistributing/ facilitating future wealth growth, and  I think a lot of coins @ the new order of magnetude have already been flushed.

The big question for me is "what really happened to the Gox coins."  I'm working on the hypothesis that the coins were stolen along time ago and have already been absorbed by the market.  If this is true, I can't see us dipping very deep and in fact think this is quite bullish medium term.  On the other hand, if a thief really has 750,000 BTC, I could imagine considerable selling pressure for some time.  

The first part of your theme seems possible........... but likely NOT complete b/c in my thinking it would take considerable time for 750k btc to be absorbed completely into the market.  The second alternative seems even less possible b/c it seems very unlikely (though possible) that the stealing would have ONLY been by one thief.

[billyjoeallen]

To be honest the only thing that can collapse the price IMO are the week hands, (non believers) stolen Bitcoin and whales redistributing/ facilitating future wealth growth, and  I think a lot of coins @ the new order of magnetude have already been flushed.

The big question for me is "what really happened to the Gox coins."  I'm working on the hypothesis that the coins were stolen along time ago and have already been absorbed by the market.  If this is true, I can't see us dipping very deep and in fact think this is quite bullish medium term.  On the other hand, if a thief really has 750,000 BTC, I could imagine considerable selling pressure for some time.  

The first part of your theme seems possible........... but likely NOT complete b/c in my thinking it would take considerable time for 750k btc to be absorbed completely into the market.  The second alternative seems even less possible b/c it seems very unlikely (though possible) that the stealing would have ONLY been by one thief.

If the coins were originally stolen in the 2011 Goxhack, then it did take a long time to absorb them into the market. A year and a half before the next ATH.

[Davyd05]

601 on stamp seems to the battleground area, for fibs it  down breaks below 38% retracement at this threshold..

However the market looks like its being hit up with waves of buy and hodlers, I mean the volume for these last dumps isn't as significant as I'd expect given the level of Fud and goxxing floating around

edit: assuming a lot, the bids seem hidden or placed neatly just above the points above 604 and 606, also being recent lows as of this run up to 710 buy a rushed hodler

[JorgeStolfi]

According to my graph below we are still at a higher price per bitcoin than would be predicted with an R squared of >0.9

Fitting trend lines to historical prices may give heart-warming plots but is not useful for predictions, unless the trends reflect some underlying fundamental variable.

Exponential trends may be caused by "infection" in a new market: as more people invest in it, the more the item is mentioned in the news and spreads by word of mouth.  Or they maybe due to the company growing, hence opening more stores/factories/products/whatever, hence growing even more. The exponential growth stops when  that market is saturated, there is no more room for the company to grow, or something else happens (e.g. devastating news) that turns investors away. 

In that chart I see three separate exponential trends, but since they are shifted (P(t) = A*Q**(t-t0) + B, instead of just P(t) = A*Q**(t-t0)) they do not look straight in the log plot. They span

  (1) from ~Jun/2012 to ~Jul/2012 (~2 months)
 
  (2) from ~Jan/2013 to ~Mar/2013 (~3 months)

  (3) from ~Oct/2013 to ~Nov/2013 (~1.5 months)

I can only guess about the underlying causes of those exponential growth periods, but the last two seem to coincide with the opening of the Chinese markets: (2) by BTC-China in Hong Kong, (3) by Huobi and OKCoin in the mainland.  Perhaps (1) is the opening of the European market?

I would guess that in each case the trading population grew exponentially because it was driven by the "infection" process above.  Perhaps (2) stopped by market exhaustion?  Spurt (3), as we know, stopped because of the PBoC decrees.

So, the apparent  multi-year exponential growth in that graph is actually due to three successive market opening events, each much bigger than the other, spaced roughly 8 months apart.   Each gave a shifted exponential, instead of a pure exponential, because it was added to the strady-state of the previous ones. 

To continue that trend, it would be necessary to have another "market opening" event, even bigger than China's.  Itseems unlikely that a large country like China, Russia or India will remove its barriers.  Perhaps the "new market" will not be geogrphic but economic/social (say, Walmart accepts bitcoin, USA declares Bitcoin legal tender or whatever). Who knows.

[Davyd05]

whatever New York State can build as a bitcoin exchange can be a new exp spurt some people just crave legitimacy of government stamps of approval lol

[Peter R]

Jorge, what looks exponential on that curve is actually super-exponential.  Straight lines represent exponential growth on a log curve.  

If you plot bitcoin statistics that correlate with adoption, you'll see fairly constant exponential growth over the last 5 years:

- blockchain.info webwallets
- transactions per day excluding popular addresses
- network hashrate

The super-exponential run-ups and crashes only seem to happen with the price.

[JorgeStolfi]

That volume is most likely just bots exchanging and re-exchanging the same coins even for a possibility of a 0.00001% gain since there is no fee. Does high volume of exchanges keep the price more steady? I think so because it creates a more baseline value of each coin at that moment (multiple "people" buying/selling around the same price).

This is an interesting experiment that economists should write theses about: two markets trading the same commodity with very different fee structures, that generate very different volumes per trader and very different trading patterns.

Are they the ones that lead bitcoin price changes? I don't think so, I really think they are reactive to other bitcoin markets which involve more human transactions, even if it means that their bots have large and quick responses to subtle changes in the other exchanges making it appear although they lead the exchange price. Just IMO.

I believe that there is substantial human intervention in those markets.  For one thing, as I wrote many times before, Huobi's trade stops almost completely at night, and OKCoin's gets down to a steady background that is a small fraction of the daytime volume (and may be fake?).  Also one sees some resitance at round prices, which humans like but robots should not care about. 

Is that volume "weaker" in some sense, just because it is caused by the no-fee structure?  It seems an interesting question for the theses above...

[ChartBuddy]

Explanation

[JorgeStolfi]

Jorge, what looks exponential on that curve is actually super-exponential.  Straight lines represent exponential growth on a log curve.  

A shifted exponential A*Q**(t - t0) + B will look exponential even in a logscale plot, if B is large enough.  Here is an example with A = 20, B = 200, Q = 1.7.

Note that the function is a SIMPLE exponential, but shifted, and the Y scale is logarithmic; yet it still looks like an exponential instead of a straight line.

[lyth0s]

According to my graph below we are still at a higher price per bitcoin than would be predicted with an R squared of >0.9

Fitting trend lines to historical prices may give heart-warming plots but is not useful for predictions, unless the trends reflect some underlying fundamental variable.

Exponential trends may be caused by "infection" in a new market: as more people invest in it, the more the item is mentioned in the news and spreads by word of mouth.  Or they maybe due to the company growing, hence opening more stores/factories/products/whatever, hence growing even more. The exponential growth stops when  that market is saturated, there is no more room for the company to grow, or something else happens (e.g. devastating news) that turns investors away. 

In that chart I see three separate exponential trends, but since they are shifted (P(t) = A*Q**(t-t0) + B, instead of just P(t) = A*Q**(t-t0)) they do not look straight in the log plot. They span

  (1) from ~Jun/2012 to ~Jul/2012 (~2 months)
 
  (2) from ~Jan/2013 to ~Mar/2013 (~3 months)

  (3) from ~Oct/2013 to ~Nov/2013 (~1.5 months)

I can only guess about the underlying causes of those exponential growth periods, but the last two seem to coincide with the opening of the Chinese markets: (2) by BTC-China in Hong Kong, (3) by Huobi and OKCoin in the mainland.  Perhaps (1) is the opening of the European market?

I would guess that in each case the trading population grew exponentially because it was driven by the "infection" process above.  Perhaps (2) stopped by market exhaustion?  Spurt (3), as we know, stopped because of the PBoC decrees.

So, the apparent  multi-year exponential growth in that graph is actually due to three successive market opening events, each much bigger than the other, spaced roughly 8 months apart.   Each gave a shifted exponential, instead of a pure exponential, because it was added to the strady-state of the previous ones. 

To continue that trend, it would be necessary to have another "market opening" event, even bigger than China's.  Itseems unlikely that a large country like China, Russia or India will remove its barriers.  Perhaps the "new market" will not be geogrphic but economic/social (say, Walmart accepts bitcoin, USA declares Bitcoin legal tender or whatever). Who knows.

I agree that the trend is shifted upward due to those few peaks, but as bitcoin becomes more widely accepted I can only imagine other similar events in the future, such as the people of a small country adopting bitcoin for all transactions (whether or not the government makes it the official currency). Either way the overall relationship is very strong and I know it will effect where I think bitcoin is headed. To each person their own opinion/predictions, but I'm throwing mine out there. The fundemental value that I'm attempt to project is the rate at which bitcoin price will increase based upon the rate of public acceptance and usage of bitcoin, which is also supported by your claim that those few peaks are from the opening of new exchanges (public acceptance and usage).

By the way isn't trending historic data the root of almost all chart analysis?

[lyth0s]

That volume is most likely just bots exchanging and re-exchanging the same coins even for a possibility of a 0.00001% gain since there is no fee. Does high volume of exchanges keep the price more steady? I think so because it creates a more baseline value of each coin at that moment (multiple "people" buying/selling around the same price).

This is an interesting experiment that economists should write theses about: two markets trading the same commodity with very different fee structures, that generate very different volumes per trader and very different trading patterns.

Are they the ones that lead bitcoin price changes? I don't think so, I really think they are reactive to other bitcoin markets which involve more human transactions, even if it means that their bots have large and quick responses to subtle changes in the other exchanges making it appear although they lead the exchange price. Just IMO.

I believe that there is substantial human intervention in those markets.  For one thing, as I wrote many times before, Huobi's trade stops almost completely at night, and OKCoin's gets down to a steady background that is a small fraction of the daytime volume (and may be fake?).  Also one sees some resitance at round prices, which humans like but robots should not care about.  

Is that volume "weaker" in some sense, just because it is caused by the no-fee structure?  It seems an interesting question for the theses above...

Well personally if I were to write a bot I would most likely turn it off while I sleep in case there is some major market change that my bot responds to that I don't agree with. IE if the exchange were to undergo a process like mt gox did, or suddenly not allow fiat withdrawl etc. I would want to keep my btc, take it off the exchange and get a better rate elsewhere for example.

Also when programming humans tend to set if/then for/loops around even numbers, which may explain the seemingly human activity around whole numbers.

Edit: These are my thoughts, I have no evidence to support my claims other than anecdotal programming experience.

[686f646c]

The first part of your theme seems possible........... but likely NOT complete b/c in my thinking it would take considerable time for 750k btc to be absorbed completely into the market.  The second alternative seems even less possible b/c it seems very unlikely (though possible) that the stealing would have ONLY been by one thief.

Let's say there is a thief and his name is not Mark. Either the coins were stolen slowly or in very short period of time.

what would you do if you were the thief?

If you stole the coins slowly over time, you would have no idea when you will be discovered. also you know that once the hack is discovered the price will tank, so the best move would be to sell them as soon as you get them.

If you stole 750k coins very quickly in 2011, it wouldn't be worth much at that time, you know the price will tank when discovered and it would reasonable to believe your hack could possibly kill bitcoin forever, you best bet would be to offload it in bulk to large buyers in the underground world.

In the unlikely event 750k coins were stolen recently, that would make it the biggest heist in history, you certainly can't think you freely can sell it on any exchange, especially since mtgox knows the addresses and can follow the blockchain. Congrats you have the same problem as satoshi.

My conclusion there is no thief with 100k's of coins.

[freebit13]

Could Gox allowing users to log in and check their balances be causing such slow markets? Many seemed to have been trying to recoup their losses by buying back in, could this lull them into a false sense of security and stop them buying new coins?

[TERA]

Everyone keeps talking about this "trend line" (im guilty of it too), but what if our trendlines are actually wrong. They are all drawn after the fact, arbitrarily, using the latest data, which may be skewed high or low if we are currently at a very high or low point. What if the price doesnt actually follow that trendline, or more importantly what if our slope of the trendline is completely wrong? The correct trendline and the correct slope would be revealed only after a major price correction (like 2011) and then we'd follow the new trendline. For example, it could be like this (in purple):





We all think the blue/green is the real trendline but then it ends up actually being the purple one after a devastating correction (a REAL bear market) is revealed.

[lyth0s]

Everyone keeps talking about this "trend line" (im guilty of it too), but what if our trendlines are actually wrong. They are all drawn after the fact, arbitrarily, using the latest data, which may be skewed high or low if we are currently at a very high or low point. What if the price doesnt actually follow that trendline, or more importantly what if our slope of the trendline is completely wrong? The correct trendline and the correct slope would be revealed only after a major price correction (like 2011) and then we'd follow the new trendline. For example, it could be like this (in purple):





We all think the blue/green is the real trendline but then it ends up actually being the purple one after a devastating correction (a REAL bear market) is revealed.

I think the purple slope is scewed due to its starting point being different than the other two

[inca]

Everyone keeps talking about this "trend line" (im guilty of it too), but what if our trendlines are actually wrong. <snip>
We all think the blue/green is the real trendline but then it ends up actually being the purple one after a devastating correction (a REAL bear market) is revealed.

Why are you using Mt Gox dip down to 1xx's to formulate your trendline?

Barring a change in the US regulatory framework I think we will grind sideways with moments of volatility before slowly making our way back towards the ATH, at which point all hell will break loose.

[ChartBuddy]

Explanation