78 percent of Bitcoin currency stashed under digital mattress, study finds

[Eugen Leitl]

http://arstechnica.com/tech-policy/2012/10/78-percent-of-bitcoin-currency-stashed-under-digital-mattress-study-finds/


78 percent of Bitcoin currency stashed under digital mattress, study finds

Significantly fewer digital coins are in circulation than previously
presumed.

by Dan Goodin - Oct 17, 2012 12:00 pm UTC

Enlarge / A long chain of Bitcoin transactions, showing each address putting
aside a small number of digital coins, helps illustrate a finding that 78
percent of them remain dormant in "savings accounts."

Dorit Ron, Adi Shamir

More than three-quarters of the digital coins in the Bitcoin digital currency
scheme aren't circulating because they remain dormant in user accounts that
have never participated in outgoing transactions, a recently published study
has found.

The figure translates to more than 7.019 million BTCs, the term used to
denote a single coin under the digital currency, which uses strong
cryptography and peer-to-peer networking to enable anonymous payments among
parties who don't necessarily know or trust each other. Based on exchange
rates listed on Mt.Goxbthe most widely used Bitcoin exchangebthe coins have a
value of more than $82.87 million. On May 13, the date the researchers
analyzed their data, there were slightly more than 9 million BTCs in
existence.

Mathematician Dorit Ron and Adi Shamir (the "S" in the widely used RSA
cryptography scheme) arrived at that finding by downloading the entire
Bitcoin history and following the trail of some 180,000 transactions. They
found there were about 3.12 million accounts, which are known as "addresses"
in Bitcoin parlance. They belonged to about 1.5 different owners, on average,
since there's no limit on how many addresses a single individual may possess.
More than 609,000 of those addresses had received a significant portion of
the outstanding BTCs without once making a payment.

"However, if we sum up the amounts accumulated at the 609,270 addresses which
only receive and never send BTCs, we see that their owners have actually put
aside in some kind of 'saving accounts' 7,019,100 BTCs, which are almost 78
percent of all existing BTCs," the researchers wrote. Almost 60 percent of
those coins were "old," meaning they were received more than three months
prior to the May 13 cut-off date for the project.

"This means that there are much fewer BTCs in circulation than previously
presumed," the researchers continued. "Yet, the total number of Bitcoins
participating in all the transactions since the establishment of the system
(except for the actual minting operations) is 423,287,950 BTCs. This implies
that each coin which is in circulation had to be moved a much larger than
expected number of times."

The report, titled Quantitative Analysis of the Full Bitcoin Transaction
Graph, isn't the first time researchers have analyzed the full available
history of Bitcoin transactions. A paper presented earlier this year by
Fergal Reid and Martin Harrigan, demonstrated how it's possible to identify
some Bitcoin users despite some of the scheme's anonymity features. Research
presented at last year's Chaos Computer Club conference also showed how to
defeat Bitcoin anonymity. Ron and Shamir's research, by contrast, analyzes
transaction history to answer questions about how typical account holders use
the Bitcoin system and how they acquire and spend the digital coin.

In addition to discovering that the vast majority of BTCs don't circulate,
the new paper also found that almost all large Bitcoin transactions stem from
a single payment made almost two years ago. They made that finding despite
apparent attempts by the account holders involved to obfuscate their actions.
The researchers started by mining the history for data that identified when
two or more addresses belonged to the same owner. They then isolated
transactions involving 50,000 BTCs or more and analyzed when the amounts were
accumulated and spent.

"We discovered that almost all these large transactions were the descendants
of a single large transaction involving 90,000 Bitcoins which took place on
November 8th 2010, and that the subgraph of these transactions contains many
strange looking chains and fork-merge structures, in which a large balance is
either transferred within a few hours through hundreds of temporary
intermediate accounts, or split into many small amounts which are sent to
different accounts only in order to be recombined shortly afterwards into
essentially the same amount in a new account.

The discovery that so little of the currency is actually in circulation could
have implications for the future of the Bitcoin system. It's unclear what
might happen if large amounts of those reserves were to begin trading
suddenly. It wouldn't be surprising to see such an influx cause a decline in
the value of each coin, similar to a 90-percent decline experienced last
year.