Shared publicly  - 
 
a survey of Bitcoin charts, 2012-07-22

The price of Bitcoin has zoomed up in the last couple of months, since the end of May, from around $5/ⓑ to around $9/ⓑ:

http://bitcoincharts.com/charts/mtgoxUSD#tgSzm1g10zm2g25

This doubtless both reflects increased interest in Bitcoin from the broader world and also stimulates increased interest.

Here are some other measurements related to Bitcoin.

currency exchange trade volume

First, trade volume on the currency exchanges. This is the sum of the values of trades of Bitcoin in exchange for some other currencies, summed up over the course of one week. The overall pattern is that trade volume has been increasing in recent weeks and is at or near and all-time high for most currencies, but not for USD.

BTC ↔ USD — ~$3.0M/week

http://bitcoincharts.com/charts/mtgoxUSD#rg360zigWeeklyzczsg2011-01-01ztgOzm1g2zm2g25zvzcv

In the most recent week for which there is complete data, more than $6.5M changed hands in trades of Bitcoin for United States Dollars. Adding in some other exchanges which also allow trades of BTC for USD (not shown), it comes up to about $7.0M in the most recent week. This, however, is an outlier and to my eyes there is no major trend over the last year. The more typical weeks saw less than $3.0M per week.

BTC ↔ EUR — ~$370K/week

In constrast to USD, for which a single exchange ("mtgox") dominates and accounts for more than 90% of all trades, there are several exchanges handling a substantial fraction of trade of Bitcoin for Euros:

http://bitcoincharts.com/charts/btcdeEUR#rg360zigWeeklyzczsg2011-01-01ztgOzm1g2zm2g25zvzcv

http://bitcoincharts.com/charts/intrsngEUR#rg360zigWeeklyzczsg2011-01-01ztgOzm1g2zm2g25zvzcv

http://bitcoincharts.com/charts/mtgoxEUR#rg360zigWeeklyzczsg2011-01-01ztgOzm1g2zm2g25zvzcv

Summed together, these show a clear trend of increasing weekly trade volume of ⓑ↔€, from ~€90K (~$120K) per week through most of the year 2012, rising over the last couple of months to ~€455 (~$550K) in the most recent week.

BTC ↔ SLL — ~$185K/week

For reasons obscure to me, one of the biggest currencies that is exchanged for Bitcoin is "Linden Dollars" from the "Second Life" multiplayer virtual world:

http://bitcoincharts.com/charts/virwoxSLL#rg360zigWeeklyzczsg2011-01-01ztgOzm1g2zm2g25zvzcv

The trend is upward, the trade volume has been more consistent week after week than most other currencies, and the volume for the most recent week was almost SLL 50M (~$185K) per week.

If I'm eyeballing this page correctly:

https://www.virwox.com/

It looks like SLL is exchanged for EUR at a volume of ~€100M (~$80M) per week‽ And USD at ~35M per week‽

BTC ↔ GBP — ~$125K/week

There are two exchanges where you can trade Bitcoin for British Pounds:

http://bitcoincharts.com/charts/mtgoxGBP#rg360zigWeeklyzczsg2011-01-01ztgOzm1g2zm2g25zvzcv

http://bitcoincharts.com/charts/intrsngGBP#rg360zigWeeklyzczsg2011-01-01ztgOzm1g2zm2g25zvzcv

The trend is definitely up over the course of the year, recently reaching ~£200K (~$125K) per week.

BTC ↔ CAD — ~$60K/week

There are two exchanges where you can trade Bitcoin for Canadian Dollars:

http://bitcoincharts.com/charts/virtexCAD#rg360zigWeeklyzczsg2011-01-01ztgOzm1g2zm2g25zvzcv

http://bitcoincharts.com/charts/mtgoxCAD#rg360zigWeeklyzczsg2011-01-01ztgOzm1g2zm2g25zvzcv

Adding the two together shows a definite trend of increasing trade volume. The most recent week is an outlier at ~CAD 100K (~USD 100K), with the weeks before that being at about ~CAD 60K (~USD 60K) per week.

BTC ↔ CNY — ~$24K/week

http://bitcoincharts.com/charts/btcnCNY#rg360zigWeeklyzczsg2011-01-01ztgOzm1g2zm2g25zvzcv

In the last two weeks of May there was an unusually high volume of trades of ⓑ↔¥. The last week of May saw ~¥750K exchanged (~$120K). However, even disregarding those two weeks as outliers, it looks to me like the weekly volume of  ⓑ↔¥ has grown from ~¥50K (~$8K) to ~¥150K (~$24K).

That's nothing compared to the volume of ⓑ↔$, of course, but it is an interesting trend, and it is interesting to wonder who is trading Bitcoin for Chinese Renminbi and why.

total transaction fees per day and estimated transaction volume per day

http://blockchain.info/charts/transaction-fees?timespan=60days&showDataPoints=false&daysAverageString=1&show_header=true&scale=0&address=

This is anemic! Serving as a transaction processor ("miner") currently means you are competing with all the other transaction processors on the globe for a share of a measely ~ⓑ20 (~$180) per day.

(The transaction processors currently also get, independent of transaction fees, a reward of ~ⓑ7200 newly minted Bitcoin (~$65K) per day shared out among them probabilistically in proportion to their computational contribution. That reward is due to become smaller and smaller over time, starting later this year when it will drop to ~ⓑ3600/day.)

Contrast with this chart which shows how much money people are moving from one person to another by means of Bitcoin payments:

http://blockchain.info/charts/estimated-transaction-volume-usd?timespan=60days&showDataPoints=false&daysAverageString=1&show_header=true&scale=0&address=

This number can't be known for certain, so what we are looking at here is an estimate provided by the folks at blockchain.info. They estimate that ~$1.5M worth of Bitcoin per day changes hands.

(Note: all of the "currency exchange trade volume" discussion above is about people trading BTC for other currencies, like USD. This one is about people sending BTC to one another over the Internet.)

nodes connected to the network

http://bitcoinstatus.rowit.co.uk/

This has declined throughout 2012 but seems to be leveling off around 15,000. Note: this does not mean there are 15,000 users or customers of Bitcoin! This number would probably be better understood as the number of redundant servers on the Bitcoin network. Any one server could probably serve at least tens of thousands of users/customers. The fact that there are 15,000 redundant servers distributing the ledger across the world and cross-checking one another's correctness — that is awesome.

total combined power of transaction processors (miners)

http://bitcoin.sipa.be/

The total combined hashing power dedicated to validating Bitcoin transactions and preventing double-spends is approaching the high that it experienced during the first Bitcoin boom in mid-2011. This is while the price of Bitcoin in USD is still around 1/3 of what it was at its height that summer. This seems like an encouraging sign -- the amount of hash power dedicated to ensuring the safety of transactions is as high as it has ever been even without the high reward and high stakes.

transaction confirmation times

http://bitcoinstats.org/

This is a live view of the network so it changes every time you look at it, but it typically shows that thousands of transactions have been posted in the last few hours, that most of them were confirmed within the first half hour, and that a few of them took hours to be confirmed.

Bitcoin market capitalization in USD

http://blockchain.info/charts/market-cap?timespan=all&showDataPoints=false&daysAverageString=1&show_header=true&scale=0&address=

This number is the total number of Bitcoin generated up until a certain time multipled by the USD/BTC exchange rate at that time. This is a better number to look at than just the price (which we started with at the beginning of this note), because new Bitcoin is generated continuously at a rate of about ⓑ7200 per day. So if the price held constant, this would suggest that the demand for Bitcoin was increasing at the same rate as this increase of supply.

On the other hand, this number does not reflect Bitcoin that is lost forever due to people accidentally deleting their private keys. Unfortunately, we have no measure of how much that has happened. That phenomenon means that there is an unknown rate at which the supply of Bitcoin is being reduced, counter to the known rate at which the supply is increasing.

Anyway, the graph is very interesting.

The absolute value of USD on the Y axis needs to be taken with a grain of salt — multiplying the total volume of Bitcoin ever created until today times the most recent trade price for USD suggests that the value of all Bitcoin ever created is ~$85M. But that statement is kind of meaningless — if you owned all of the Bitcoin in the world, you couldn't sell it for $85M. So, that number is meaningful only as a reference to itself at other times, if that.

previous posts in this series

Here are some of my greateat hits/misses in posting about Bitcoin:

• 2011-11-14 "Oh boy, the price of Bitcoin in USD has fallen to $2.20.com. If I had any more USD to risk I would be all over that." — https://plus.google.com/108313527900507320366/posts/dPvh9mZdpmG

• 2011-12-19 Bitcoin rising from its own ashes — https://plus.google.com/108313527900507320366/posts/QDga9NM3Dgp

• 2011-12-22 the Bitcoin protocol in one lesson — https://plus.google.com/108313527900507320366/posts/GFrJHiYfPAG

• 2012-04-30 Bitcoin miners deploying FPGA — https://plus.google.com/108313527900507320366/posts/2ztAhLnXQKm

• 2012-05-11 a Botnet herder mining Bitcoin — https://plus.google.com/108313527900507320366/posts/1oi1v7RxR1i


Disclosure: the author owns some Bitcoin
19
4
Adam Langley's profile photoGavin Andresen's profile photoZooko Wilcox-O'Hearn's profile photoGlenn Willen's profile photo
20 comments
 
"for a share of a measely ~ⓑ20 (~$2) per day"

That's ~$200/day, no? Still measly to be sure.
 
+Zooko Wilcox-O'Hearn I assume when you talk about transaction processors fighting over 20BTC/day, you're deliberately leaving out what they're actually fighting over right now, which is the 7200BTC newly produced per day, on the grounds that the latter source won't last. But it's important to note that anyway, I think.

Also, note "$24_k_/week" for CNY, and "~¥150K (~_¥_24K)."
 
You think that kind of volatility is a good thing?
 
The rise in BTC/SLL trading is likely due to the persistent difficulties in BTC/USD exchange. For various reasons, it can be much more expedient to convert BTC->SLL->USD than BTC->USD.
 
Glenn: thanks for the error correction. Fixed. And you are exactly right about my leaving out the reward on the grounds that it won't last, and that I should have mentioned it anyway.
 
Glenn: Okay I added a paragraph about the reward and made a few other minor edits.
 
Ramon: that's interesting. If that's the major reason for it, then I wonder if BTC ↔ SLL's ~$185K/week should mostly be added to the BTC ↔ EUR number (since SLL is apparently traded for EUR more than for USD).
 
I'm hoping we won't see a repeat of last summer's price bubble every time a new country discovers Bitcoin.  And RE: transaction fees: fixing miners and clients so there is a market price for transaction fees develops is very high on the development priority list; see https://gist.github.com/2961409 
 
Good Review of the Exchange flows.  Can you please do this every month?
 
Thanks for you thorough review of the current state ofthe bitcoin market.
One small remark,

Summed together, these show a clear trend of increasing weekly trade volume of ⓑ↔€, from ~€90K (~$70K) per week through most of the year 2012, rising over the last couple of months to ~€455 (~$370K) in the most recent week.
90k euro is around 120k usd at the current rate, 455k euro is around 555k usd.
 
Henk: thank you for the error correction! Fixed. Sheesh that was a bad one. Okay, everyone who looked at my original article please note the multiple errors detected by other readers so that you are not misled.
 
Patty: I'm unlikely to put in the time every month to write something like this, or even to remember every month to think about it. :-)
 
Gavin: FWIW, I in general like your ideas about making a market for transaction fees. However, also FWIW, I find it terrifying that we have to risk making all Bitcoin users vulnerable to remote exploit (i.e., we have to make changes to the mainline bitcoin code) in order to deploy such improvements.
 
If you have a magical way of improving software without changing it, I want to hear about it!  For this particular change, I'm much more worried about making a change that is incompatible with old software and causes people's transactions to never confirm or a transaction-spam-DoS attack than introducing a remote exploit.
 
Well the sort of magic that I'm imagining is along the lines of the "authority separation" that you and the other Bitcoin developers have already been working on. Bitcoin that requires multiple signatures to be transferred is one. If you had that, would it be possible to upgrade the software in one of your Bitcoin processes but not the other, and have the new one execute the new improved transaction-fee-choosing algorithm, without being vulnerable to any bug in the new one robbing you of your entire stash?

I guess for this particular improvement a simpler separation would work: if you use software to spend your Bitcoin, just don't upgrade that software! Miners can upgrade and start discriminating among transaction fees in the new way, without controlling any Bitcoin that could be extracted by a thief. Payer processes, who control Bitcoin that is being hunted by thieves, would probably be safer if they didn't change the software they are running.
 
Jon: that's why I tried to make it clear that this is not a measure of users, but is perhaps better understood as a measure of servers.
Add a comment...