The Chinese traders and their robots have decided that 1 BTC is worth 4800 CNY, because they have realized that the Chinese traders and their robots have decided that 1 BTC is worth 4800 CNY, because they have realized that the Chinese traders and their robots have decided that 1 BTC is worth 4800 CNY, because they have realized that the Chinese traders and their robots have decided that 1 BTC is worth 4800 CNY, because they have realized that the


(image from http://www.boston.com/bostonglobe/ideas/theword/2009/01/bootstraps_and.html)

Volume of BTC trade to/from USD and other national currencies (kBTC):


            01/14  01/15

  Bitstamp   12.3    8.2
  BTC-e       9.8    7.2
  MtGOX       6.4    4.7
  BitFinEx    7.1    3.7
  Bitcoin.DE         0.4
  CaVirtEx           0.3
  Kraken             0.3
  CampBX             0.1
  CryptoTrade        0.0

  SUBTOTAL   35.6   24.9

  Huobi      91.4   70.2
  OKCoin     22.2   24.8
  BTC-China   5.6    3.5

  SUBTOTAL  119.2   98.5

  TOTAL     154.8  123.4


Numbers collected by hand from the Bitcoinwisdom charts and added by hand, may be wrong. 

They do not incluce trade between BTC and other cryptocoins such as LiteCoin.

The Coinbase volume is not available at Bitcoinwisdom.

About the FBI bitcoins:  I live far from the US and have no inside knowledge, but based on my understanding of how governments work, I would guess that:

The people in charge of the sale of those bitcoins do not stand to profit from it (at least legally).

Therefore, they do not particularly care to get the best possible price for them.

Their primary concern is to follow the rules so that they won't be charged with blatant errors, negligence, favoritism, etc.  Selling government property through private deals is highly against the rules, I can't imagine them doing it.

They will not get blamed if the coins are auctioned for $500 and the buyers immediately resell  them for $1000; that sort of thing happens all the time in public auctions.

They will try to avoid obvious mistakes that they can be blamed for (like failing to publicize the auction or doing so with too short notice).  But they will not look for smart trading strategies, like selling litte by little or waiting for the market price to rise.   (For a civil servant, being dumb is OK; trying to be smart and failing is very, very bad.)

They are also not concerned about what the auction could do to the market price.  A speculator's loss is always someone else's profit, so a price crash will only affect the distribution of wealth, not its sum. 

On the other hand, making hoarded goods available for use or consumption is always a good thing for the world as a whole.  (Granted that it may not apply to bitcoins, given that their utility is independent of the quantity in circulation.)

Thus I expect that they will auction the SilkRoad stash in largish lots (just small enough to have competitive auctions), and let the market do whatever it will do.

You don't need to actually do the transaction on the blockchain, or wait for confirmation, if the accounts in the two exchanges belong to trusting partners (or to the same person) and the receiving account has enough balance to keep on trading as if the transfer has taken place. 

Just as when you draw $100 cash from an ATM of your bank in Mongolia, you do not have to wait until your cash is physically shipped there from your home country, or even wired to Mongolia through the central banks.  Your bank just moves numbers in its ledgers.

Such transfers can be kept virtual and private as long as the Mongolia branch has (or can borrow) enough cash locally to keep its ATMs stocked, and is confident that the balance of all virtual transfers can be realized eventually.
Your USD withdrawals from MtGOX may take weeks or years, but the exchange owners probably can transfer money from their bank (or take out a suitcase full of cash) in a couple of hours, if they decided to do so.

One must be aware that cryptocoin exchanges are still unregulated, unaudited, and have no ethical standards to uphold.

Could it be be a client with special privileges? Like, someone inside MtGOX and/or Bitstamp?

Arbitrage trading requires the ability to transfer cash or coins quickly between markets. One way, as the poster suggested, is to have a pile of coins and a pile of cash in each exchange. Then one can do instantaneous "virtual transfers" if required by the chosen arbitrage strategy, and do real transfers later, once in a while, only to rebalance the piles.  Does this make sense?

According to Wikipedia, in mainland China the 2014 Chinese New Year holidays are 31 January – 6 February
http://en.wikipedia.org/wiki/Chinese_new_year

I wish I was! But no, Gimp is the work of many great Gnu/Linux hackers out there...

use "Export" rather than "Save" to save in a format other than .xcf

So that the Bears do not feel slighted by that extrapolation wallpaper:



(At its peak, WorldCom was worth over 100 billion USD. AFAIK it was the largest bankruptcy in US history.)

EDIT: bitcointalk seems to be truncating the image for some reason.  You may get it here:
http://www.ic.unicamp.br/~stolfi/temp/worldcomm-closing-price-log-extrap.png

Volume of BTC to/from USD and other national currencies, for 2014-01-14 (kBTC):


  Bitstamp   12.3
  BTC-e       9.8
  BitFinEx    7.1
  MtGOX       6.4

  SUBTOTAL   35.6

  Huobi      91.4
  OKCoin     22.2
  BTC-China   5.6

  SUBTOTAL  119.2

  TOTAL     154.8


Numbers collected by hand from the Bitcoinwisdom charts, may be wrong.

The Coinbase volume is not available at Bitcoinwisdom. Other exchanegs seem to have much less than 500 BTC volume each (not included above).

The Huobi volume is somewhat below its daily average; the same may be true for the other exchanges as well.

... and UP ~ 100 CNY/day since 12/17, and DOWN ~ 100 CNY/day since 01/05, and ...

 

Well, it depends on when you start to compute the mean...

Start at 2013-01-01 and the mean of log increments is positive; start at 2013-11-18, and it is zero; start at 2013-11-29, and it is negative. 

If the past is relevant, the recent past (say, the last two months) should have more weight than the remote past, no?  So why is the mean since 2011 more likely to be correct than the mean since 2013-11-29?

But if the mean increment is indeterminate to that degree, it becomes part of the zero-mean random term, just increasing its variance.

That observation is the key to rising from Level One to Level Two.

BTW, even if one starts at 2011, the mean increment per time step is not "strongly" positive: it is still small compared to the standard deviation of the random term.

Another observation:

The plots of the main indicators (running means) look displaced from the actual price plots.  The reason is that a mean price that is computed from data up to a certain time is "logically located" at some earlier time, which is the mean of the weight indices.

Specifically, suppose a mean price V is computed by V = SUM{ W(k)*P(t-k) : k in 0..N-1 }, where P(i) is the price at time i and W(0..N-1) are weights that add to 1.  Then V should be plotted at time t - m where m = SUM{ W(k)*k : k in 0..N-1 }.   

If the weights are positive (say, uniform 1/N or exponentially decreasing) then m is positive, so the plot should be displaced to the right.  In that case, the indicator V is in fact using data up to time t to "predict" a price in the past, namely P(t-m).

To get a true predictor for the price P(t+1), one must use positive and negative weights that still add to 1 but are such that the mean index m, as defined above, is -1  For example W(0)=2 and W(1)=-1 will give the simplest (two-point) linear extrapolation predictor. The mean V computed with these weights should then be plotted at time t+1 rather than t.

Levels of Enlightenment for market trading:

  * At Level of Enlightenment Zero, you buy or sell at current price based on what other people tell you.

  * At Level of Enlightenment  One, you draw lines, triangles and dinosaurs over the charts, look for trends and patterns, in an attempt to predict future prices from the set of past prices.

  * At Level of Enlightenment Two, you realize that the market does not give a damn for the past price history.  Any trends are due to the trading community's perception of external events. When the external conditions change (e.g. when the traded item is banned in Mongolia, or featured in a TV serial), the trends immediately change in ways and amounts that are very hard to predict.  The best one can say is that the price P(now + t), in log scale, is P(now) plus a random variable with zero mean whose variance increases linearly with t, irrespective of its past history.

  * At Level of Enlightenment Three you know more complicated formulas that use past prices to give predicitions for future prices that are significantly tighter than those of the Level Two model above. But I haven't got to that Level yet.

I suspect that there is a maximum Level of Enlightenment N-1, after which one's understanding of markets goes back to Level Zero.

Huobi's volume was abnormally low in their morning hours, and price was steady at 5100 BTC since their previous night. Then volume jumped up and price plunged to 4925 (eating a 300 BTC wall @ 5000 in a few seconds).

Indeed, in the 1h charts their volume is very low today, compared to that of the previous 4 days
(which varies with local time in a characteristic way).

It is even lower than the volume on Jan/10, another anomalous day.

The price also has been surprisingly stable for the last two days.  Price stability and low volume, which is cause and which is effect? Or are both effects of some external cause?

Yes.  Question 1: Why stop at 2017?



Question 2: What will I be able to buy with 10,000,000,000,000,000 dollars in 2024?

 

Thanks! My "vertically shift" suggestion was to format it as follows

except with smaller dots (or dashes) between the two columns...

As for which end of the interval is open or closed: it will be confusing to readers either way, so it is indifferent.

The Salomonic solution is to split orders that are eactly equal to the break price (say 4850) and count half their volume on each side of it (half above and half below 4850).  But this solution may be even more confusing to readers...

However, the criterion should be symmetric around the spread, since one usually reads the asks upwards and the bids downwards. 

Thanks again, and all the best.

The poor do not undertand speculative markets and are asily swindled.  If they learn that someone got rich by investing in X, many will sell their houses and sustenance cows to invest in X.  (TelexFree, a recent Brazilian Ponzi scheme, was most popular in the poorest parts of the county. I have seen claims that half the adult population of a state in the deep Amazon invested in it.)

It was talked several months ago, I still prefer the consistency to consider it is break or not.
The graphic looks like[...]
So it won't be changed, maybe would be an option.
[/quote]
Why not shift the price column down by half a line? Then it would look sort of like this


           4900 -
                  653
           4850 -
                  233
           4800 -
                  25
4778.22
                  12
           4750 -
                  63
           4700 -
                  423
           4650 -
                  1024

except that it would be squeezed vertically so that both columns are single-spaced.