Rally from $500 happened on very low volume, crash looming?

[rpietila]

Data used: Bitstamp's hourly volume weighted average historical trades Nov 1, 00:00 - Nov 30, 09:00.

Total volume during that period: BTC953,782.

Code:

price        volume     mid point    BTC      $m
$200-299     BTC215,619 $250.00   215619 $53.90
$300-399     BTC208,384 $350.00   208384 $72.93
$400-499     BTC135,057 $450.00   135057 $60.78
$500-599     BTC98,430  $550.00    98430 $54.14
$600-699     BTC86,344  $650.00    86344 $56.12
$700-799     BTC46,202  $750.00    46202 $34.65
$800-899     BTC73,779  $850.00    73779 $62.71
$900-999     BTC47,990  $950.00    47990 $45.59
$1000-1099   BTC36,849  $1050.00   36849 $38.69
$1100-1199   BTC5,127   $1150.00    5127  $5.90

During November, 59% of the trades (by BTC volume) happened at prices lower than $500. Conversely, only 9% happened on $900 or higher. The average price for all trades was $517.

[Carra23]

It will be more useful if you can predict where the pullback will go to and from where. We all know that at some point it will go down a bit.

[fred1111]

It could mean the opposite, too. When everyone waits for a crash, expect many trapped bears.
Thanks for the data!

[traderCJ]

I see it differently ..

The price/volume ratio only varied by a factor of two over all of your data, except the last datapoint.  There could be several factors to blame:

1) Breaking over into the low 1000s occurred during Thanksgiving week.  Wait until mid next week when fiat clears the exchanges.
2) There is considerable psychological resistance to buying anything "virtual" and "faith based" that is considerably more than $1k (irrational market participants).

[mrjeff]

Is this data directly available from bitstamp or was it sourced elsewhere?  Just curious where to look for the raw data.

[tutkarz]

the amount traded bitcoins is decreasing because the price is higher but traded dollars are more or less the same

$200-299   BTC215,619 200x215000 = 43 mln
$300-399   BTC208,384
$400-499   BTC135,057
$500-599   BTC98,430
$600-699   BTC86,344
$700-799   BTC46,202
$800-899   BTC73,779
$900-999   BTC47,990
$1000-1099   BTC36,849 1000 x 36000 = 36mln
$1100-1199   BTC5,127

(don't have time to calculate all)

also i don't know if you should make percent based groups not 100 dollars

[Kj1]

could you work in percentages?

eg 200 -> 220 (+10%): volume =
220 -> 242 (+10%): voluime = ...

Also, using USD instead of BTC would help.
Check bitcoinity options (show volume in Currency)
plot graph over 6m, you'll see the volume keeps increasing

[sgbett]

Code:

                        mid point    BTC      $m
$200-299     BTC215,619 $250.00   215619 $53.90
$300-399     BTC208,384 $350.00   208384 $72.93
$400-499     BTC135,057 $450.00   135057 $60.78
$500-599     BTC98,430  $550.00    98430 $54.14
$600-699     BTC86,344  $650.00    86344 $56.12
$700-799     BTC46,202  $750.00    46202 $34.65
$800-899     BTC73,779  $850.00    73779 $62.71
$900-999     BTC47,990  $950.00    47990 $45.59
$1000-1099   BTC36,849  $1050.00   36849 $38.69
$1100-1199   BTC5,127   $1150.00    5127  $5.90

you all posted whilst i was building the table... but yeah usd volume is interesting... maybe only support until 1000-1100 range.

be nice if thats the new floor

[keewee]

Data used: Bitstamp's hourly volume weighted average historical trades Nov 1, 00:00 - Nov 30, 09:00.

Total volume during that period: BTC953,782.

Price;     Volume               USD Value
                                      (approx. millions)
$200-299   BTC215,619      53
$300-399   BTC208,384      73      (186 below $500)
$400-499   BTC135,057      60
$500-599   BTC98,430        54
$600-699   BTC86,344        56
$700-799   BTC46,202        34
$800-899   BTC73,779        62      (296 above $500)
$900-999   BTC47,990        45
$1000-1099   BTC36,849     39
$1100-1199   BTC5,127       6

During November, 59% of the trades (by BTC volume) happened at prices lower than $500. Conversely, only 9% happened on $900 or higher. The average price for all trades was $517.

I've bolded the USD figures for those trades. and that has remained pretty consistent until $1100 was breached. Also, Nov 30 has only just begun in the US and we don't know what sort of trading day we will see yet. About US$100 million more has been traded above $500 than below. I think the drop in btc volume is probably because people are still thinking of the value of btc in $US terms


Edit: sgbett and I had the same thought 

[sgbett]

It will be more useful if you can predict where the pullback will go to and from where. We all know that at some point it will go down a bit.

One year ago I predicted that it will cross $100, hit $300, crash, and stabilize several months at around $120.

Then I earned the 1000s of btc that I sold 2 weeks ago

This time it is possible to be different. I try to give good advice to become a more trusted member of the community.

If history teaches anything, its that "past performance is not a guarantee of future returns".

What I mean is that everyone is expecting the run-up/crash cycle we have seen before, and so the market being the market, it will probably do what is least expected

Which I think is what you are saying - in which case I agree!

[Vycid]

Thanks for collating, data is always good.

I think it's extremely important to keep in mind that MtGox has quite recently started offering 0% trade fees. So it's really not so much of a surprise that the Bitstamp volume from $1100-$1199 has been thin.

[oakpacific]

No money in !=more money out I guess.

It's difficult for the money to both get in and out thus the low volume and high price.

[rpietila]

1. It is undeniable that volume has dropped both in BTC and in USD terms. It is not insignificant that it drops by 20%, 30% or so. These are the warning signs that should not be discarded.

2. It would be mathematically better to work on percentages, but I wanted a visually pleasing way + there was a technical reason (truncate to -2 decimals) that I chose this one.

3. Data is public in:
Bitcoincharts.

4. In Bitstamp, Thanksgiving is not a big thing.

5. The lowered volume indicated that there is an overhang on the sell side. Current holders are unwilling to sell as much as is otherwise rational, because they want a better timing. That we went +500% and they are still waiting.

6. USD volume has certainly increased during this period. This proves that there has been much new interest. Since the price has been climbing exponentially (with a daily exponent of 6.3% over 27 days), they have generally invested all of it quite quickly. This can change in a heartbeat.

7. On average, bitcoin's value appreciation works so that during every double, existing holders sell 17%. In November this has definitely not been so. There will have to be a constant exponential increase of new money coming in (via a dysfunctional or non-existent infrastructure), or a crash. Probably both. And several times a year. At this point, anyone who denies the existence of either, is delusional, and deserves the pain.

[rpietila]

you're hoping for a crash to get back in 

Pros don't hope. They adjust the bets so that they win regardless. (If they are in the position to move the markets, they will)

[mrjeff]

Thanks, I was just looking at the data and it seems.

• We've had about 14 hours of trading where the weighted average price was between $1100 and $1199.
• We've had 80 hours of trading where the weighted average was a price of $800-899
• We've had 180 hours of trading where the weighted average was between 200 and 300.

.

I'm not a statistician but I'm curious if the hours of trading at a particular price should be noted at all when looking at total volume for a certain price level?

[accord01]

It's only been at 1000+ for 24 hrs?  Of course volume is going to be lower.  Future "crashes" won't be as apparent as time go on and will look more and more like corrections, because whales aren't dumb enough to lose so much value for their coins. It is so much more profitable, due to the strength of the market, to sell in smaller amounts preferably at peak times.  Any major sell off at low liquidity times for less value is either the result of boredom or malice (manipulation), and it will be less effective as time goes on.  

So it looks like if you sold at anything below 900, well maybe buy back in asap, and short it at 1800?  Other wise, choo choo mutha fucka!

[elebit]

Low volumes on Bitstamp are to be expected, since the rise from 500 to 1000 took place on BTCChina.

Both Gox and Stamp were mostly silent as they watched the Chinese bull run. Of course the price takes a discontinous jump then.

It would be interesting to see if your numbers still hold true if you look at the four big exchanges together. I suspect that trade volume was larger this time than in March (thereby implying that the correction will be shorter and faster) but it would be interesting to find out if that's true!

[trace666]

While Bitstamp is an important part of the BTC market, it is just that.
To come to any conclusions here, we need to look at a bitter share of the total volume.
Here is an indication of the current distribution of BTC trading volume:

BTCChina	44%
MtGox (USD only)	16%
Bitstamp	14%
Other exchanges	26%

This is based on bitcoincharts.com data.

[rpietila]

Thanks, I was just looking at the data and it seems.

• We've had about 14 hours of trading where the weighted average price was between $1100 and $1199.
• We've had 80 hours of trading where the weighted average was a price of $800-899
• We've had 180 hours of trading where the weighted average was between 200 and 300.

.

I'm not a statistician but I'm curious if the hours of trading at a particular price should be noted at all when looking at total volume for a certain price level?

That depends what you want to prove. I want to raise the awareness of the scenario that we have been going too far too fast, which means that odds are no longer wildly stacked on our favor (as usual), just reasonably stacked. Both the fact that we cross price levels quickly and that the volume is low, testify to the increased risk.

We are at 3.5x the exponential trendline just now. Iirc, the intraday bubble top in April was 4.3x. Especially worrisome thing is that not even one is analyzing these things with any scientific/mathematical fashion except me. The complacency has indeed reached the levels of April, where any newb thinks that he has become a genius by buying in 9 days before, and any cashout by older members is ridiculed.

[Vycid]

While Bitstamp is an important part of the BTC market, it is just that.
To come to any conclusions here, we need to look at a bitter share of the total volume.
Here is an indication of the current distribution of BTC trading volume:

BTCChina	44%
MtGox (USD only)	16%
Bitstamp	14%
Other exchanges	26%

This is based on bitcoincharts.com data.

Wow, thank you, I did not know it had become so lopsided.

What is the date range for this?