I am a black belt karate informatics expert (kara = empty, te = hand).

Today (Sunday Mar/02) trade volume was extremely low: only ~94 kBTC, the lowest since the Chinese New Year holidays (Jan/29--Feb/05).  Since that week, the volume has been below 100 kBTC only once, on Feb/19 (~95 kBTC).  Today's total was less than 1/10 of Tuesday's peak (~948 kBTC).

Compared to yesterday, volume fell 55% outside China (33 to 15), 57% in China (183 to 79), and 57% overall.  China's slice was 84% of the total.

In China, Huobi decreased the most, 61% (from 104 to 41) but was still ahead of OKCoin (35 kBTC).

Outside China, Bitstamp decreased 55% (from 14 to 6.44),  BTC-e 52% (9.55 to 4.58), and Bitfinex 64% (8.62 to 3.13).  Their ranking did not change but Bitfinex fell further behind BTC-e.

Daily volumes of BTC trade to/from USD and other national currencies (in kBTC):


             !    Sat !    Sun !    Mon !    Tue !    Wed !    Thu !    Fri !    Sat !   Sun !                     
  EXCHANGE   !  02/22 !  02/23 !  02/24 !  02/25 !  02/26 !  02/27 !  02/28 !  03/01 ! 03/02 ! Currencies considered

  Bitstamp   |  18.08 |  20.87 |  51.84 | 115.76 |  49.11 |  17.05 |  22.63 |  14.30 |  6.44 | USD                 
  BTC-e      |  18.68 |  14.24 |  20.78 |  73.86 |  33.61 |  14.60 |  12.74 |   9.55 |  4.58 | USD,EUR,RUR         
  BitFinEx   |  14.73 |  15.80 |  27.16 |  94.55 |  37.71 |  12.04 |  15.29 |   8.62 |  3.13 | USD                 
  Kraken     |   1.08 |   1.06 |   1.06 |   2.23 |   1.87 |   0.89 |   1.34 |   0.39 |  0.32 | EUR                 
  Bitcoin.DE |   0.61 |   0.50 |   0.73 |   2.23 |   1.00 |   0.50 |   0.70 |   0.31 |  0.32 | EUR                 
  CaVirtEx   |   0.20 |   0.23 |   0.30 |   2.15 |   0.32 |   0.31 |   0.33 |   0.15 |  0.19 | CAD                 
  CampBX     |   0.26 |   0.22 |   0.37 |   0.42 |   0.36 |   0.05 |   0.07 |   0.08 |  0.07 | USD                 

  SUBTOTAL   |  53.64 |  52.92 | 102.24 | 291.20 | 123.98 |  45.44 |  53.10 |  33.40 | 15.05 |                     

  Huobi      | 154.12 | 185.15 | 163.47 | 352.76 | 250.64 | 123.38 | 122.00 | 104.12 | 40.64 | CNY                 
  OKCoin     |  87.71 | 141.90 |  84.03 | 275.68 | 150.41 |  83.60 | 129.17 |  73.29 | 34.99 | CNY                 
  BTC-China  |   9.29 |  10.42 |  13.12 |  27.72 |  14.72 |   6.71 |   7.17 |   5.02 |  2.64 | CNY                 
  Bter       |   0.48 |   0.62 |   0.53 |   1.10 |   0.53 |   0.31 |   0.29 |   0.30 |  0.24 | CNY                 

  SUBTOTAL   | 251.60 | 338.09 | 261.15 | 657.26 | 416.30 | 214.00 | 258.63 | 182.73 | 78.51 |                     

  TOTAL      | 305.24 | 391.01 | 363.39 | 948.46 | 540.28 | 259.44 | 311.73 | 216.13 | 93.56 |                     



All numbers were collected by hand from the site http://bitcoinwisdom.com. Beware of possible errors.

For each exchange, the numbers include only the trade volume to/from the currencies listed in the rightmost column. Trade between BTC and other cryptocoins, such as LiteCoin, is NOT included.

Dates on the header line are UTC. Specifically, "01/15" means "from 01/15 00:00:00 UTC to 01/15 23:59:59 UTC". (Beware that Bitcoinwisdom uses your local time, so the date may appear to be off by 1 day.  For example, if you are 2 hours west of Greenwich, it may show "01/14 22:00" when the UTC time is "01/15 00:00".)

Ha! You just can't imagine.

It will be MUCH worse than that.

If you guys don't stop picking on my person, I will have to take an extremely harsh punitive measure that you will not ever forget.  You have been warned.

Chinese Slumber Method prediction for Monday Mar/03

Trade volume V_d for today Mar/02 was ~41 kBTC, and volume V_h in the official Slumber hour (19:00--19:59) was 52 BTC; so, by the new rule V_h/V_d < 0.005, it is definitely a True Slumber Point.

The price at the official Slumber hour today (Mar/02 19:00--19:59) was 3427 CNY.  There is now a clear decreasing trend among the last four True points (Feb/27--28, Mar/01--02).  Therefore, it seems better to abandon the constant trend that I had been assuming for the last segment, starting Feb/20, and assume a new linear trend starting at Feb/27.  The least-squares line of those four points is p(d) = 3521.20 - 30.30*(d - 27), where {d} is the day number since Feb/01. (This line also happens to fit the Feb/26 point, even though it is nominally a False one). Thus

Prediction valid for: Monday 2014-03-03, 19:00--19:59 UTC (not before, not after)
Huobi's predicted price: 3400 CNY.
Bitstamp's predicted price: 556 USD.

In retrospect, is not worth fitting any trend to the entire Empty Gox Chaos period that followed the Second Karpeles Catastrophe (namely, the six days Feb/20--25).

The prediction is the blue rectangle almost outside the right margin in the chart below.  The orange and grey dots are the True and False Slumber Points, and the orange line is the new trendline.


The price ratio between Huobi and Bitstamp was 6.11. That is so close to the factor R = 6.12 that I have been using, that I prefer to stick to the latter.  Thus, the prediction for Bitstamp is simply that for Huobi, divided by R = 6.12.

NOTE: I stand by my predictions, but I am not responsible for any errors that the market may make.

Checking the Chinese Slumber Method predictions for Mar/02

(It is called Chinese Slumber Method because it can put half of China's propulation to sleep.)

The last prediction was not bad but a bit too bullish, again.

Prediction posted on: Saturday 2014-03-01, 23:37 UTC
Prediction was valid for: Sunday 2014-03-02, 19:00--19:59 UTC

Huobi's predicted price: 3515 CNY
Huobi's actual price (L+H)/2: 3427 CNY
Error: 88 CNY

Bitstamp's predicted price: 572 USD
Bitstamp's actual price (L+H)/2: 561 USD
Error: 11 USD

The Huobi prediction in question is the rightmost blue rectangle on the chart below.  The light blue-gray rectangles are the previous predictions.  The orange and grey dots are the True and False Slumber Points, the mean Huobi prices at 19:00 UTC every day.  The orange line is the trend that was assumed for the prediction.


The following chart shows the Bitstamp prices and predictions. The orange and grey dots are Huobi's prices at 19:00 UTC every day, divided by R = 6.12.  The orange line is the trend used in the prediction, which is Huobi's trend divided by R (6.40 for Feb/07--09, 6.12 for all other times).

Sorry for the duplicate post, my browser went kaputt during the first attempt and I though it didn't go through.

Thanks for the link.

Right, and I forgot also that Roger Ver did a similar audit of MtGOX and certified that they were fully solvent.  

For over 10 years I have been explaining to people down here that if a fully-digital voting machine prints the total votes in two different formats, and the printouts match, it does not follow that the totals are correct.  Most people understand that, unfortunately not the people who are in charge of our voting system.

So Mr. Antonopoulos was shown two numbers by Coinbase, the total BTC balance of their clients and the total number of BTC in their cold storage wallets.  He checked the latter, and since the two numbers matched, he vouched for Coinbase not being Karpelian.

At least  he did a more thorough audit than Ver, who was shown bank statements about the value of "A" and vouched that "(A+B)-(X+Y)" was positive.

And neither Ver nor Antonopoulos told us the actual numbers...


This is a new exchange that vowed to be transparent, right?


The "etc" does not include BTC-e, does it? Or Bitstamp, or Bitfinex?

AFAIK, those three are now the largest exchanges outside China, and the rest is small crumbles.  

Thanks for the link.

Right, and I forgot also that Roger Ver did a similar audit of MtGOX and certified that they were fully solvent. 

For over 10 years I have been explaining to people down here that is a fully-digital voting machine prints the total votes in two different formats, and the printouts match, it does not follow that the totals are correct.  Most people understadn that, unfortunately not the people who are in charge of our voting system.

So Mr. Antonopoulos was shown two numbers by Coinbase, the total BTC balance of their clients and the total number of BTC in their cold storage wallets.  He checked the latter, and since the two numbers matched, he vouched for Coinbase not being Karpelian.

At least  he did a more thorough audit than Ver, who was shown bank statements about the value of "A" and vouched that "(A+B)-(X+Y)" was positive.

And neither Ver nor Antonopoulos told us the actual numbers...


That site is unreachable now, may be a problem with my network here.  Will check when it comes up.  Isn't that the new exchange that vowed to be transparent?


The "etc" does not include BTC-e, does it? Or Bitstamp, or Bitfinex?

AFAIK, those three are now the largest exchanges outside China, and the rest is small crumbles.   

Thanks.  It loads now.

(The page went blank it seems.)

Very interesting.

Didn't Karpeles brag somewhere of having been through tight situations that we cannot imagine, or something like that?

Well... none of the other exchanges has volunteered to give even a rough idea of their financial position, especially how many BTC they owe and how many BTC they own. 

I count that as "real and material" evidence that they must be in some sort of Karpelian situation themselves. 

Oops, one of the lemmings got scared by something and ran off the cliff...

Or simply an indication that it is Sunday, and many traders remembered that they have a family... 

I guess the reporter had to make up something positive quick, to make good on the payola she got from the store.

Huobi has been the largest exchange in the world by volume, by far, for several months.

Its volume has been steadily shrinking since the historic records Tuesday Feb/25, like that of all exchanges.

That historic record was apparently caused by the week-long turmoil in all markets following MtGOX closure.

China has 60% to 80% of the total trade volume, but this percentage goes up and down. It has been increasing of late, but mainly because trade outside China has shrunk more than trade in China.

Its trade volume is usually 8x to 10x that of Bitstamp. If that translates into liquidity too, then Huobi will have a much larger effect on Bitstamp than the other way around.

Namely, when an arbitrager buys an amount of BTC on Huobi and sells the same amount on Bitstamp, the drop at the Bitstamp is usually much bigger than the rise at Huobi.  So the effect of arbitrage is to pull Bitstamp closer to Huobi, rather than the other way around.

The problem on Safari occurs only with hours "8" and "9", on the 1h charts, they display as "0" on the crosshair cursor's bottom tab.  It does not occur on the 30m charts, so "8:0" and "8:30" display correctly.

It occurred to me that in some languages/systems, a number that starts with "0"  is interpreted as octal rather than decimal.  In those platforms, "00"through "07" give the same result in octal and decimal,  "11" to "23" are interpreted as decimal, but "08" and  "09" would be invalid.

Perhaps Safari's javascript differs from the javascript in other browsers in that detail?  Sometimes the interpretation of numbers that start with "0" is determined by a global variable, and that variable may have a different default value on Safari?

I do not know whether the Chinese use robots or semi-automatic tools.  But I would not call them "high-fequency".  This term usually means a trading robot that does many transactions per second, to exploit very small margins.  All Huobi clients together do perhaps a few hundred transactions per minute, and many of these look rather human (with round numbers, for instance).